Lumos: An Open-Source Device For Wearable Spectroscopy Research

Lumos: An Open-Source Device for Wearable Spectroscopy Research
AMANDA WATSON∗ and CLAIRE KENDELL∗ , University of Pennsylvania, USA

ANUSH LINGAMOORTHY, Drexel University, USA
INSUP LEE, University of Pennsylvania, USA
JAMES WEIMER, Vanderbilt University, USA
Fig. 1. Lumos is a wearable optical spectrometer that enables non-invasive health monitoring in the real-world. Left: the
wristband form factor is familiar and can be worn when moving around in the real-world. Right: the fingertip form factor is
more suited to clinic settings and discrete measurements.
Spectroscopy, the study of the interaction between electromagnetic radiation and matter, is a vital technique in many
disciplines. This technique is limited to lab settings, and, as such, sensing is isolated and infrequent. Thus, it can only provide
a brief snapshot of the monitored parameter. Wearable technology brings sensing and tracking technologies out into everyday
life, creating longitudinal datasets that provide more insight into the monitored parameter. In this paper, we describe Lumos,
an open-source device for wearable spectroscopy research. Lumos can facilitate on-body spectroscopy research in health
monitoring, athletics, rehabilitation, and more. We developed an algorithm to determine the spectral response of a medium
with a mean absolute error of 13nm. From this, researchers can determine the optimal spectrum and create customized sensors
for their target application. We show the utility of Lumos in a pilot study, sensing of prediabetes, where we determine the
relevant spectrum for glucose and create and evaluate a targeted tracking device.
CCS Concepts: • Human-centered computing → Ubiquitous and mobile devices.
Additional Key Words and Phrases: spectroscopy, wearable technology, health tracking
∗ Both authors contributed equally to this research. 187

Authors’ addresses: Amanda Watson, aawatson@seas.upenn.edu; Claire Kendell, ckendell@seas.upenn.edu, University of Pennsylvania, USA;
Anush Lingamoorthy, Drexel University, USA, aln57@drexel.edu; Insup Lee, University of Pennsylvania, USA, lee@seas.upenn.edu; James
Weimer, Vanderbilt University, USA, james.weimer@vanderbilt.edu.
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that
copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first
page. Copyrights for components of this work owned by others than ACM must be honored. Abstracting with credit is permitted. To copy
otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from
permissions@acm.org.
© 2022 Association for Computing Machinery.
2474-9567/2022/12-ART187 $15.00
https://doi.org/10.1145/3569502
Proc. ACM Interact. Mob. Wearable Ubiquitous Technol., Vol. 6, No. 4, Article 187. Publication date: December 2022.
187:2 • Watson and Kendell, et al.
ACM Reference Format:

Amanda Watson, Claire Kendell, Anush Lingamoorthy, Insup Lee, and James Weimer. 2022. Lumos: An Open-Source Device
for Wearable Spectroscopy Research. Proc. ACM Interact. Mob. Wearable Ubiquitous Technol. 6, 4, Article 187 (December 2022),
24 pages. https://doi.org/10.1145/3569502
1 INTRODUCTION
Spectroscopy, the study of the interaction between electromagnetic radiation and matter, is a vital technique in
many disciplines. In medicine, spectroscopy has facilitated substantial advancements in diagnostic technology.
Developments in spectroscopy have brought about biomedical imaging technology, including X-ray, MR spec-
troscopy, and CT scan technology. Additionally, analysis of biological samples relies heavily on spectroscopic
techniques to measure the biomarkers that track the progression of diseases. In its current instantiation, its utility
is limited because it only provides a brief snapshot of the monitored parameter. Data in isolation, not in a time
series, does not provide the trends clinicians rely upon to make informed decisions.
Wearable technology brings sensing and tracking technologies to the masses. Integrating this technology into
our daily lives has collected a wealth of previously untapped data. Capturing this amount of data has created
longitudinal datasets ripe for analysis. These datasets present opportunities for the application of machine
learning and data analytic techniques. This data, combined with machine learning and artificial intelligence,
can support clinical decisions, opening the door to more accurate diagnoses. More so, it has transformed how
people live, impacting their daily routines, interactions with others, and health monitoring. These qualities make
wearable technology the perfect vehicle to collect continuous data from spectroscopy devices.
The traditional benchtop spectrometer is already being minaturized [7, 8, 40, 77]. Sampling and testing are
being done in the lab, clinic, and field via portable spectrometers [23, 27, 56, 67]. Optical sensing [17, 58]
has demonstrated the viability of a wearable spectral device but is limited by its sensing spectrum. Providing
researchers, clinicians, scientists, and even the general public with a more comprehensive wearable spectrometer
will increase the amount of meaningful data collected as well as lead to novel clinical applications. Compared to
prior work, our approach creates a ready-to-use research platform that can be utilized to develop these novel
applications and longitudinal datasets.
In this paper, we address the following research questions:
RQ1: How can we redesign the traditional benchtop spectrometer while accounting for the many constraints
of a wearable device including battery life, computing constraints, and overall footprint?
RQ2: How can we determine the spectral response of a medium while adapting to dynamic environments
given a limited set of light emitting diodes and photodiodes?
RQ3: How can this device be leveraged to address research challenges in target applications?
The potential benefits of integrating spectroscopy into wearable technology are substantial. In this work,
we present Lumos, a wearable spectroscopy device that enables noninvasive, real-time, and continuous health
monitoring. To make Lumos available to the research community, we open-sourced the hardware designs and
algorithms used in this work [72]. This includes the components, schematics for the circuitry, Gerber files for
printing the PCB, and code. We provide the CAD files for the 3D printed form factors, wristband and finger
clamp, to house the circuitry. Lumos uses off-the-shelf components that can be combined in the manner we
describe in this paper for research purposes or be tailored to fit a targeted application. We evaluated our device
on its accuracy when compared with a traditional spectrometer, its power consumption, and its reactions to
temperature and fluid.
We developed an algorithm to detect the spectral response of a medium. This determines the optimal spectrum
for further evaluation when creating a wearable spectroscopy device for a target application. This algorithm
seeks to construct the spectral response of only the medium. It adapts to light leakage from the environment
Lumos: An Open-Source Device for Wearable Spectroscopy Research • 187:3
providing maximal invariance to environmental disturbances. It adjusts its light intensity to standardize readings
across skin tone, change in pressure applied to the device, blood perfusion, etc. Additionally, we estimate the
theoretical response of our device without a medium to remove the bias from our readings. We evaluate this
algorithm on its accuracy, resolution, and resistance to outside sources of noise.
Lumos is an open source device that facilitates wearable spectroscopy research in target applications. To
demonstrate its utility in a target application, we conducted a pilot study. In this pilot study, we used the Lumos
device in a medical application: prediabetes monitoring. First, we monitored blood glucose via a glucometer and
compared it to readings from our device. We used the spectral response estimation algorithm to determine the
optimal wavelengths for tracking glucose. Then we created a customized device tailored to those wavelengths
and demonstrated that the device could be used to track changes in glucose. While we are not currently able
to accurately calculate discrete glucose readings, we can determine the change in glucose over time. This is
sufficient to track the progression of prediabetes for our target application.
Specifically, our contributions are summarized as follows:
(1) An open-source device to perform wearable optical spectroscopy research. This device is designed to get
researchers up and running quickly in new studies and be easily tailored for custom applications.
(2) An algorithm to estimate the spectral response of a medium to determine the optimal spectrum for a given
application while providing maximal invariance to environmental disturbances.
(3) A pilot study, prediabetes monitoring, to demonstrate the viability of using Lumos for a target applications.
The remainder of our paper is structured as follows: Section 2 summarizes the work related to this paper. In
Section 3, we introduce the Lumos hardware, including the components used, its form factors, and the underlying
theory that supports its operation. Section 4 describes the wavelength algorithm. In Section 5, we evaluate the
hardware and our algorithm. Section 6 details our pilot study. Finally, we provide discussion and the future work
of this paper in Section 7, and wrap up with our conclusion.
2 RELATED WORK
In this section, we introduce spectroscopy and present how it has been traditionally used as well as how it is
evolving. Then, we discuss wearable spectral and optical sensing that has brought heart rate, blood oxygenation,
and pulse oximetry to the masses. Finally, we examine applications of wearable spectroscopy that have been
studied in the research lab.
2.1 Spectroscopy
Spectroscopy studies the interaction between electromagnetic radiation and matter. It is a vital technique in
many disciplines, including medical imaging, molecular analysis, and remote astronomical sensing. The most
common types of spectroscopy include atomic spectroscopy [32], ultraviolet and visible spectroscopy [76], infrared
spectroscopy [75], Raman spectroscopy [75], and nuclear magnetic resonance [25]. Traditionally, spectrometry,
the measurement of the interaction between electromagnetic radiation and matter, is done on benchtop requiring
large, heavy, and complex machines with operators that have been specially trained. These benchtop services
are characterized by the precision and accuracy of their measurements, but due to the complicated processes,
procedures can be time-consuming and inconvenient. Furthermore, these techniques cannot be directly applied
to living organisms as normal physiologic functions must be maintained. Hence, samples are obtained from
living organisms, but this is not an entirely benign process as it typically requires invasive procedures that cause
subject discomfort.
Work has already begun to miniaturize the traditional benchtop spectrometer [7, 8, 40, 77]. Many companies [23,
27, 56, 67] have developed portable spectrometers to bring testing directly into the lab, clinic, and field. Despite
this, there is still a need to track real-world, longitudinal continuous data on patients. For example, the most
common issue with pain management protocols is the insufficient treatment of pain with opioids. The reasons
for insufficient pain management can be caused by one or more of the following: intolerance of side effects from
the medications, desire to minimize medication intake, cost of medication, and external sources of influence [12].
Current techniques for monitoring adherence involve lengthy questionnaires and invasive testing. Whereas
a wearable spectrometer could solve this adherence problem by continuously monitoring patient opioid use.
Additionally, it could help identify drug diversion, a major concern for clinicians and the Drug Enforcement Agency
(DEA) [44]. Further, opioid management during in-patient hospice care is an exceptional clinical challenge [63].
The goal is to provide sufficient analgesia for conditions that can be particularly painful. Balanced against this
is the desire for a patient to maintain normal cognitive function. Additionally, during hospice care, the goal is
to provide the best quality care while minimizing invasive techniques. Rather than drawing blood samples to
monitor a patient’s opioid levels, a non-invasive monitor would greatly benefit both the patient and clinician. In
addition to the above examples, alcohol, ketones [30], blood pH [4], glucose [18], as well as responsiveness to
medical treatment and medications [68] could all benefit from being monitored in a continuous, non-invasive,
outpatient manner.
2.2 Other Approaches to Spectral/Optical Sensing

Spectral and optical wearable sensors allow for the continuous sensing of various biomarkers. Some biomarkers
are directly measured through the skin by placing an optical sensor on a subject’s wrist and sampling the response
in a portion of the electromagnetic spectrum. Other biomarkers are more complicated and, in their current state,
require the collection of a biofluid before analysis can be completed. The basics of wearable optical sensing
involve a light source to introduce light into the body through the skin and a sensor to detect changes in the
light due to absorbance and scattering. The most widely available wearable optical sensing devices track a
wearer’s heart rate and blood oxygenation [17, 58]. Heart rate is obtained by utilizing a single light source and a
photodetector to measure small changes in blood volume in the capillary layer. Pulse oximetry leverages a dual
light source and photodetector to sense variations in the optical characteristics of hemoglobin during oxygenation
and deoxygenation [33, 46, 54, 62]. Additionally, researchers have begun the development of wearable sensors
that detect various ions in sweat using a UV light source with a photodetector [3, 6, 48]. This allows for the
tracking of ions such as sodium, potassium, and heavy metals, which, when imbalanced, can lead to impaired
organ system function. More complex wearable optical sensors take the basics and add a chemical reaction.
Optical sensing [31, 43, 69], a sensing method that detects light intensity, has brought progress to the wearable
detection and monitoring of pH, electrolytes [39, 43], and glucose[21, 28, 65]. Hyperspectral Imaging is a related
form of sensing that extends beyond typical the typical bands of red, green, blue, and has been used in various
applications, including HyperCam, to detect spectral variance [22].
2.3 Applications of Wearable Spectroscopy

In research, many forms of spectroscopy have been explored for wearable sensing [75]. Most importantly, near-
infrared and functional near-infrared spectroscopy has been utilized for physiologic measurements [9, 74, 75].
Near-infrared spectroscopy (NIR) utilizes the wavelength range from 750-2500 nm of the electromagnetic spectrum
and has been commonly used to measure oxygenated and deoxygenated hemoglobin through the skin [75].
Functional near-infrared spectroscopy (fNIRS) uses a subset of the NIR spectrum to sense the oxygenation of
hemoglobin [16, 36]. Diffuse optical technologies is a model-based technique for NIR measurement done via
continuous-wave, frequency-domain, and time-domain spectroscopy [11, 34, 60, 68]. This technique has been
used to measure the concentrations and change of hemoglobin oxygenation, water, and lipids [68]. Here, we will
present some applications that are becoming possible due to wearable spectroscopy.
Non-invasive blood sugar monitoring has been an elusive objective. Historically, a blood sample is required to
detect the concentration of glucose. This is a painful process that creates a barrier to consistent and effective blood
sugar monitoring. NIR spectroscopy, a non-invasive modality, has shown potential for detecting glucose [50]. To
date, attempts are complicated and plagued by obstacles such as interfering absorption [75] and motion artifacts.
Through our research, we have demonstrated the ability to noninvasively and continuously monitor changes in
glucose in limited circumstances.
An increased concentration of lactate indicates insufficient blood flow or excessive metabolism. This can
have many causes and even more numerous complications. One example is the end organ damage caused
by sepsis. Traditional spectroscopy is excellent at detecting the concentration of lactate in a blood sample.
Unfortunately, ascertaining the clinical relevance of a single data point can be challenging as it lacks context.
Research into a wearable lactate sensor has shown that the NIR region holds promise for detecting changes in
lactate concentrations. Budidha et al. [9] showed that changes of 1 mmol/L of lactate in blood analogous solutions
are detectable.
The most studied wearable spectroscopy devices use functional near-infrared spectroscopy to detect brain
activity. When a region of the brain is more metabolically active, it has increased oxygen requirements. This is
detected and provides use in brain-computer interfaces [74]. Diffuse correlation spectroscopy and speckle contrast
spectroscopy are applications of functional near-infrared spectroscopy that have been used to measure deep
tissue blood flow [16]. Continuous wave diffuse optical imaging implemented in a wearable device has allowed
for the measurement of oxygenation changes in breast tumors during chemotherapy. This device uses six LED
and photodiode pairs [68]. Theoretically, this can provide clinicians with the ability to assess the effectiveness of
their treatment and modify it in real-time.
Wearable technology provides the ability to simultaneously attach sensors to many areas of the body. This
allows for the capture of a diverse set of physiologic characteristics in real-time. Further, it promotes research
outside of the lab for situations in which it is unreasonable to use traditional benchtop methods. Spectroscopy is a
powerful tool that is essential for analyzing and identifying human disorders. Thus far, it has been an underused
modality limited to providing transient snapshots of data. Advances in wearable technology have allowed for the
application of spectroscopy techniques to health sensing. Moving forward, wearable spectroscopy provides the
opportunity to continuously track the progression of disease or even health of a subject. It creates an opportunity
to collect massive amounts of high-value longitudinal data. That amount of data could be used to gain remarkable
insights into human health and performance.
3 LUMOS HARDWARE
In this section, we discuss the hardware that makes up the Lumos device. We first examine the theory of
spectroscopy and its applications in the wearable domain. Then, we describe the two 3D printed form factors
created to house our device. Finally, we discuss the off-the-shelf components and circuitry used to create our
device.
3.1 Theory of Operation

Spectroscopy, as previously stated, is the study of the interaction between electromagnetic radiation and matter.
It has two main interactions: absorption and emission of photons, particles of light. The measurement of the
the photons gives information about the medium [70]. Absorption is when electromagnetic radiation, such as
light, is absorbed via a change in energy as it passes through a medium [55]. The wavelengths that are absorbed
or partially absorbed would present a lower number of photons at the detector, while the wavelengths that
were transmitted would present a higher number of photons. This is depicted in Figure 2a. The figure displays a
multicolored light shining on and interacting with a medium. That medium absorbs a portion of the spectrum,
(a) Absorption Spectroscopy (b) Emission Spectroscopy
Fig. 2. Theory of Spectroscopy
yellow wavelengths. The remaining light is read by the sensor on the opposite side of the medium, showing lower
counts at the sensors near the yellow wavelengths. Emission is when the electromagnetic radiation applied to a
medium results in the release of photons due to absorption of photons that were applied [42]. This is depicted in
Figure 2b. The figure displays a multicolored light with various center wavelengths shining on and interacting
with a medium. This medium emits photons which are visible to the human eye as green light when illuminated
by the multicolored light. This release, as well as any transmitted light, is captured by the detector. The emission
of photons must obey the conservation of energy. Thus, the emmision would not be greater than the energy
absorbed. These measurements grant us insight into the medium itself. Mediums can have both absorption and
emission occurring at once. Thus, we use this theory as a basis for our sensor and algorithm design in this paper.
Traditionally, optical spectrometers measure the photons either transmitted through a medium or reflected
from a medium. Transmission-based spectrometers place the sensor on the opposite side of the medium from
the light source. Reflectance-based spectrometers place the sensor and light source on the same side of the
medium. Each method has its advantages and disadvantages. To discuss the differences between the two in a
wearable configuration, we will use pulse oximetry as an example. Pulse oximetry is a non-invasive application of
spectroscopy to monitor oxygen saturation. Studies have shown both transmission and reflection-based devices
can provide accurate oxygen saturation measurements [45]. Transmission-based devices, such as the finger clamp
commonly used in medicine, are best used in thin areas where the entirety of the medium can be perfused with
light. While this limits the on-body locations in which the sensor can be placed and the wearers’ movement,
this style of sensor is trusted in medical settings. Reflection-based pulse oximetry is generally housed in a
smartwatch. It is useful when light cannot fully travel through the medium to show an optical response on the
other side [14]. The light source and sensor are placed side by side for reflection-based measurements, which
gives more flexibility for on-body locations of measurement [14, 46]. However, light source and sensor alignment
are more complicated with this method. As such, a small change in the angle of a surface mount light emitting
diode (LED) can drastically change a measurement. As both methods have pros and cons, we seek to incorporate
both into our designs. This will require additional data processing as well as separate form factors.
When designing Lumos, we pursued designs that would measure absorption and emission as well as measure
photons through transmission and reflectance. Next, we will discuss the form factors we created to make this
possible.
(a) Finger Clamp Form Factor (b) Watch Form Factor
Fig. 3. Lumos Device Form Factors
3.2 Form Factors

As we developed Lumos, we pursued designs already commonly used in wearable technology. This led us to
two form factors, a wristband form factor with a similar design to a smartwatch and a finger clamp form factor
with a similar design to a pulse oximeter. These designs are shown in Figure 3. The finger clamp leverages
transmission spectroscopy and places the light source on the opposite side of the medium from the detector. The
wristband leverages reflectance spectroscopy by placing the light source and photodetector on the same side.
Both the wristband and finger clamp allow for emission and absorption spectroscopy based on the spectrum
input. These form factors offer users a more comfortable experience when using our devices. Furthermore, as
they are mainly 3D printed, with a few additional easy-to-find components, they do not require a large time
or resource investment. The 3D print schematics are available on our Github [72] with links to the auxiliary
components such as springs and pins that were used.
3.2.1 Finger Clamp. The first form factor is the finger clamp, and it is shown in Figure 3a. It is similar to
commercially available pulse oximeters in that it rests on the distal phalanges of the finger. This form factor
was chosen because it is comfortable, easy to use, and generally familiar to people. It leverages transmission
spectroscopy in which light shines on one side of a medium, with the detector on the other side reading the
photons from absorption or emission. The finger clamp consists of the 3D printed housing, a spring to keep
the housing closed, and a bevel pin to open and close around. When designing this form factor, we addressed
two main challenges: consistent pressure from the clamp without causing discomfort and light leakage from
the environment. Consistent pressure is needed because as pressure varies, the measurements also fluctuate. So,
we introduced a spring and pin to standardize the amount of pressure applied while remaining comfortable to
the wearer. Light leakage is the environmental light sensed by the detector. If too much light leaks in, it can
obscure subtle changes in our measurements. To address this, we closed off the sides of our form factor as shown
in Figure 3a. Additionally, through hole or surface mount LEDs can be used in this form factor.
3.2.2 Wristband. The second form factor is a wristband, as shown in Figure 3b. This is similar to a smartwatch in
how it rests on the wrist. It has the advantages of being familiar to the user, comfortable, and can be worn all day.
The wristband form factor is based on reflectance spectrometry, so the light source and detector are on the same
side of the medium. Because of this, we are less limited in the locations this sensor can be placed. For example, if
we extend the straps, it could be attached to the upper arm or leg. This form factor consists of the 3D printed
housing, straps, and a skin-safe encapsulant. The skin-safe encapsulant separates the electronic components from
the skin, which is essential in a wearable device. It is a transparent silicon that provides a moisture barrier from
any sweat on the skin. The entire assembly protrudes from the bottom of the wrist band to ensure tight contact
with the skin. Additionally, this encapsulant allows for the device to be worn tightly on the skin, decreasing the
amount of light leakage and standardizing the amount of pressure applied.
LEDs Photodetector
415nm
12 - Channels
450nm
ADC
480nm
515nm
I2C Com
LED 555nm MCU
DRIVER
590nm
630nm
660nm
910nm
Sensor
Fig. 4. Block diagram of Lumos Device. The surface mount LEDs generate light and the photodetectors read the light through
a medium. These readings are sent to the controller board, where the ADC samples the data and communicates it to the
collection system.
3.3 Components
Both form factors described above are based on the same electrical components; they are housed in different
configurations. At a high level, the Lumos device consists of a light source and a detector. The light source
we describe includes commercially available surface mount LEDs that cover the entire visual spectrum and
LED drivers to control the intensity of the light. The LEDs are configurable based on the spectral needs of
the application. The detector consists of the sensor with the spectral sensor, microcontroller, communication
component, and battery. To house these electronics, we developed a small custom printed circuit board (30mm x
30mm x1.6mm) to ensure a small footprint to fit into our form factors. Figure 4 shows the component diagram
for our device.
3.3.1 Light Source. The goal of the light source is to cover the targeted spectrum and appropriately illuminate
the medium. This is accomplished through an array of surface mount LEDs and LED drivers. In the finger clamp,
the light source sits on top of the form factor. In the smartwatch, it is integrated into the same printed circuit
board (PCB) as the photodetectors.
1.0 1.0
415 415
0.8 445 0.8 445
Normalized Intensity
Normalized Intensity
470 480
515 515
0.6 0.6
555 555
590 590
0.4 0.4
630 630
680 680
0.2 0.2
0.0 0.0
400 500 600 700 800 400 500 600 700 800
Wavelength (nm) Wavelength (nm)
(a) LED Wavelengths (b) PD Wavelengths
Fig. 5. Spectral Response of our LEDs and PDs
To build a comprehensive system, in this paper, we selected an LED array that covered the visual spectrum.
Each LED [53] produces light in a continuous spectrum centered around one peak wavelength. We have eight
LEDs in the visual spectrum with center wavelengths around 415nm, 450nm, 470nm, 530nm, 568nm, 599nm,
633nm, and 660nm. The LEDs do have spectral ranges that overlap. This is shown in Figure 5a. In this figure, the
total spectrum covered is shown as well as the overlap between LEDs. To handle the complexity that the overlaps
may cause, we developed an algorithm that we will discuss in detail in Section 4.
To appropriately perfuse the medium, we use LED drivers to tailor the intensity of each LED. This is important
so that the intensity of the LED is high enough for a meaningful measurement but not too high that it will
oversaturate the photodiode. The LED drivers also allow us to have consistent readings between all LEDs. We
will provide more information on the intensity calibration algorithm that we use to tailor the intensity of each
LED in Section 4.
3.3.2 Detector. The detector senses the response from the light source applied to the medium. It consists of
the spectral sensor, microcontroller, and communication component. The spectral sensor is integrated into our
custom PCB. The microcontroller and communication component are housed on a separate board.
We chose the A7341 spectral sensor [2] because it has a large sensing range and can easily be integrated into a
wearable device. It has eight channels in the visible range, with peaks centered around 415nm, 445nm, 480nm,
515nm, 555nm, 590nm, 630nm, and 680nm. The sensing channels are displayed in Figure 5b. The dimensions of
this device are 3.1mm x 2mm x 1mm. Thus, it fits into our form factors nicely, as shown in Figure 3a and Figure
3b. The AS7341 needs 1.8VDD for operation. The max current draw from the AS7341 is 300µA making it low
power. We will evaluate the power consumption as well as battery life in Section 5.
We chose the Arduino Nano 33 IoT to power our AS7341 and communicate data from the AS7341 to the data
collection system. The Arduino Nano 33 IoT has a Cortex M0+ SAMD21 processor and NINA-W102-00B WiFi/BLE
Module. It has the ability to send data out over WiFi using MQTT to the data collection platform [73] as it
eliminated our need for an additional device such as a smartphone to route the data to the data collection system.
The Arduino and the entire device are powered by a 400 mAh lithium-ion battery. This gives it approximately
five hours of battery life with an evaluation of this in Section 5. The Arduino supplies 3.3V, which is more than
needed for our spectral sensor. As such, there is a voltage regulator on the AS7341 so the Arduino can safely
power the spectral sensor. In future iterations of this prototype, we will be looking to shrink this chip to be more
compact and tailored to our device, as this Arduino is excessive. This will lead to reduced power consumption
and overall space footprint.
4 SPECTRAL RESPONSE OF MEDIUM

In this section, we detail the algorithm used to determine the spectral response of a medium. This allows
us to target that spectral response for a specific application, increasing battery life, decreasing computation
requirements, and reducing the overall footprint. First, we need to calibrate the intensity of our LEDs to account
for environmental light, skin tone, change in pressure applied against the device, and blood perfusion. Second,
we eliminate environmental light leakage. Third, we remove the reading we theoretically expect to see if there
was no medium. Finally, we identify the spectrum of interest we should target moving forward.
4.1 Intensity Calibration

LED light intensity is affected by many factors, including environmental light, skin tone, change in pressure
applied against the device, and blood perfusion. To remove the impact of these factors, each LED’s intensity is
calibrated for each person and environment. This one-time calibration process occurs when the user first puts it
on. It takes approximately 2-5 seconds to calibrate each time. The calibration process is necessary to mitigate
over and under saturation on the detector side. Overall, the calibration process adjusts the light intensity until
a targeted reading is obtained at the photodetector. We calibrate each LED’s brightness on the photodetector
nearest to its wavelength. For example, the 470nm LED will be calibrated with the 480nm PD.
Our photodetectors have a 16-bit resolution giving us a maximum value of 65,535. To leave room to sense
both absorption and emission, we choose a target value of approximately 23 of our maximum reading or between
42,000 counts and 44,000 counts. This gives our sensor a higher sensitivity while leaving ample room for emission
readings. This can be adjusted based on the needs of the application. The iterative calibration process is performed
by adjusting the current to the LED using an 8-bit LED driver that adjusts the light intensity until it reaches our
targeted value. This adjusts the LED with a resolution of 0-255, approximately 0-50mA. At each iteration (𝑖), we
calculate the error between the targeted reading (𝑟𝑡 ) and the current reading (𝑟𝑖 ). A proportional controller is
then used to determine the adjustment required for each LED to achieve the desired count value. A proportional
gain (𝐾𝑝 ) is used to determine the ratio of output response to the error signal. At each iteration, we calculate the
intensity (𝐼 ) with the following equation
𝐼𝑖+1 = 𝐾𝑝 (𝑟𝑡 − 𝑟𝑖 ) (1)
Then after the new intensity is calculated, we adjust the LED and repeat the process until the targeted reading
is reached. In general, the entirety of this iterative process occurs in approximately 2-15 iterations or 3-5 seconds.
We evaluate the accuracy and time needed to run the intensity calibration in Section 5.
4.2 Environmental Light Leakage

The intensity calibration accounts for the initial environmental light leakage, but as environments change, so
does the light leakage. To handle this dynamically, we sample our sensor in two LED states: ON and OFF. When
the LED is ON, we read the combination of the spectral response of the medium and the environmental light
leakage. When the LED is OFF, we only read the environmental light leakage. The time spent in an ON state
is determined by each LED’s rise time and the integration time of each of our PDs. The rise time is the time it
takes for the LED to reach maximum intensity based on the current calculated in the intensity calibration. The
integration time is the amount required for each PD to record a reading. The rise time and integration time are
gathered from the datasheets. Then, to eliminate the environmental light leakage from our reading, we take the
average of the two OFF states that surround an ON state and subtract that from the ON state. This is done for
every PD for each LED.
4.3 Theoretical Approximation

To determine the spectral response of a medium, we compare the readings from Lumos to the expected readings
if no medium was present. To do this, we approximate the readings from Lumos if there is no medium. This
requires an approximation for each combination of LED and PD. To approximate each combination, we first
estimate the spectral response of each component as a Gaussian. Second, we calculate the expected spectral
response of each combination. Finally, we adjust our theoretical approximation to account for leakage current.
4.3.1 Gaussian Estimation. Theoretically, the spectral response of our LEDs and PDs are approximately a
Gaussian [47]. To estimate the Gaussian, we need three metrics: intensity, center wavelength, and full width
half maximum. The intensity (𝐼 ) is gathered from the intensity calibration step described above. The Gaussian
estimation will be scaled to the calibrated intensity. The center wavelength and full width half maximum are
available from the datasheets accompanying each LED and PD. The center wavelength (𝐶𝑊 ), is the wavelength
at which the LED or PD is at maximum intensity. The full width half maximum (𝐹𝑊 𝐻 𝑀), is the width of the
spectrum at half of√the full intensity. When calculating the Gaussian, the 𝐹𝑊 𝐻 𝑀 relates to the standard deviation
𝜎 by 𝐹𝑊 𝐻 𝑀 = 2𝜎 2 ln 2. Then, we estimate the spectral response (𝑆𝑅) via the following equation:
(𝑥 − 𝐶𝑊 ) 2

𝑆𝑅 = 𝐼 e exp − 𝐹𝑊 𝐻 𝑀 (2)
2( 2.35 ) 2
The estimated curves for all LEDs are shown in Figure 5a and for all PDs are shown in Figure 5b.
Fig. 6. Calculation of Theoretical Approximation for 530nm (Green) LED
4.3.2 Spectral Response Calculation. Next, we calculate the expected response for each combination of LED and
PD. To do this, for each LED, we overlay each PD. In Figure 6, we show this example for the 515nm LED. This
LED is shown with each PD, giving us eight graphs. When the LED and PD overlap, we expect a spectral response
from the Lumos device. In Figure 6, we see that the LED overlaps with the 445nm, 480nm, 515nm, 555nm, 590nm,
and 630nm PDs. For each overlap, we calculate the area under the overlap of the two curves to determine the
expected response of each PD. This is normalized to the total possible response from each PD, i.e., the area under
the PD curve. We show the estimated responses for the 515nm LED in Figure 7a. Then, we scale the estimated
response to the counts we read from the Lumos device, also shown in Figure 7a. In total, for the eight LEDs and
eight PDs, we calculated 64 expected responses.
4.3.3 Leakage Current Adjustment. Photodetectors are not perfect and a known issue is that they produce a
leakage current when any photons excite the photodetector [71]. This causes small readings in channels that are
not being excited. The readings are linearly related to the number of photons hitting the sensor, i.e., as more
photons hit the sensor, the higher the readings. We take an experimental reading without a medium to calculate
the leakage current adjustment. We look to the PD readings where the estimated response is zero and average
them. This average becomes our current leakage adjustment which is added to all theoretical estimations where
the estimated response is zero. This is shown in Figure 7a.
No Medium
40000 1.0
Leakage Cur
Relative Intensity [A.U.]
Spectral Response
4000
Scaled 0.8
30000 Theoretical 2000
0
0.6
Counts
20000 −2000
0.4 −4000
10000 680
0.2 630
590
415 555
0 450
470 515 PD
0.0 530 480
450 500 550 600 650 568 445
LED 599
Wavelength(nm) 633
660
415
(a) Scaling and Leakage Current Adjustment at 515nm LED (b) Spectral Response of No Medium
Fig. 7. Spectral Response Calculation
4.3.4 Overall Spectral Response. Finally, we calculate the spectral response of the medium by combining all the
PD readings from all of the LEDs. We show the overall response of air or no medium in Figure 7b. The overall
response is represented by the entire surface in the figure. The peaks and troughs represent wavelengths with
high responses that should be targeted for higher resolution sensing. Peaks represent absorption wavelengths,
while troughs represent emission wavelengths. In this figure, we see very little response, which is expected as it
is the spectral response of air. We evaluate various mediums in Section 5.
5 EVALUATION
In this section, we evaluate the Lumos Device and the spectral response algorithm. First, we analyze our device.
We begin by evaluating its response to temperature and fluids. Then, we examine its power requirements. Then,
we compare it to a benchtop spectrometer. Second, we explore the results of our spectral response algorithm.
We evaluate the time required for the intensity calibration. Then, we determine the accuracy of our device as
compared with mediums with known spectral responses.
5.1 Temperature Experiment

Temperature can affect sensor readings even when not outside of the normal operating range [35, 64]. To under-
stand the effect of temperature on our device, we evaluated the sensor’s response to hot and cold temperatures.
The sensor, AS7341, is guaranteed to work within the range −30◦𝐶 to 70◦𝐶. Since we do not expect our sensor to
experience the extremes of this range when worn on the body, we evaluate our device between 0◦𝐶 and 45◦𝐶.
These temperatures were selected as they are on the edges of normal environmental conditions [59]. For any
wearable device, it is important that it withstand normal environmental conditions that humans experience.
26000
Spectral Response (counts)

Spectral Response(Counts)
30000
25500
Lumos 1
Lumos 1
28000 Lumos 2
25000 Lumos 2
Lumos 3
Lumos 3
Spray
24500 26000
24000 24000
−15 0 15 30 45 60 75 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
Temperature ( ∘ C) Time (minutes)
(a) Results from Temperature Experiment (b) Results from Fluid Experiment
Fig. 8. Temperature and Fluid Experiment Results
To evaluate our device in varying temperatures, we chose a narrow spectrum light source (445nm) that would
illuminate a single channel. We tracked the counts on that channel as we varied the temperature between 0◦𝐶
and 45◦𝐶. This experiment was performed with our sensor in a black box to minimize ambient light. To increase
the temperature, we used a hair dryer to blow hot air through a hole in the side of the box while not increasing
the ambient light. To decrease the temperature, we put the black box into a freezer. In both scenarios, we tracked
the temperature with a thermometer. Furthermore, we repeated this experiment across multiple devices to show
inter-device repeatability.
We display the results from the experiment in Figure 8a. We show the main channel (445nm) from three
devices. The other channels had low counts that did not have noticeable changes related to temperature. The
three devices all read the same LED but, due to their placement, received different numbers of photons. This
accounts for the difference in counts between the devices. Over the 90◦𝐶 temperature change, we saw a less
than 200 count change in our readings on any of the devices. This accounts for a less than one percent change
in the overall reading. While we do see a change in the counts, it is small enough that it should not affect the
performance of our sensor.
5.2 Fluid Experiment

In addition to temperature, fluids can also impact the sensors. For example, sweat can coat the encapsulant,
causing differences in sensor readings. To evaluate Lumos’s response, we simulated different levels of sweat.
We used our wrist band form factor with a waterproof encapsulant to do this. This was done by spraying the
encapsulant with water. We repeated this three times to simulate minimally sweaty to very sweaty, with very
sweaty being someone dripping in sweat. Each spray was approximately .75 ml. As above, we performed this
study in a black box to reduce light leakage. Figure 8b shows the results from the fluid experiment in the 445nm
wavelength. All other wavelengths followed a similar trend. These results show no significant effect, less than
100 counts, on the spectral response readings until the third spray. This puts 2.25 ml of water onto our sensing
area, well past what can be expected from sweating.
5.3 Power
We measured the power consumption of the Lumos device using a digital multimeter to measure the power drawn
when all LEDs are at max intensity, PDs are sampled at 10Hz, and our communication protocols are functioning
as normal. This is the max power draw from our device. With these settings, the Lumos device consumes 85 mA
(425mW) of current. This gives us approximately five hours of battery life with a 400 mAh lithium-ion battery.
This is more than sufficient for a normal research study. Longer studies are possible with a higher capacity
battery. The Arduino Nano IoT is the most power-hungry part of our device. Replacing this with a lighter-weight
component would increase the battery life. In practice, our system could sample data less frequently or transmit
data to the data collection system at higher intervals. A sleep functionality could also be implemented so that the
device, including the LEDs, PDs, and processing circuitry, do not need to be on when not in use. In Section 7 we
discuss opportunities to upgrade the hardware, including some that would increase battery life.
5.4 Intensity Calibration
250
200
0.1
Intensity
150 0.01
0.001
100 0.0001
50
0
0 10 20 30 40 50 60 70
Steps
Fig. 9. Intensity Calibration at Each 𝐾𝑝 for 530nm LED
To determine the best 𝐾𝑝 for our system, we tested four values, 𝐾𝑝 = [0.1, 0.01, 0.001, 0.0001]. An initial default
intensity value of 𝐼 = 255, maximum intensity, was used to start this test. We show the results for how 𝐼 changed
using the 530nm LED at each step for each 𝐾𝑝 in Figure 9. The stars show when the calibration algorithm is
successfully finished. 𝐾𝑝 = [0.1, 0.0001] did not converge to our targeted value even after running for over 100
iterations. 𝐾𝑝 = 0.01 converged the fastest with an average of 4.8 steps. 𝐾𝑝 = 0.001 converged in approximately
44 steps. We chose the setting 𝐾𝑝 = 0.01 for our device as it converged to the targeted value in the fewest number
of steps. This took approximately 2.4 seconds. We show the results from each 𝐾𝑝 in Figure 1. This setting is
adjustable based on the needs of the application. Anything above 0.1 or below 0.0001 does not converge, so we
did not test those values.
Table 1. Time and Computational Needs for Each 𝐾𝑝 s
𝑲𝒑 # Steps SD Steps Time (s)

0.1 ∞ ∞ ∞
0.01 4.8 2.9 2.4
0.001 44.4 41.3 22.2
0.0001 ∞ ∞ ∞
5.5 Comparison with Spectrometer

In order to validate the AS7341 spectral sensor, we compared it to a ground truth spectrometer, the Ocean Optics
Vis-NIR fiber optic spectrometer [52], as seen in Figure 10b. This spectrometer has a range from approximately
450nm to 1118nm with a resolution of 1.37nm. In this study, we generated colors that were similar to what our
photodetectors read using an iPad screen. We pointed the iPad at both the ground truth spectrometer and the
AS7341 and recorded the readings from both. In Figure 10a, the ground truth spectrometer readings are plotted
alongside the AS7341 sensor readings. For the most part, we saw that the AS7341 readings followed the same
curve as the ground truth spectrometer and gave confidence in moving forward with this device. We did see
more error in the 630nm and 680nm readings. We believe the higher Lumos counts in the red spectrum were due
to the PDs on our device having a larger sensing range in that spectrum.
(a) Comparison with Benchtop Spectrometer (b) Experimental Setup
Fig. 10. Spectrometer Comparison Experiment
Given the comparisons to the benchtop spectrometer, we see that our sensor does accurately detect the
wavelength and intensity of light. However, the limitations of our spectral device are evident. For this study,
we selected colors that would match the center wavelength channels of the spectral sensor. However, the color
generation for this study was based on hex codes and used a combination of wavelengths to achieve a color [66]
which is different than LEDs that are centered around one wavelength. The sensor picked up these combinations
but with much less granularity than the ground truth spectrometer. Moving forward from this study, we chose
visible LEDs that were centered around a specific wavelength [15, 49, 53, 61]. We understand this device lacks
the precision and accuracy of the benchtop spectrometer. However, this iteration of the sensor allows the user to
conveniently wear it around the finger or wrist. Additionally, sensing granularity and accuracy can be improved
with additional customized hardware which will be discussed in Section 7. The goal of the device was to use
off-the-shelf components to create a wearable sensor that can easily be tailored for research or personal use. We
continued with this sensor for our prototype since it met these requirements.
415nm LED 450nm LED 470nm LED 530nm LED

Spectral Response (Counts)
Exp Exp Exp Exp

40000
Theory Theory Theory Theory
30000
20000
10000
568nm LED 599nm LED 633nm LED 660nm LED

Spectral Response (Counts)
Exp Exp Exp Exp

40000
Theory Theory Theory Theory
30000
20000
10000
0
5
5
5
5
5
5
5
5
0
41
44
48
51
55
59
63
68
41
44
48
51
55
59
63
68
41
44
48
51
55
59
63
68
41
44
48
51
55
59
63
68
Wavelength(nm) Wavelength(nm) Wavelength(nm) Wavelength(nm)
Fig. 11. Comparison of Experimental Measurements and Theoretical Approximation for Each LED
5.6 Theoretical Estimation

We evaluated our theoretical approximation by comparing them to experimental measurements. To record the
experimental measurements, we placed the Lumos device in a black box. There were openings in the box for the
wires with mitigations to remove all environmental light. We took measurements when the LED was off, and
all channels showed zero counts. To have a consistent intensity across all of the LEDs, we used our intensity
calibration algorithm to assure 44,000 counts. No medium was placed between the spectral sensor and LEDs.
Figure 11 shows the comparison between the theoretical approximations and experimental measurements at each
LED. The theoretical approximations have a 0.944 Pearson correlation with the experimental measurements.
5.7 Spectral Response of Medium

We conducted experiments to understand the Lumos device’s ability to identify the spectral response of a medium.
We started by characterizing the source and detector output with no medium and then moved on to characterize
the change in output response when a medium was placed between the source and detector. We consider no
medium to be air. The physical mediums we used were six colored light filters: purple, blue, green, yellow, orange,
and red. The mediums were placed into the fingertip form factor for this experiment. In order to characterize
these, we collected data to show the response of each of our LEDs when measured by all visible photodiode
channels for each filter, including with no filter. The algorithm discussed in Section 4 is applied to calculate
the spectral response of the mediums or lack thereof. We compare this to the known spectral response of the
mediums that we collected from the datasheets.
This experimental data collection was done in a controlled lab setting. The AS7341 spectral sensor was placed
in a black box with an LED. The ambient light was mitigated as much as possible. When we took readings with
the LED off, all channels showed zero counts. The results of this study are shown in Figure 12. The results from
the air medium are shown in Figure 7b. We compared the peaks in these figures to the center wavelengths of the
Purple Blue Green
Spectral Response
Spectral Response
Spectral Response
15000 40000 40000
35000
10000 30000 30000
25000
5000 20000 20000
15000
0 10000 10000
−5000 5000
0
0
680
650 650
630
600 590 600
550 415 555 550
450 450 515 PD 450
500 500 PD 470
500 500 PD
530 480
550 568 550
450 445 450
LED 600 LED 599
633 415 LED 600
650 660 650
Yellow Orange Red

Spectral Response
Spectral Response
Spectral Response
30000 20000 25000
15000 20000
20000
10000
15000
10000 5000
10000
0
5000
0 −5000
−10000 0
−10000 −15000 −5000
680 680 680

630 630 630
590 590 590
415 555 415 555 415 555
450
470 515 PD 450
470 515 PD 450
470 515 PD
530 480 530 480 530 480
568 445 568 445 568 445
LED 599
633 415
LED 599
633 415
LED 599
633 415
660 660 660
Fig. 12. Spectral Response of Mediums
filters, demonstrating a mean absolute error of 13nm with a standard deviation of 8nm. We further compared this
to peak detection before interpolation and showed a mean absolute error of 17nm with a standard deviation of
11nm. Thus, we see that interpolation improves our center wavelength detection. In some cases, the filters, such
as yellow, covered the entire spectrum after the initial peak, creating an extended peak. We see our algorithm
detect this, but as we do not have a center wavelength to compare against, we removed it from our analysis.
6 PILOT STUDY
Lumos can be used in many different application scenarios, including but not limited to health monitoring,
athletics, and rehabilitation. This section discusses a pilot study that demonstrates this, specifically prediabetes
tracking. We start by motivating our study and describing the current standard of care in diabetes monitoring.
Then, we describe how Lumos facilitated our sensor design and creation. Finally, we evaluate our device and
system for accuracy.
6.1 Background and Motivation

Diabetes is a disease that impacts over 400 million people worldwide [51]. It causes a number of complications,
including shortened life expectancy, decreased quality of life, blindness, cardiovascular issues, and more. In the
United States, over $300 billion yearly is spent to manage and treat diabetes and its numerous complications.
Prediabetes is a common precursor to diabetes that is characterized by higher than normal blood sugar levels
but lower blood sugar levels than with diabetes [41]. It is possible to treat and even reverse prediabetes through
lifestyle changes, but most people with prediabetes are unaware of their condition. To determine if a person
(a) Glucose Form Factor Design (b) Glucose Monitor Prototype
Fig. 13. Glucose Device Design and Initial Prototype
has prediabetes, blood glucose tests are run in the clinic. This process is invasive as it requires drawing blood
for accurate measurements. Moreover, it is time and resource intensive creating a barrier to prompt testing. By
tracking blood glucose via wearable technology, as well as its overall trends, we can potentially monitor the
progression of prediabetes outside of the clinic.
Traditional glucose monitoring requires a blood sample and a glucometer. These monitors have been tested,
evaluated, and proven effective for diabetes monitoring and management. New devices have come on the market
to try to less invasively monitor blood glucose. Continuous Glucose Monitors (CGM) collect interstitial fluid
through a needle inserted into the skin to provide continuous glucose monitoring [51]. Breath Ketone Analyzers
estimate glucose levels by measuring the ketones in your breath. Research has begun into less invasive optical
methods [1, 24, 29] in the visible [38] and near-infrared range [75] that have been done in lab settings and
wearable devices. Specifically, Lepore et al. [38] shows the fluorescence emission of glucose oxidase peaks at
around 520nm and increases in the presence of glucose in a lab setting. Dantu et al. [13] found that the ratio of
the 510nm wavelength to 475nm wavelength to be an effective metric for tracking glucose. The found that as
glucose concentration increases, the ratio decreases, using Beer-Lambert’s law. This work was not in a wearable,
but using a light source and smartphone. Additionally, 650nm wavelength of light being shown on a solution has
been shown to increase in transmittance of photons, with glucose and was done using a handheld device [5]. A
noninvasive continuous blood glucose monitor can lead to advancements in timely prediabetes identification.
The goal of this study is to demonstrate a proof of concept for such a device by leveraging Lumos.
6.2 Spectral Response Identification

In order to identify the optimal wavelengths to monitor glucose, we ran a preliminary study. In this study, we
monitored the glucose of a single person over the course of an hour. This study determined that three wavelengths:
470nm, 515nm, and 680nm, showed promising responses for change in glucose. To validate this, we customized
our sensor to target 470nm, 515nm, and 680nm. These wavelengths do not align with the original surface mount
LEDs used in Lumos. As these components are different, we developed a new custom form factor shown in Figure
13a. This form factor was designed to house a single through hole LED while maintaining continuous pressure
on the finger. Additionally, this form factor is designed such that it can be worn on three fingers simultaneously.
We chose to run each LED on separate fingers to simultaneously collect data from all three sensors without the
data loss that switching LEDs could cause. Our prototype is shown in Figure 13b.
6.3 Results
We tested our proof of concept sensor in comparison with a commercially available glucometer [10] in a user
study. This study used a single participant where we attached the Lumos devices on the participant’s left hand as
shown in Figure 13b. The 680nm LED was attached to the pointer finger, the 515nm LED to the middle finger,
and the 475nm LED to the ring finger. Blood glucose readings were drawn from the fingers on the right hand.
This setup is shown in Figure 13b. Each study was performed over approximately 30 minutes to one hour. We
repeated this study five times. Our participant began in a fasted state and we started the studies in the morning.
This appears as a low starting glucose readings. When the study began, they drank a sugary drink (soda), and we
recorded the effect over time of that on their blood sugar. Approximately every three minutes, we took a new
blood glucose reading. We continuously sampled the three Lumos devices at approximately 2 Hz.
(a) 475nm LED (b) 515nm LED (c) 680nm LED
Fig. 14. Results from Glucose Experiments
In Figure 14 we show the raw data collected from each LED with its targeted PD. The ground truth blood glucose
is shown by the red points. It is measured by a commercially available glucometer. The optical spectroscopy
readings from the 480nm, 515nm, and 680nm photodetector channels are shown by the blue lines. In Figure 14a
and Figure 14c, the change in counts positively trends with the change in glucose. In Figure 14b, the change in
counts negatively trends with the glucose readings. In this figure, the y-axis of the glucose readings is flipped.
We performed further analysis to quantify the correlations between the glucose measurements and the readings
from our sensor at our three target wavelengths. To compare these two data points, we interpolated the missing
glucose values using a linear interpolation. We found that 470nm and 515nm were statistically significant with
a p-values less than 0.05 and Pearson correlations of 0.843 and -0.927 respectively. The negative correlation
showing that the glucose reading and Lumos’s reading are highly correlated but in the the opposite directions
from each other. 680nm showed relatively low correlation when compared with our ground truth glucose values
with a Pearson correlation of 0.359. The lack of correlation for the 680nm channel encourages future work in the
surrounding wavelengths including expanding into the infrared spectrum. These results show that readings from
our sensor may be used to track overarching glucose trends. They prompt additional research to continue to
improve the noninvasive monitoring of glucose and thus track the progression of prediabetes.
7 DISCUSSION AND FUTURE WORK

In this work, we demonstrated Lumos, an open-source device for wearable spectroscopy. It enables real-time,
non-invasive, and continuous health monitoring outside of the clinic or lab. We acknowledge that the Lumos
device is low fidelity compared to a benchtop spectrometer, but it allows users to determine what wavelengths
to target for an application and then build a higher fidelity sensor around those wavelengths. We envision the
Lumos Device being used for research purposes. It is a powerful tool that can help researchers determine the
wavelengths of interest for applications. Once the wavelengths have been identified, more targeted devices can
be developed. This will help to improve accuracy, battery life, and a form factor can be developed that is suited
specifically for the application.
7.1 Device Improvement

While our device provides the ability to perform wearable optical spectroscopy, it is limited in many facets. The
size can be decreased and the sensor data can be improved in terms of granularity and overall range. This will
be accomplished through customization and miniaturization. Decreasing the size of the device will increase its
portability and ease of placement on the body. Increasing the granularity of the device will improve sensing
via more data points that will unlock details about the photons being detected with the goal of improving the
accuracy of Lumos. Increasing granularity beyond the visible spectrum will bring further insights into the makeup
of the medium in other ranges. These works have shown absorption peaks of glucose in the NIR range and we
plan to implement this in our work [26, 75].
In order to improve the size of the device, customization is key. This device can be shrunk by swapping the
Arduino for a board targeted to our specifications. Beyond that, to further shrink the prototype, a customized
circuit could be designed for parts of this system and further tailor our design to our specifications. The next
iteration of form-factors for the wrist-based reflective device will include better protection against cross-talk
between the LEDs and photodetector. An analysis of the effects of distance and angle between the LEDs and
photodetector is needed to determine the optimal placement of components to maximize count values. We also
plan to include the frequency and pulse width of each LED as part of the current calibration step to make the
device more adaptable to various body compositions while decreasing battery consumption.
The current sensor can only sense eight center wavelengths in the visible spectrum, which reduces the
granularity of our device significantly. In the next version of our device, we want to increase this granularity
by sensing more center wavelengths along the visible spectrum. We also want to target specific LEDs with
a controllable intensity that will be most effective for specific and additional biomarkers and spectrometry
technology. Additionally, our current sensor has one specific Near Infrared Reflectance (NIR) channel that it
senses, and we want to include NIR sensing in the next version of this device. More than that one channel, we
want to expand the sensor beyond the visible range with much more granularity in order to sense more of the
electromagnetic spectrum. In expanding the spectrum, both our sensor and source require improvements that
span additional wavelengths.
Overall, our current device has potential improvements that will unlock key advancements in the optical
sensing of biomarkers in the body. Among these, decreasing the size, increasing the granularity and range, and
improving its basic specifications, such as power consumption, will greatly improve the performance and results
of the device.
7.2 Algorithm Improvement

Our algorithm details the spectral response of a medium and gives insight into target applications. In this paper,
we demonstrated good overall performance of our algorithm. It can be improved in two facets. First, we can
continue to improve on the theoretical model. Second, we can upgrade our wavelength detection. Our theoretical
approximation approximates the response of a perfect set of LEDs and PDs. In the real world, this is not feasible,
and as such, we see errors in our approximations. As there is some error inherently in each LED and PD, we
can add additional estimations to account for this error. This will allow for a more accurate approximation. The
leakage current bias is currently being calculated based on the experimental reading. This could be modeled
and calculated as we calculate the rest of our theoretical response. Each PD should have a baseline leakage
current bias that would then be scaled based on the intensity of the light being shown on our PD. The wavelength
detection algorithm finds the peaks of the interpolated data using SciPy. Essentially, if the interpolated data in
Figure 12 crosses a certain threshold and is after a certain amount of samples from the previous peak, then a peak
is identified and recorded. In order to improve on this, we can detect not only a peak but a range of wavelengths
for further analysis. To accomplish this, we should detect a range similar to a full-width half max surrounding
the center wavelength where the counts are significant.
7.3 Real World Deployment

The Lumos device has been tested in a lab setting. To further improve our device, we plan to test it in the real
world. Lumos will behave similarly to IoT devices, such as smartwatches, by wirelessly collecting signals across
multiple wavelengths that can be used to determine biomarkers such as heart rate, respiration rate, blood oxygen
levels (SpO2), and glucose. This will facilitate future studies on a broad patient population with varying human
physiology, skin pigmentation, and comorbidities. The device has a maximum battery life of five hours with
continuous usage of all LEDs. But like many real world deployments, we can duty-cycle the data collection and
transfer process. This will significantly increase the battery life and allow longer user studies to be performed.
Limiting the wavelengths sampled can also increase the battery life. By collecting data at certain intervals,
running the device on low power mode, and limiting the spectrum sampled to what is needed for the study, we
can extend the battery life by up to sixty hours, almost twice the amount of a regular smartwatch.
The wrist form factor has the ability to collect data for multiple wavelengths during high-motion scenarios
such as running, working out, or playing a sport. To fully utilize this form factor, it needs to be resistant to motion
artifacts. Motion artifacts affect specific wavelengths significantly, while others are resistant to them [19, 57].
Testing the various wavelengths we utilize while inducing motion artifacts could help create smarter wearable
sensors with adaptable wavelengths depending on the user’s state of motion. For example, green wavelengths
are more resistant to motion artifacts and have a good signal-to-noise ratio but are easily absorbed by the skin,
thereby limiting their penetration depth [37]. Red Wavelengths are great at monitoring heart rate and have deeper
penetration than green but are highly sensitive to motion noise [20]. Different wavelengths are also absorbed by
the skin differently based on melanin content.
8 CONCLUSION
In this paper, we present Lumos, an open-source device for wearable spectroscopy research. Using our custom
PCB and off-the-shelf components, Lumos demonstrates five hours of battery life. We developed an algorithm
to determine the spectral response of a medium while remaining invariant to environmental disturbances with
a 13nm mean absolute error. From this, researchers can determine the optimal spectrum to target for their
application facilitating on-body spectroscopy research in health monitoring, athletics, rehabilitation, and more.
We show this utility in a pilot study, sensing prediabetes, where we determined three relevant wavelengths,
470nm, 515nm, and 680nm, for glucose and created and evaluated a targeted tracking device.
ACKNOWLEDGMENTS
The authors would like to thank Xian Li for his early work in sensor selection, Hyonyoung Choi for allowing us
to use his camera and photography expertise, and Kyle C. Quinn for his efforts in proofreading and editing. This
research was supported in part by the National Science Foundation under grants NSF-1915398, NSF-2125561 and
by the National Institute of Health under grants NIH 1R01EB029363-01 and NIH 1R01EB029767-01.
REFERENCES
[1] Nurul Akmal Binti Abd Salam, Wira Hidayat bin Mohd Saad, Zahariah Binti Manap, and Fauziyah Salehuddin. 2016. The evolution
of non-invasive blood glucose monitoring system for personal application. Journal of Telecommunication, Electronic and Computer
Engineering (JTEC) 8, 1 (2016), 59–65.

[2] Adafruit 2022. AS7341: 11-Channel Multi-Spectral Digital Sensor. https://ams.com/documents/20143/36005/AS7341_DS000504_3-00.pdf.
[3] Mahmoud Al-Omari, Kivanc Sel, Anja Mueller, Jeffery Edwards, and Tolga Kaya. 2014. Detection of relative [Na+] and [K+] levels in
sweat with optical measurements. Journal of Applied Physics 115, 20 (2014), 203107.
[4] M Kathleen Alam, Mark R Rohrscheib, James E Franke, Thomas M Niemczyk, John D Maynard, and M Ries Robinson. 1999. Measurement
of pH in whole blood by near-infrared spectroscopy. Applied spectroscopy 53, 3 (1999), 316–324.
[5] Haider Ali, Faycal Bensaali, and Fadi Jaber. 2017. Novel approach to non-invasive blood glucose monitoring based on transmittance and
refraction of visible laser light. IEEE access 5 (2017), 9163–9174.
[6] Pavel Anzenbacher Jr, Fengyu Li, and Manuel A Palacios. 2012. Toward wearable sensors: fluorescent attoreactor mats as optically
encoded cross-reactive sensor arrays. Angewandte Chemie 124, 10 (2012), 2395–2398.
[7] Krzysztof B Beć, Justyna Grabska, and Christian W Huck. 2021. Principles and applications of miniaturized near-infrared (NIR)
spectrometers. Chemistry–A European Journal 27, 5 (2021), 1514–1532.
[8] Matthew G Blain, Leah S Riter, Dolores Cruz, Daniel E Austin, Guangxiang Wu, Wolfgang R Plass, and R Graham Cooks. 2004. Towards the
hand-held mass spectrometer: design considerations, simulation, and fabrication of micrometer-scaled cylindrical ion traps. International
Journal of Mass Spectrometry 236, 1-3 (2004), 91–104.
[9] Karthik Budidha, Mohammad Mamouei, Nystha Baishya, Meha Qassem, Pankaj Vadgama, and Panayiotis A Kyriacou. 2020. Identification
and quantitative determination of lactate using optical spectroscopy—towards a noninvasive tool for early recognition of sepsis. Sensors
20, 18 (2020), 5402.
[10] CareTouch 2022. CareTouch Blood Glucose Monitor Kit. https://www.amazon.com/Care-Touch-Diabetes-Testing-Kit/dp/B076VSN7TR/
[11] Albert E Cerussi, Vaya W Tanamai, David Hsiang, John Butler, Rita S Mehta, and Bruce J Tromberg. 2011. Diffuse optical spectroscopic
imaging correlates with final pathological response in breast cancer neoadjuvant chemotherapy. Philosophical transactions of the royal
society A: mathematical, physical and engineering sciences 369, 1955 (2011), 4512–4530.
[12] Charles Chabal, Louis Jacobson, Anthony J Mariano, and Edmund Chaney. 1998. Opiate abuse or undertreatment? The Clinical Journal
of Pain 14, 1 (1998), 90–91.
[13] Vishnu Dantu, Jagannadh Vempati, and Srinivasan Srivilliputhur. 2014. Non-invasive blood glucose monitor based on spectroscopy
using a smartphone. In 2014 36th annual international conference of the IEEE engineering in medicine and biology society. IEEE, 3695–3698.
[14] ACM Dassel, R Graaff, M Sikkema, A Meijer, WG Zijlstra, and JG Aarnoudse. 1995. Reflectance pulse oximetry at the forehead improves
by pressure on the probe. Journal of clinical monitoring 11, 4 (1995), 237–244.
[15] Digikey 2022. LED Emitters- Infrared, UV, Visible. https://www.digikey.com/en/products/filter/led-emitters-infrared-uv-visible/94
[16] Tanja Dragojević, Joseph L Hollmann, Davide Tamborini, Davide Portaluppi, Mauro Buttafava, Joseph P Culver, Federica Villa, and
Turgut Durduran. 2018. Compact, multi-exposure speckle contrast optical spectroscopy (SCOS) device for measuring deep tissue blood
flow. Biomedical optics express 9, 1 (2018), 322–334.
[17] Fatema El-Amrawy and Mohamed Ismail Nounou. 2015. Are currently available wearable devices for activity tracking and heart rate
monitoring accurate, precise, and medically beneficial? Healthcare informatics research 21, 4 (2015), 315–320.
[18] Annika MK Enejder, Thomas G Scecina, Jeankun Oh, Martin Hunter, WeiChuan Shih, Slobodan Sasic, Gary L Horowitz, and Michael S
Feld. 2005. Raman spectroscopy for noninvasive glucose measurements. Journal of biomedical optics 10, 3 (2005), 031114.
[19] Litong Feng, Lai-Man Po, Xuyuan Xu, Yuming Li, and Ruiyi Ma. 2014. Motion-resistant remote imaging photoplethysmography based
on the optical properties of skin. IEEE Transactions on Circuits and Systems for Video Technology 25, 5 (2014), 879–891.
[20] Louise Finlayson, Isla RM Barnard, Lewis McMillan, Sally H Ibbotson, C Tom A Brown, Ewan Eadie, and Kenneth Wood. 2022. Depth
penetration of light into skin as a function of wavelength from 200 to 1000 nm. Photochemistry and Photobiology 98, 4 (2022), 974–981.
[21] Rodolfo J Galindo and Grazia Aleppo. 2020. Continuous glucose monitoring: the achievement of 100 years of innovation in diabetes
technology. Diabetes research and clinical practice 170 (2020), 108502.
[22] Mayank Goel, Eric Whitmire, Alex Mariakakis, T Scott Saponas, Neel Joshi, Dan Morris, Brian Guenter, Marcel Gavriliu, Gaetano
Borriello, and Shwetak N Patel. 2015. HyperCam: hyperspectral imaging for ubiquitous computing applications. In Proceedings of the
2015 ACM International Joint Conference on Pervasive and Ubiquitous Computing. 145–156.
[23] Gossen 2022. Gossen MAVOSPEC BASE Innovative Spectrometer. https://www.tequipment.net/Gossen/MAVOSPEC-BASE/Light-
Meters/Illuminance-Meters/.
[24] Weixi Gu, Yuxun Zhou, Zimu Zhou, Xi Liu, Han Zou, Pei Zhang, Costas J. Spanos, and Lin Zhang. 2017. SugarMate: Non-Intrusive Blood
Glucose Monitoring with Smartphones. Proc. ACM Interact. Mob. Wearable Ubiquitous Technol. 1, 3, Article 54 (sep 2017), 27 pages.
[25] Robin Kingsley Harris. 1986. Nuclear magnetic resonance spectroscopy. (1986).
[26] Rolamjaya Hotmartua, Pujianto Wira Pangestu, Hasballah Zakaria, and Yoke Saadia Irawan. 2015. Noninvasive blood glucose detection
using near infrared sensor. In 2015 International Conference on Electrical Engineering and Informatics (ICEEI). IEEE, 687–692.
[27] ILT350 Chroma Meter 2022. ILT350 Chroma Meter. https://www.intl-lighttech.com/products/ilt350-chroma-meter.
[28] Arvind Jina, Michael J Tierney, Janet A Tamada, Scott McGill, Shashi Desai, Beelee Chua, Anna Chang, and Mark Christiansen. 2014.
Design, development, and evaluation of a novel microneedle array-based continuous glucose monitor. Journal of diabetes science and
technology 8, 3 (2014), 483–487.

[29] Omar S Khalil. 2004. Non-invasive glucose measurement technologies: an update from 1999 to the dawn of the new millennium. Diabetes
technology & therapeutics 6, 5 (2004), 660–697.
[30] Svetlana Khaustova, Maxim Shkurnikov, Evgeny Tonevitsky, Viacheslav Artyushenko, and Alexander Tonevitsky. 2010. Noninvasive
biochemical monitoring of physiological stress by Fourier transform infrared saliva spectroscopy. Analyst 135, 12 (2010), 3183–3192.
[31] Jayoung Kim, Alan S Campbell, Berta Esteban-Fernández de Ávila, and Joseph Wang. 2019. Wearable biosensors for healthcare
monitoring. Nature biotechnology 37, 4 (2019), 389–406.
[32] Richard P Kingsborough, Alexandra T Wrobel, Devon Beck, Lauren Cantley, Shane Tysk, and Roderick Kunz. 2020. Chemical sensing
via a low SWaP wearable spectrometer. In Chemical, Biological, Radiological, Nuclear, and Explosives (CBRNE) Sensing XXI, Vol. 11416.
International Society for Optics and Photonics, 114160O.
[33] Ryan R Kroll, J Gordon Boyd, and David M Maslove. 2016. Accuracy of a wrist-worn wearable device for monitoring heart rates in
hospital inpatients: a prospective observational study. Journal of medical Internet research 18, 9 (2016), e6025.
[34] Michele Lacerenza, Mauro Buttafava, Marco Renna, Alberto Dalla Mora, Lorenzo Spinelli, Franco Zappa, Antonio Pifferi, Alessandro
Torricelli, Alberto Tosi, and Davide Contini. 2020. Wearable and wireless time-domain near-infrared spectroscopy system for brain and
muscle hemodynamic monitoring. Biomedical Optics Express 11, 10 (2020), 5934–5949.
[35] V Lakshminarayanan and N Sriraam. 2014. The effect of temperature on the reliability of electronic components. In 2014 IEEE international
conference on electronics, computing and communication technologies (CONECCT). IEEE, 1–6.
[36] Etienne Lareau, Frederic Lesage, Philippe Pouliot, Jerome Le Lan, Mohamad Sawan, and Dang Nguyen. 2011. Multichannel wearable
system dedicated for simultaneous electroencephalography/near-infrared spectroscopy real-time data acquisitions. Journal of biomedical
optics 16, 9 (2011), 096014.
[37] Jihyoung Lee, Kenta Matsumura, Ken-ichi Yamakoshi, Peter Rolfe, Shinobu Tanaka, and Takehiro Yamakoshi. 2013. Comparison between
red, green and blue light reflection photoplethysmography for heart rate monitoring during motion. In 2013 35th Annual International
Conference of the IEEE Engineering in Medicine and Biology Society (EMBC). 1724–1727. https://doi.org/10.1109/EMBC.2013.6609852
[38] M Lepore, M Portaccio, E De Tommasi, P De Luca, U Bencivenga, P Maiuri, and DG Mita. 2004. Glucose concentration determination by
means of fluorescence emission spectra of soluble and insoluble glucose oxidase: some useful indications for optical fibre-based sensors.
Journal of Molecular Catalysis B: Enzymatic 31, 4-6 (2004), 151–158.
[39] Huijie Li, Guangfu Wu, Zhengyan Weng, He Sun, Ravi Nistala, and Yi Zhang. 2021. Microneedle-Based Potentiometric Sensing System
for Continuous Monitoring of Multiple Electrolytes in Skin Interstitial Fluids. ACS sensors 6, 6 (2021), 2181–2190.
[40] Linfan Li, Tsung-Chi Chen, Yue Ren, Paul I Hendricks, R Graham Cooks, and Zheng Ouyang. 2014. Mini 12, Miniature Mass Spectrometer
for Clinical and Other Applications Introduction and Characterization. Analytical chemistry 86, 6 (2014), 2909–2916.
[41] YanFeng Li, Linda S Geiss, Nilka R Burrows, Deborah B Rolka, and Ann Albright. 2013. Awareness of prediabetes—United States,
2005–2010. Morbidity and Mortality Weekly Report 62, 11 (2013), 209.
[42] LibreTexts 2022. Spectroscopic Methods. https://chem.libretexts.org/@go/page/70697.
[43] Julia Madden, Conor O’Mahony, Michael Thompson, Alan O’Riordan, and Paul Galvin. 2020. Biosensing in dermal interstitial fluid
using microneedle based electrochemical devices. Sensing and Bio-Sensing Research 29 (2020), 100348.
[44] Laxmaiah Manchikanti, Standiford Helm 2nd, Bert Fellows, Jeffrey W Janata, Vidyasagar Pampati, Jay S Grider, and Mark V Boswell.
2012. Opioid epidemic in the United States. Pain physician 15, 3 Suppl (2012), ES9–38.
[45] Paul D Mannheimer. 2007. The light–tissue interaction of pulse oximetry. Anesthesia & Analgesia 105, 6 (2007), S10–S17.
[46] Yitzhak Mendelson, R James Duckworth, and G Comtois. 2006. A wearable reflectance pulse oximeter for remote physiological
monitoring. In 2006 International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE, 912–915.
[47] Ivan Moreno, Chang-Yu Tsai, David Bermúdez, and Ching-Cherng Sun. 2007. Simple function for intensity distribution from LEDs. In
Nonimaging Optics and Efficient Illumination Systems IV, Vol. 6670. SPIE, 137–143.
[48] Deirdre Morris, Shirley Coyle, Yanzhe Wu, King Tong Lau, Gordon Wallace, and Dermot Diamond. 2009. Bio-sensing textile based patch
with integrated optical detection system for sweat monitoring. Sensors and Actuators B: Chemical 139, 1 (2009), 231–236.
[49] Mouser 2022. Infrared Emitters. https://www.mouser.com/c/optoelectronics/infrared-data-communications/infrared-emitters/
[50] UA Müller, B Mertes, C Fischbacher, KU Jageman, and K Danzer. 1997. Non-invasive blood glucose monitoring by means of near infrared
spectroscopy: methods for improving the reliability of the calibration models. The International journal of artificial organs 20, 5 (1997),
285–290.
[51] David M Nathan. 2015. Diabetes: advances in diagnosis and treatment. Jama 314, 10 (2015), 1052–1062.
[52] Ocean Insight 2022. Preconfigured Ocean FX VIS-NIR Spectrometer. https://www.oceaninsight.com/products/spectrometers/high-
speed-acquisition/ocean-fx/ocean-fx-vis-nir-1/
[53] Martina O’Toole and Dermot Diamond. 2008. Absorbance based light emitting diode optical sensors and sensing devices. Sensors 8, 4
(2008), 2453–2479.
[54] Jakub Parak, Adrian Tarniceriu, Philippe Renevey, Mattia Bertschi, Ricard Delgado-Gonzalo, and Ilkka Korhonen. 2015. Evaluation of
the beat-to-beat detection accuracy of PulseOn wearable optical heart rate monitor. In 2015 37th Annual International Conference of the
IEEE Engineering in Medicine and Biology Society (EMBC). IEEE, 8099–8102.

[55] Pasco 2022. Spectroscopy. https://www.pasco.com/products/guides/what-is-spectroscopy
[56] Nadia K Pervez, Tanya C Garza, Michael J Gazes, James I Scholtz, Mark F Comerford, and Ioannis Kymissis. 2017. Integrated Light
Management as a Path to Miniaturizing Spectrometers. In Applied Industrial Optics: Spectroscopy, Imaging and Metrology. ATh2A–3.
[57] David Pollreisz and Nima TaheriNejad. 2019. Detection and removal of motion artifacts in PPG signals. Mobile Networks and Applications
(2019), 1–11.
[58] SP Preejith, Annamol Alex, Jayaraj Joseph, and Mohanasankar Sivaprakasam. 2016. Design, development and clinical validation of a
wrist-based optical heart rate monitor. In 2016 IEEE International Symposium on Medical Measurements and Applications (MeMeA). IEEE,
1–6.
[59] Colin Raymond, Tom Matthews, and Radley M Horton. 2020. The emergence of heat and humidity too severe for human tolerance.
Science Advances 6, 19 (2020), eaaw1838.
[60] Darren M Roblyer. 2020. Perspective on the increasing role of optical wearables and remote patient monitoring in the COVID-19 era
and beyond. Journal of Biomedical Optics 25, 10 (2020), 102703.
[61] Roithner 2022. Roithner LaserTechnik GmbH. http://www.roithner-laser.com/
[62] Markus Rothmaier, Bärbel Selm, Sonja Spichtig, Daniel Haensse, and Martin Wolf. 2008. Photonic textiles for pulse oximetry. Optics
express 16, 17 (2008), 12973–12986.
[63] Leah Sera and Mary Lynn McPherson. 2018. Management of opioid-induced constipation in hospice patients. American Journal of
Hospice and Palliative Medicine® 35, 2 (2018), 330–335.
[64] Abhishek B Sharma, Leana Golubchik, and Ramesh Govindan. 2010. Sensor faults: Detection methods and prevalence in real-world
datasets. ACM Transactions on Sensor Networks (TOSN) 6, 3 (2010), 1–39.
[65] Sanjiv Sharma, Ahmed El-Laboudi, Monika Reddy, Narvada Jugnee, Sujan Sivasubramaniyam, Mohamed El Sharkawy, Pantelis Georgiou,
Desmond Johnston, Nick Oliver, and Anthony EG Cass. 2018. A pilot study in humans of microneedle sensor arrays for continuous
glucose monitoring. Analytical Methods 10, 18 (2018), 2088–2095.
[66] Kevin AG Smet. 2020. Tutorial: The LuxPy Python toolbox for lighting and color science. Leukos 16, 3 (2020), 179–201.
[67] StellarRAD 2022. StellarRAD+Color Handheld-Colorimeter. https://www.shopstellarnet.com/stellarrad-color-handheld-colorimeter.
[68] Fei Teng, Timothy Cormier, Alexis Sauer-Budge, Rachita Chaudhury, Vivian E Pera, Raeef Istfan, David A Chargin, Samuel Brookfield,
Naomi Yu Ko, and Darren M Roblyer. 2017. Wearable near-infrared optical probe for continuous monitoring during breast cancer
neoadjuvant chemotherapy infusions. Journal of biomedical optics 22, 1 (2017), 014001.
[69] Hazhir Teymourian, Farshad Tehrani, Kuldeep Mahato, and Joseph Wang. 2021. Lab under the skin: microneedle based wearable devices.
Advanced healthcare materials 10, 17 (2021), 2002255.
[70] Nikolai V Tkachenko. 2006. Optical spectroscopy: methods and instrumentations. Elsevier.
[71] UNLV 2022. Things you should know about photodiodes. https://www.physics.unlv.edu/~bill/PHYS483/LED_PIN.pdf
[72] Amanda Watson. 2022. Lumos. https://github.com/aawatson22/Lumos.
[73] Amanda Watson, Hyonyoung Choi, Insup Lee, and James Weimer. 2021. Raproto: an open source platform for rapid prototyping of
wearable medical devices. In Proceedings of the Workshop on Medical Cyber Physical Systems and Internet of Medical Things. 1–6.
[74] Dominik G Wyser, Olivier Lambercy, Felix Scholkmann, Martin Wolf, and Roger Gassert. 2017. Wearable and modular functional
near-infrared spectroscopy instrument with multidistance measurements at four wavelengths. Neurophotonics 4, 4 (2017), 041413.
[75] Jyoti Yadav, Asha Rani, Vijander Singh, and Bhaskar Mohan Murari. 2015. Prospects and limitations of non-invasive blood glucose
monitoring using near-infrared spectroscopy. Biomedical signal processing and control 18 (2015), 214–227.
[76] LDS Yadav. 2005. Ultraviolet (UV) and visible spectroscopy. In Organic Spectroscopy. Springer, 7–51.
[77] Zongyin Yang, Tom Albrow-Owen, Weiwei Cai, and Tawfique Hasan. 2021. Miniaturization of optical spectrometers. Science 371, 6528
(2021), eabe0722.

Lumos: An Open-Source Device For Wearable Spectroscopy Research

Uploaded by

Copyright:

Available Formats

Lumos: An Open-Source Device For Wearable Spectroscopy Research

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Lumos: An Open-Source Device For Wearable Spectroscopy Research

Uploaded by

Copyright:

Available Formats

Lumos: An Open-Source Device for Wearable Spectroscopy Research

AMANDA WATSON∗ and CLAIRE KENDELL∗ , University of Pennsylvania, USA

∗ Both authors contributed equally to this research. 187

ACM Reference Format:

2.2 Other Approaches to Spectral/Optical Sensing

2.3 Applications of Wearable Spectroscopy

3.1 Theory of Operation

(a) Absorption Spectroscopy (b) Emission Spectroscopy

Fig. 2. Theory of Spectroscopy

(a) Finger Clamp Form Factor (b) Watch Form Factor

Fig. 3. Lumos Device Form Factors

3.2 Form Factors

(a) LED Wavelengths (b) PD Wavelengths

Fig. 5. Spectral Response of our LEDs and PDs

4 SPECTRAL RESPONSE OF MEDIUM

4.1 Intensity Calibration

𝐼𝑖+1 = 𝐾𝑝 (𝑟𝑡 − 𝑟𝑖 ) (1)

4.2 Environmental Light Leakage

4.3 Theoretical Approximation

Fig. 6. Calculation of Theoretical Approximation for 530nm (Green) LED

Relative Intensity [A.U.]

Fig. 7. Spectral Response Calculation

5.1 Temperature Experiment

Spectral Response (counts)

Fig. 8. Temperature and Fluid Experiment Results

5.2 Fluid Experiment

5.4 Intensity Calibration

Fig. 9. Intensity Calibration at Each 𝐾𝑝 for 530nm LED

Table 1. Time and Computational Needs for Each 𝐾𝑝 s

𝑲𝒑 # Steps SD Steps Time (s)

5.5 Comparison with Spectrometer

(a) Comparison with Benchtop Spectrometer (b) Experimental Setup

Fig. 10. Spectrometer Comparison Experiment

415nm LED 450nm LED 470nm LED 530nm LED

Exp Exp Exp Exp

568nm LED 599nm LED 633nm LED 660nm LED

Exp Exp Exp Exp

5.6 Theoretical Estimation

5.7 Spectral Response of Medium

Purple Blue Green

Yellow Orange Red

680 680 680

Fig. 12. Spectral Response of Mediums

6.1 Background and Motivation

(a) Glucose Form Factor Design (b) Glucose Monitor Prototype

Fig. 13. Glucose Device Design and Initial Prototype

6.2 Spectral Response Identification

(a) 475nm LED (b) 515nm LED (c) 680nm LED

Fig. 14. Results from Glucose Experiments

7 DISCUSSION AND FUTURE WORK

7.1 Device Improvement

7.2 Algorithm Improvement

7.3 Real World Deployment

Engineering (JTEC) 8, 1 (2016), 59–65.

technology 8, 3 (2014), 483–487.

IEEE Engineering in Medicine and Biology Society (EMBC). IEEE, 8099–8102.

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.