Novel Wireless Capsule Endoscopy Diagnosis System

with Adaptive Image Capturing Rate

Zhi Jin, Tammam Tillo, Eng Gee Lim, Zhao Wang and Jimin Xiao

Department of Electrical and Electronic Engineering, Xi’an Jiaotong-Liverpool University, Suzhou, China

Keywords:

Wireless Capsule Endoscopy, Smart Image Capturing Rate, Image Recognition Technique.

Abstract:

Wireless Capsule Endoscopy (WCE) is a device used to diagnose the gastrointestinal (GI) track, and it is one

of the most used tools to inspect the small intestine. Inspection by WCE is non-invasive, and consequently it

is more popular if compared to other methods that are traditionally adopted in the examination of GI track.

From the point of view of the physicians, WCE is a favorable approach in increasing both the efﬁciency and

the accuracy of the diagnosis. The most signiﬁcant drawback of WCE is the time consumption for a physician

to check all the frames taken in the GI track, in fact it is too long, and could be up to 4 hours. Many anomaly-

based techniques were proposed to help physician shorten the diagnosis time, however, these techniques still

suffer from high false alarm rate, which limits their actual use. Therefore, in this paper we propose a two stage

diagnosis system that ﬁrstly uses a normal capsule to capture the whole GI track, and then we use an automatic

detection technique that detects anomalies with high false alarm rate. The low speciﬁcity of the ﬁrst capsule

ensures that no anomalies will be missed in the ﬁrst stage of the process. The second stage of the proposed

diagnosis system uses a different capsule with adaptive image capturing rate to re-capture the GI tract. In this

stage the capsule will use high image capturing rate for segments of GI tract where an anomaly was detected

in the ﬁrst stage, whereas, in the other segments of the GI tract a lower image capturing rate will be used in

order to have better use of the second capsule’s battery. Consequently, the second generated video, which will

be inspected by the physician, will have higher resolution sequence around the areas with suspected lesion.

1 INTRODUCTION

The history of endoscope use to inspect internal or-

gans can be traced back to the 19th century, in

1806, when German scientist Philipp Bozzini ﬁrst in-

vented the endoscope to inspect human bladder and

bowel(Litynski, 1996). Since then, different types

of endoscopy tools are constantly developed and im-

proved, so for example the gastroscopy is used to de-

tect gastric lesions, and colonoscopy for the intesti-

nal lesions. In the past ten years endoscopy-related

technology developed rapidly, and one of the most

promising technology in this ﬁeld is Wireless Cap-

sule Endoscope (WCE). One of the beneﬁts of WCE

is that it serves inspecting the stomach and small in-

testine for illnesses. Unlike the usage of conventional

endoscopic methods, in the application using WCE,

there is no hose, which enables the patient to main-

tain normal life activities. Moreover, since the cap-

sule endoscope is small enough to be swallowed, it

signiﬁcantly reduces the pain caused by traditional

endoscope to the patient, even in comparison with en-

doscope with soft pipes. Furthermore, in comparison

with endoscopy, colonoscopyand other traditional en-

doscope, where due to hose length and bending re-

strictions which limits the depth range of inspection

of the human body, the WCE could be used to inspect

the small intestine. Consequently, it becomes one of

the most effective tools to check the whole section

of the 5-7 meters small intestine(Triester et al., 2006),

where it has been approved by the medical profession,

due to its convenience, hygiene, and effectiveness.

The WCE is a capsule shaped device equipped

with small-sized electronic circuitry, which includes

the built-in LED that light the internal of the GI tract,

the imaging system which captures images, a vari-

ety of sensors, battery, the transmitter module and the

antenna and some other supporting components.The

most popular WCE, developed and manufactured by

Given Imaging (Given, ). Other manufacturer, such

as Olympus Pharmaceutical Company produces their

own capsule. This latter produces the M2A-capsule

endoscope (Olympus, ), with a size of 11 × 27mm.

After being swallowed, it continuously works for 7 to

143

Jin Z., Tillo T., Lim E., Wang Z. and Xiao J..

Novel Wireless Capsule Endoscopy Diagnosis System with Adaptive Image Capturing Rate.

DOI: 10.5220/0004346201430147

In Proceedings of the International Conference on Computer Vision Theory and Applications (VISAPP-2013), pages 143-147

ISBN: 978-989-8565-47-1

 2013 SCITEPRESS (Science and Technology Publications, Lda.)

8 hours in the human body. During this period the

WCE camera will take pictures of the digestive sys-

tem at the rate of two frames per second, and even-

tually captures approximately 50, 000 color pictures.

The capsule shell is made of special biological mate-

rials resistant to stomach acid and powerful digestive

enzymes, and it moves , inside the GI tract, thanks to

the gastrointestinal peristalsis.

In general, there are some limiting factors to the

WCE, the ﬁrst one is the low quality of the obtained

images by the capsules. This limits the effectiveness

of the image processing techniques, and consequently

it may increase the probability of misdiagnosis. The

main constraint for not being able to increase the res-

olution is the limited capacity of the battery. The sec-

ond limiting factor comes from the fact that the move-

ment of the capsule, inside the GI tract, dependentson

the peristalsis, which means that the instant capsule

speed varies due to the different internal structure of

different individuals. Therefore, the current capsule

system, with the existing image processing methods

and image capturing rate cannot ensure high accuracy

of diagnosis. In fact, when the capsule moves fast,

misdiagnosis may exist.

In (Ping et al., 2011) a comparative study between

the capsule endoscope and the double-balloon en-

teroscopy was carried out, the comparison is in terms

of clinical and economic impact of both methods. In

this study it was reported that the success rate of diag-

nosis is only 81.73% when the WCE was only used.

However, it is stated that the accuracy will increase to

90.56% with the help of double balloon endoscopy.

To overcome some of the limitation of current

WCE, intelligent image processing method could be

used to improvethe rate of successful diagnosis.Given

Imaging Ltd, published a patent (Podilchuk, 2007),

in which feature matching is applied on a number

of captured images with the WCE and an existing

database of images. This technique can be summa-

rized as following, the captured images by the WCE

are compared and matched with a database of im-

ages, using processing system that relay on lesions

features in both sets of images. Then it highlight the

images which are consistent with the required fea-

tures. Thereby, the sensitivity of the system to the sus-

pected lesions has been increased, and this helps doc-

tors by shortening the detection time. However, this

method still cannot overcome the weaknesses caused

by the fast moving of the capsule in the areas with

suspected lesions, and the limited number of captured

pictures when the image capturing rate is ﬁxed. Thus,

although this method improves detection, it still has

large possibility of misdetection.

To have more captured images is also an effective

way to improve the detection rate of lesions. How-

ever, the storage and capturing capability of the cap-

sule itself is limited. In fact, how to effectively ad-

just the image-capturing rate without a substantial

increase in the energy consumption of the capsule,

is one important factor that determines capsule efﬁ-

ciency. In the patent (Han et al., 2010), a technique

to adjust the image capturing rate is proposed based

on the observation that the speed of movement of the

capsule, inside the GI tract, is different in different

organs. So ﬁrstly, the time required for the capsule to

reach different organs will be estimated. Then, via a

built-in chip, the image capturing rate is adjusted once

it reaches a certain preset time. This method compen-

sates the fast movement of the capsule in some di-

gestive organs, therefore, the possibility of successful

diagnoses of illnesses is increased. However, tuning

the image capturing rate of this method is not based

on suspicious lesion areas. That is to say, the num-

ber of generated images for the lesions and the nor-

mal regions is the same in the same organ. So this

will increase the burden of post-image processing and

screening.

Thus, in general, the existing WCE is still unsatis-

factory, due to many reasons, among them the image

quality, and the image capturing rate which cause high

probability false detection and misdetection. There-

fore, this becomes the major obstacle for doctors to

detect the patient’s condition quickly and effectively.

This paper proposes a new WCE system with adaptive

image capturing rate, based on two capsules inspec-

tion paradigm. Thus the image capturing rate will be

tuned during the second stage of inspection, based on

the outcomes of the inspection with the ﬁrst capsule.

The ﬁrst inspection serves to identify the suspected

lesion regions.

2 NEW PROPOSED

METHODOLOGY

The proposed WCE system and detection method

aims to improve the amount of effective information

and increase the temporal resolution of the captured

video in the suspicious tracts of the digestive system.

This objective is achieved by adjusting the image cap-

turing rate of the capsule. Fig.1 shows the ﬂowchart

of the proposed diagnosis system, which aims through

the use of the ﬁrst capsule to detect any suspected

lesion regions through the use of image recognition

technique, and consequently to appropriately adjust

the capture rate of the second capsule. The details of

main elements of the system as shown in Fig.1 are as

following: D11 is the ﬁrst swallowed capsule, which

VISAPP2013-InternationalConferenceonComputerVisionTheoryandApplications

144

is a traditional capsule with ﬁxed image capturing rate

and it includes a wireless transmitter module, and this

capsule is used during the ﬁrst diagnosis stage. D12 is

the second swallowed capsule whose capture rate can

be modulated, and it contains a wireless transmitter

and receiver module. This capsule is used for the sec-

ond stage of the diagnosis procedure, and this stage

could start after completing the ﬁrst diagnosis stage,

or while it is still running. D1 and D2 are the pa-

tient carried built-in receiving apparatus with antenna

arrays, D3 and D4 are the storage units for captured

images, D5 and D6 are image feature identiﬁcation

devices, D7 is used for storing the captured imaged

after image feature matching with image recognition

system database, D8 is capsule capture rate controller

and D9 is a transmitter.

Shown in the drawing process, the patient needs to

swallow the capsule 1, D11, which can transmit sig-

nal to external body device. The transmission mod-

ule inside can convert the captured images into wire-

less signals and then transmit them across the human

body to the patient carried receiving apparatus D1.

Images stored in the storage unit D3 will be deter-

mined whether has feature consistency with the fea-

ture images of various types of gastrointestinal lesions

by D5. Then these matched images are marked and

the two previous images of the matched image will

also be marked. After swallowing the ﬁrst capsule,

D11 for a certain period of time, patient can swal-

low capsule 2, D12. Different from the Capsule1, the

Capsule2 can be controlled by receiving external cap-

ture rate changing signal to adjust the capture rate in

different intestinal regions. When D2 receives image

signals captured by the Capsule2, these images will

be identiﬁed by the image features device D6 and the

results of comparison with marked images Capsule1.

Once the image captured by capsules 2 is consistent

with the feature image, the rate controller D8 sends

the signal to speed up the capture rate of the capsule.

Fig.2 shows a schematic diagram of the existing

design of the capsule; this kind of capsule will be used

in the ﬁrst stage of the diagnosis. In this ﬁgure, 1 point

to the built-in optical system of a capsule, 2 stands for

an image sensor, 3 is the image sensor controller, 4 is

built-in capsule microprocessor, 5 is an illumination

system, 6 is power supply device to provide energy

to the various components of capsule, 7 is a wireless

transmitting device, 8 is an antenna. The built-in mi-

croprocessor controls the wireless receiving module

7 and the image sensor controller 3. Meanwhile, the

image sensor controller 3 controls the image sensor 2.

The wireless transmitter module 7 transmits the sig-

nal to the patient carried receiving device, for further

analysis and processing by the operation processing

system.

Fig.3 shows a schematic diagram of the Capsule2,

which is used in the second stage of the proposed di-

agnosis process. In this ﬁgure, 1 represents the built-

in optical system of the capsule, 2 is the image sen-

sor, 3 is the image sensor controller, 4 is built-in cap-

sule microprocessor, 5 is an illumination system, 6 is

power supply device to provide energy to the various

components of capsule, 7 is a wireless transmitting

device, 8 is an antenna. Different from the Capsule1,

in the Capsule2 the wireless receiver and transmitter

module 7 not only transmits the captured images to

the carried receiver,but also serves to receivean exter-

nal control signal to adjust the image capturing rate,

thereby changing the capsule capture rate in the dif-

ferent tracts of the human digestive system. In Cap-

sule2 the built-in microprocessor controls the wireless

receiving module 7 and the image sensor controller 3,

and the illumination system 5.

Image matching

㜦೺1

Receiver

Optical

module

Transmitter

㜦೺2

Optimal

module

Transmitter

Image matching

Receiver

(Capsule 1)

Receiver

(Capsule 2)

Storage

unit

Storage

unit

image

feature

identificatio

n device

image

feature

identificatio

n device

Storage unit

(

suspicious lesion

area images

)

Capture

rate

controller

Transmitter

Time interval

D1 D3

D11

D12

Capsule 2

Capsule 1

Figure 1: The block diagram of the overall system.

1237

4 5

Figure 2: The main structure of Capsule1.

1237

4 5

Figure 3: The main structure of Capsule2.

NovelWirelessCapsuleEndoscopyDiagnosisSystemwithAdaptiveImageCapturingRate

145

A B C D E

No.

Mark

Two before the

image

One before the

image

Lesion region

image

Lesion region

image

Normal

Figure 4: The adopted method for marking the suspected lesion area in few consecutive frames.

3 IMPLEMENTATION

As shown in Fig.1, the captured images obtained by

the ﬁrst capsule are transmitted through the human

body to the external receiving apparatus D1, those im-

ages are stored in the storage unit, D3. Then every

photo stored in D3 will be analyzed to detect vari-

ous gastrointestinal lesions by the image processing

system; this system will be tuned so as to have low

speciﬁcity, in order to increase the chances of detect-

ing lesions. Moreover, to reduce the probability of

missing any lesions the system will mark few frames

before and after the frame that contains the suspected

lesions and these images will be stored into the stor-

age unit D7. For example, let us suppose that among

the frames f

, f

, and f

, the frame f

is suspected

of having some lesions, in this case all the three im-

ages will be stored into storage unit D7. As for the

Capsule2, D12, the patent will swallow it after a pe-

riod of time of swallowing Capsule1. At beginning,

Capsule2 will take pictures by low default speed (this

speed could be set by doctors). Then the transmit-

ted images by Capsule2 will be compared with those

stored in D7. So, when some transmitted images are

positively matched with the feature images in D7, the

image capturing rate controller, D8, will transmit the

signal to wireless capsule endoscope D12 to increase

the capturing rate through transmitter D9. Therefore,

the capsule can take pictures in suspected lesion area

in a faster rate than before. Furthermore, when there

is no more match between transmitted image by Cap-

sule2 and feature image, the capture rate controller

D8 will transmit new signal to change the capture rate

back to the normal low speed in order to save battery’s

energy.

The process of lesion marking is shown in Fig.4.

So, let us suppose that in series of captured images

of digestive tract from frame A to E, there is one le-

sion area in image C and D (region 2: represented as

the dark area). In this case, when the image feature

identiﬁcation device detects the lesion in picture C

and D, in the video sequence generated by Capsule1,

the picture A, B, C and D will all be marked. All

marked images will be stored in a special database of

image feature recognition apparatus, and when using

the second capsule, i.e., Capsule2, all the new cap-

tured images will be compared with marked images

database directly. At this point if some captured im-

ages by the second capsule match the marked images

of the ﬁrst capsule, then a signal is transmitted to in-

crease the Capsule2 image capturing rate around the

areas with suspected lesions. This will ensure that

higher quality images will be generated around those

areas, for latter analysis by the physician. When the

built-in reception module in Capsule2 receives a rate

control signal P25, as shown in Fig.7, the capsule will

increase the imaging rate, otherwise it will keep the

normal image capturing rate.

Using the image feature

matching apparatus

Start

Yes

Matched with lesion

characteristics images

Receive the data

Yes

Marking the captured images

Store the two images before the

matched images into image

feature matching apparatus

End

P11

P12

P13

P14

P15

Figure 5: The ﬂow chart of the working principle of the ﬁrst

capsule.

The ﬂow chart reported in Fig.5 shows the work-

ing principle of the ﬁrst capsule. When the data is

received, the image feature matching apparatus will

be used to match the captured image, by the capsule,

with the lesion characteristics images P12. In this

matching process, each captured image will be ana-

lyzed to see whether it matches the stored feature im-

ages in P13. If a match is found these images will

be marked, P14, and stored into another database of

image feature, P15. In Fig.6, the ﬂow chart of the

VISAPP2013-InternationalConferenceonComputerVisionTheoryandApplications

146

working principle of the second capsule is reported.

This has similar working paradigm with the ﬁrst cap-

sule, except for the image capturing speed, which is

tuned by D8 (in Fig.1). This latter unit will transmit

a signal to the second capsule to increase the image

capturing rate when a match is found with the stored

feature images P13.

Yes

Start

Receive the data

Using the image feature

matching apparatus

Received captured image

matched with two images

before the marked image

Yes

Transmit the capture rate

control signal

End

P21

P22

P23

P24

Figure 6: The ﬂow chart of the working principle of the

second capsule.

੖

ᱟ

NoYes

Start

Receive the capture rate

control signal

Increase capture rate

End

P25

P26

Normal capture rate

Figure 7: The control mechanism of the image capturing

rate in Capsule2.

4 CONCLUSIONS

This paper proposes a new wireless capsule endo-

scope system, which comprises two diagnosis stages.

The ﬁrst one uses a normal capsule to capture the

whole GI track. The generated video sequence will

be analyzed to detect anomalies with high false alarm

rate. The second stage of the proposed diagnosis sys-

tem uses a different capsule with adaptive image cap-

turing rate to re-capture the GI tract. In this stage the

capsule will use high image capturing rate for seg-

ments of GI tract where an anomaly was detected in

the ﬁrst stage, whereas, in the other segments of the

GI tract a lower image capturing rate will be used

in order to have better use of the second capsule’s

battery. Consequently, the second generated video,

which will be inspected by the physician, will have

higher resolution sequence around the areas with sus-

pected lesion. Therefore, this new type of wireless

capsule endoscope system and method will be more

efﬁcient for clinical applications.

ACKNOWLEDGEMENTS

This work is partially supported by the Natu-

ral Science Foundation of Jiangsu province (No.

BK2010251 and BK2011352), Suzhou Science

and Technology Bureau (No.SYG201011 and

SYG201211), and XJTLU Research Development

Fund (No.10-03-16.).

REFERENCES

Given, I. http://www.givenimaging.com.

Han, Z.and Binghe, J., Yongyou, L., and Tie, C. (2010). Im-

age capture rate controllable capsule endoscope. State

Intellectual Property Ofﬁce of the P.R.C. Native Lan-

guage Patents (SIPO).

Litynski, G. S. (1996). Highlights in the History of

Laparoscopy: The Development of Laparoscopic

Techniques– a Cumulative Effort of Internists, Gyne-

cologists, and Surgeons. Barbara Bernert Verlag.

Olympus. http://www.olympus-global.com/en/global.

Ping, W., Hong, F., Qiang, G., and et al (2011). Clinical

and health economic evaluation of the capsule endo-

scope (ce) or/and double-balloon enteroscopy (dbe) in

diagnosis of intestine disease. China Journal of En-

doscopy, 17.

Podilchuk, C. (2007). Method and system for image recog-

nition using a similarity inverse matrix. United States

Patent application publication (US 2008/0133521

A1).

Triester, S. L. J., G., L., Gurudu, S., and et al (2006). meta-

analysis of the yield of capsule endoscopy compared

to other diagnostic modalities in patients with non-

stricturing small bowel crohns disease. Am J Gas-

troenterol, 101:954–964.

NovelWirelessCapsuleEndoscopyDiagnosisSystemwithAdaptiveImageCapturingRate

147