IM
US 200505172541
c») United States
2 Patent Application Publication 0) Pub. No.: US 2005/0051725 A1
Nakagaki et al, 3) Pub, Date: Mar. 10, 2005
6» Publication Classeation
6) Ine Gown 23/00
(2) US.cL 250311
(76) tavestors: Ryo Nakagakt, Kawasaki OP), Yu
Takagh, Kamakura (P); Hiroto 6 ABSTRACT
Okada, Yokohama (IP); Hiroshi
een ee) Lis possible to reliably and efficiently determine whether a
specimen contains viruses, bacteria, etc. and, if it does,
iden thee types, regardless ofthe observer. Furthermore,
DATABASE!
ss Aas
RECRERTEAMET sroor a xnaus, _ Sersumitlseedtncernm nw eed
{300 NORTH SEVENTEENTH STREET . -- —
— Ba en) and an imaging condition used when the microscopic thing
Ce api Nos a0oi8ado ‘Sopra me dt se Ay ie te
oF
——Lr— {nt cated ne compas hinges sed
J 18
101 107
M16 M16 ‘17
2
OVERALL STORAGE
ConTROL| |S™ONA
IMAGING ott
CONTROL,
UNIT
110
PROCESSING | | INPUT/OUTPUT
4 UNIT. UNIT
CAMERA 112 113Patent Application Publication Mar. 10,2005 Sheet 1of7 US 2005/0051725 Al
FIG.1
18
107
-—_—
Ms 16 a7
OVERALL
STORAGE |
comet] | unit | |paTaBase
IMAGING —
Si
CONTROL]
UNIT
[
110
112
CAMERA
PROCESSING
UNIT
INPUT/OUTPUT,
UNIT
113Patent Application Publication Mar. 10,2005 Sheet 2 of 7 US 2005/0051725 AL
FIG.2
TET OPERATOR ENTER KEYWORD |—~~201
INTO INPUT UNIT
SELECT ONE OR MORE 202
REFERENCE MICROBES BASED [~~~
ON KEYWORD
RETRIEVE PRETREATMENT 203
METHOD, IMAGING CONDITIONS, }——~
AND CAPTURED IMAGE OF EACH
‘SELECTED MICROBE
DISPLAY RETRIEVED 204
INFORMATION ON OUTPUT UNIT
¥
LET OPERATOR SELECT ONE OR 205
MORE OF SELECTED REFERENCE ~~
MICROBES FOR COMPARISON
HAVE ALL
REFERENCE MICROBES SELECTED
‘AT STEP 205 BEEN COMPARED.
WITH SAMPLE?
PRETREAT SAMPLE
OBSERVE AND CAPTURE IMAGE
OF SAMPLE UNDER RETRIEVED
IMAGING CONDITIONS
COMPARE CAPTURED IMAGE a
WITH THAT RETRIEVED FROM | —_-209
DATABASE AND DISPLAY
COMPARISON RESULT
ee)Patent Application Publication Mar. 10,2005 Sheet 3 of 7 US 2005/0051725 AL
FIG.3
IMAGE] IMAGING |PRETREAT-
NAME] 'Sata | CONDITIONS | MENT |SYMPTOMS| FEATURES.
ACCELERATION | STEP
‘| | VOLTAGE: 50kV-
MIUS||G) | MAGNIFICATION a
FEVER | SPHERICAL:
‘ABOUT 20 NM
IN DIAMETER
VOMITING | ELLIPSOIDAL
[=] [ACCELERATION |STEP1
DIARRHEA | SPECKLED
VIRUS| VOLTAGE:40kV_
@ MAGNIFICATION: a
Ge 404 401
PRETREATMENT
METHOD:
IMAGE CONDITIONS: ..Patent Application Publication Mar. 10,2005 Sheet 4 of 7 US 2005/0051725 AL
FIG.5A
128 PIXELS
mS
505 :
508 128 PIXELS
a
}r28 PIXELS
CALCULATE
FEATURE
VALUE VECTOR
Mt
\ \
503 502 —
FIG.5B
CALCULATE CALCULATE
FEATURE FEATURE
VALUE VECTOR | VALUE VECTOR
Ma 2Patent Application Publication Mar. 10,2005 Sheet 5 of 7 US 2005/0051725 AL
FIG.6
START
601
RECEIVE CAPTURED IMAGE [~~~
SE See
RECEIVE IMAGE OF REFERENCE 602
VIRUS (OR MICROBE)
RETRIEVED FROM DATABASE
Beams see tees tee tee seems tes cee ees ieee
FIND EACH VIRUS IMAGE WITHIN | —_-603
CAPTURED IMAGE
J
CALCULATE FEATURE VALUE 604
VECTOR OF epee VIRUS |}
ed
CALCULATE FEATURE VALUE 605
VECTOR OF EACH VIRUS IMAGE [~~
FOUND AT STEP 603
a
CALCULATE DEGREE OF
SIMILARITY BETWEEN FEATURE | __606
VALUE VECTOR OF EACH FOUND
VIRUS IMAGE AND THAT
OF REFERENCE VIRUS IMAGE
OUTPUT CALCULATED DEGREE 607
OF SIMILARITY aPatent Application Publication Mar. 10,2005 Sheet 6 of 7 US 2005/0051725 AL
FIG.7
706
REFERENCE VIRUS.
TO LS COMPARED: VIRUS Z
PRETREATMENT
METHOD:
IMAGING.
CONDITIONS:
DEGREE OF
SIMILARITY:
703 [NUMBER DEGREE OF] 708
PRETREATMENT SIMILARITY,
METHOD: 502 S| XxX
IMAGING 2 yy
CONDITIONS: 705
704Patent Application Publication Mar. 10,2005 Sheet 7 of 7 US 2005/0051725 AL
FIG.8
801 B01,
. C4 ~ E_|
-800
_ 117
118
FIG.9
901(A) 901(8) , 901(C)
902(A)
903(C)
I \ \_ |
903(A) '902(B) *903(B) ‘902(C)US 2005/0051725 AL
"TRANSMISSION ELECTRON MICROSCOPE,
SYSTEM AND METHOD OF INSPECTING A.
‘SPECIMEN USING THE SAME,
BACKGROUND OF THE INVENTION
The present invention relates to 4 transmission
croscope system for observing an obtained speci=
ten under 1 microscope to determine whether it contains
microorganisms and, iF it does, 10 identify theit types in
‘order to identify the cause of a disease or cary out food
jon supervision. The present invention also relates to
an inypection method using this transmission electton miero-
scope system,
{0002} When dealing with a disease or a food poisoning
‘case caused by a virus or bacterium, the frst thing to do is,
to identify the ease of the symptoms that the people oF
livestock are suffering, In such a ease, a Sample of fecal
matter, ele i gathered from a patient, et. and this sample
is checked tose if t contains aay vins or bacterium. Since
viruses and bacteria are very small, observing them usvally
requires a microscope which allows observation of minute
‘samples at high resalution, such asa transmission electron
microscope, Since transmission electron microscopes form
fan image using electrons transmitted through the sample,
they’ ate also suitable for observing substances having &
minute thee-dimensional stucture, sich a8 prot
[0003] For example, the H-7600 transmission electron
microscope catalog from Hitachi High-Techoologies Cote
poration lists traasmission eleciron microscopes for such &
Porpose. A transmission electon microscope i made up of
anelectron gun, an radiation system elect lens, a sample
holder stage, an imaging system lens, a camera, a vacuum
pumping system, 4 contol system, ete
[0004] When viruses in a specimen are observed using.
‘such microscope fecal matter or living tise, forexample,
's gathered ffom a test subject and pretreated 80 that it can
he actually observed under transmission electron micro-
scope a6 4 specimen.
[0005] Samples for observation are broadly classified into
three types: (1) negatively stained samples, 2) stained sliced
‘samples, and (8) frozen sliced samples.
[0006] (1) A acgatively stained sample is prepared by
purifying and concentrating living tissue, fecal mater, ete
hy use of a reagent or 4 centrifugal separator and then
‘mounting it on a mesh for eletron micrascopes. Represen-
tative examples of negatively stained samples ae fine pranv
lar specimens of viruses, et. Tungstic acid is usually used
as the stining agent. When this type of sample contains
Viruses, a levee (or a low wall) formed around each virus,
providing contrast between each virus and its levee.
[0007] (2) A stained sliced sample is prepared by slicing
living tssie of an animal o plat with a diamond cate, te
to produce 1 slice having 1 thickness of a few tens of
nanometers and then mounting the slice on a mesh for
‘electron microscopes. Before and when slicing the living
Tissue, it_must be subjected to processes such as fixing,
dehydrating, embedding, ad cutting. Generally, the tissue
ris he sttined so as to provide enough contrast Io observe
its survcture under an electroa microscope, Suitable staining
agents include reagents containing a heavy metal, sich a6
uranium acetate, lead citrate, lead hydroxide, and lead
Mar. 10, 2005
scctate. Usually, the tissue is double-stained with uranium
Sod lead. Staining is required since a living body is primarily
made up of light elements, such as hydrogen, oxygen,
tathon, and nitrogen, That is these elements only exhibit a
small scattering power with fespect to electron beams and,
forthermore, they differ oaly a lille in such power, resulting
{in very low image contrast. What is stained in the tissue is,
its protein; the higher the protein concentration, the more
heavily the tissue is stained. As a resull, the obtained
electron microscope image has a conteast according t0 the
protein concentration,
[0008] (3) A frozen sliced sample is prepared by bring
living tissue into contact with a copper block, which has.
been eooled down by liquid helium or liquid nitrogen so a8
to feze the tissue and then slicing the frozen iss by use
ff oooling stage and a microtome. This ype of sample was
Sevised to observe a tissue structure in an active state
Therefor, the tissue is neither fixed nor sained, aad itis
‘observed under #eryo-elecron microscope equipped with
cooling sample stage. Since the sample tissue i not stained,
the contrast of the obiained image is low.
[0009] A sample pretreated as described above for obser=
vation is mounted onthe stage of anclectron microscope and
observed by magnilyng its image a few tens of thousands to
' fow hundreds of thousinds of times. When identifying
‘microorganisms, suchas viruses contained in the simple, 4
person observes the magnified image to determine whether
it includes virus images based on the characteristic shape
snd internal structure of each vis, and, if the image
contains any virus image, to identify the type of virus
Further, when analyzing the structure of 2 protei(s) in the
sample, several magnified images are taken ofthe sample to
bre observed while the sample stage ofthe electron micro-
‘cope is being tilted. The taken images are subjected to CT
(computer tomogeaphy) processing to provide a fine three
dlimensionalstractare having dimensions on the order of &
few tens of nanometers
[0010] Recently, a cansmission electron microscope bas
been provided with (orto include) a contol personal com
puter having autofocus and imaging (contol) capabilities to
increase tbe eficency of the observation work. However,
the user must sill determine whether a sample contains
riiccoorganisms, sich a vires, snd to identify thei Iypes
bused on captured images. Acta, the user compares cach
possible virus image agains! images of previously found
Viruses listed in books and other documents o identify each
Jmaged microorganism (wins).
[0011] A problem with such » method is that it takes
‘onserable time to identify each virus, This is because the
Caphired image includes images of various things other than
microorganisms to be identified, making it necessary to
iseriminate each target microorganism from them. Further
‘more, manually searching a huge amount of past microoe
anism dlta requires substantial time and labor. Further,
there is another problem in that virus idenification may or
‘may not he accurate depending on the skill of the observer
or the degree of fatigue suffered by the observer
[0012] Furter, stil another problem arises when itis
necessary to handle a new virus, ete. That is, viruses and
microbes grow and mutate into new forms, meaning that a
‘new (undiscovered) virus may suddenly appear. AS a result,
the documents at hand might not lista virus curently underUS 2005/0051725 AL
‘observation, which may make it necessary to consume a
Targe amount of time t identity the views.
[0013] A similar problem arses, for example, when it is
recessary to determine whether a ease ofa disease found in
Japan has been caused by a microorganism which usually
‘exists locally in some other part of the world, such as Africa,
When this occurs, a sample is usually seat large research
institutes in the world to identify the microorganism which
has caused the ease. This may lead fo delay in identifying
tnd handling the microorganism,
‘SUMMARY OF THE INVENTION
[0014] The present invention provides. transmission elec
tson microscope systom that is capable of easily determining
‘whether a specimen contains mieroorganisms sad, if it does,
‘enifying their types, thereby solving the above-described
an problems. The present invention alsa proves an inspee-
tion method using sucha transmission electron microscope
system,
[0015] Further, the present invention also provides 3 net-
work system capable of quickly determining whether a virus
in specimen is identical to a vies nowly discovered at 3
remote location by gathering information on the new virus
through a wide area network,
[0016] _A transmission electron microscope systom of the
present invention comprises: means for accessing a database
Which stores transmitted electron images of difieent types
‘of microorganisms and specimen pretreatment methods and
aging conditions for capturing the transmitted electron
images; an input unit for inputting « keyword for selecting
a microscopic thing whose presence ar absence ina specie
men isto be determined; means for receiving a transmited
‘leetron image of the mietoscopie thing and specimen
prireatment method and an imaging condition selected
based om the input keyword; an imaging unit for capturing.
an image of the specimen under the received imaging
‘condition after the specimen is pretreated by’ the received
pretreatment method; a comparison tit for comparing the
image captured by the imaging unit with the image received
from the database; and an output unit for outputting @
‘comparison result reacived from the comparison uni
[0017] Further, the present invention also provides 4 net-
work system in whieh the above-mentioned. transmission
‘electron microscope system and one or more database
‘reation sites ate comnected fo 4 wide area network. This
system is characterized in that the transmission electron
microscope system accesses a database created from all or
‘some of the original databases stored in the above-men-
tioned database creation sites
[0018] The transmission electron microscope system of
the present invention makes it easy to determine whether 8
specimen contains microorganisms and, iit does, to identify
their types. Furthermore, iis possible to quickly determine
whether a virus in a specimen is identical to a vis newly