by Alexandra Bustos Iliescu
What if losing a limb no longer meant a lifetime of limited mobility? What if advanced technology could not only restore movement but also enhance the quality of life for those with physical disabilities? These questions formed the core of a thought-provoking session at the AI for Good Global Summit 2024 where experts from academia and industry joined to discuss the latest innovations in prosthetic and rehabilitation engineering. Moderated by Adjunct Professor and co-Director Olivier Lambercy from the Rehabilitation Engineering Laboratory at ETH Zurich, the session delved into how robotics and AI are said to revolutionize healthcare, offering new solutions that could improve the quality of life for people with physical disabilities.
Dima Gazda, CEO and co-Founder of Esper Bionics, introduced the work of his company, which specializes in robotic limbs for individuals without hands or legs. Gazda highlighted the immense potential of health technology, driven by the collection of scale body data and the integration of electronics within the human body. Esper Bionics’ ecosystem comprises assistive devices, wearable sensors, and a software platform that enhances the control and functionality of these devices. Gazda underscored the emotional journey of individuals using their technology, from initial despair to renewed hope and functionality.
He shared a poignant observation: “The hope to restore the body leads to the hope to win enemies that is 30 times bigger than you.”
Gazda highlighted the company’s distributed operations, with headquarters in New York, an R&D facility in Berlin, and manufacturing facilities in Ukraine. He emphasized the significant impact of their work in Ukraine, where many individuals have lost limbs due to conflict.
“By Wall Street Journal, there are more than 50,000 Ukrainians who lost their hands or legs during the last two years, and we are the company that can help most of them,” Gazda stated.
Next, Professor Robert Riener, Full Professor in the Department of Health Sciences and Technology at ETH Zurich, emphasized the importance of high-dose, user-oriented therapy for stroke patients and others with motor impairments. Riener pointed out that traditional training methods, limited to one hour per week or day, are insufficient. Instead, continuous, intensive training is necessary, akin to how children learn motor skills.
Riener highlighted the significance of user-oriented design in rehabilitation technology. He explained that developers must engage with patients and clinical personnel to create effective solutions.
Riener noted, “We can apply AI based on sensor technology to measure and monitor their human behavior and detect their preferences to allow a well-defined, user-oriented application of high-tech.”
Amine Metani, Co-founder and Chief Scientific Officer at Kurage, presented their groundbreaking device, Neuroskin. This AI-powered electrical muscle stimulator compensates for damaged parts of the brain, spinal cord, or nerves to restore movement. Metani emphasized the vital importance of physical activity for individuals with motor handicaps to prevent secondary diseases and improve quality of life.
Metani described the working mechanism of Neuroskin, which involves sensors embedded in clothing items like pants, shoes, and belts. These sensors transmit data continuously to an AI system that analyzes the movement and determines the necessary assistance. The AI then directs an electrical stimulator to send precise pulses to elicit muscle contractions, aiding in activities such as standing up, walking, climbing stairs, and even riding a bike. This technique, known as functional electrical stimulation (FES), has been around for decades but has now been made user-friendly through AI advancements.
Metani shared impressive results from using Neuroskin: “By using Neuroskin for 30 minutes a day, 5 days a week, stroke survivors in the acute recovery phase have been able to improve their walking speed and endurance by over 80%.” This achievement, eight times greater than standard rehab therapy, highlights the transformative potential of AI in rehabilitation.
Professor Heike Vallery from RWTH Aachen University and TU Delft contextualized the technological advancements enabling these breakthroughs. She noted the shift from rigid robotic structures to more compliant, interactive devices that encourage active participation and learning through error. Vallery showcased innovative projects like the gyroscopic actuator, which assists with balance, and the robot ball “Fizzi,” a hybrid exercise device and pet that motivates users to engage in their rehabilitation exercises. These examples illustrate the ongoing efforts to make rehabilitation technologies more accessible, user-friendly, and engaging.
Vallery explained that traditional robotic devices often impose a fixed gait pattern, which limits the user’s ability to make errors and actively participate in the training process. She highlighted the importance of physical compliance in new devices, allowing for finer force control and more natural movement. Vallery also discussed the need to make these devices portable and accessible, moving beyond the confines of rehabilitation clinics to enable home use.
A key theme across the presentations was the push to extend rehabilitation technologies beyond the clinic to patients’ homes. This shift aims to address the socioeconomic pressures on healthcare systems and the need for continuous, intensive therapy even outside hospital settings. The speakers emphasized the importance of user-centered design, robustness, and ease of use to ensure these technologies are effective and safe for home use.
The panelists discussed various challenges in making rehabilitation technologies suitable for home use.
Vallery emphasized the need for simplicity and user-friendliness, stating, “They should not be a barrier in setting it up and then actually getting the individual to learn how to use it.”
The panel discussion delved into the challenges and opportunities of integrating AI in neuro-rehabilitation. The experts agreed that while AI holds immense potential, ensuring safety, robustness, and user trust remains critical. Professor Riener pointed out the need for clear benchmarks and success criteria, emphasizing the importance of collaborative development with clinicians and patients to ensure the technologies meet real-world needs. Metani and Gazda echoed these sentiments, stressing the importance of usability and engagement in designing effective rehabilitation devices. They highlighted the role of continuous feedback and data analysis in refining and personalizing these technologies to better serve users’ needs.
