The 4th ASIAN AND PACIFIC RIM SYMPOSIUM ON BIOPHOTONICS

www.apbp2009.org

Date : 27-29, May 2009

Location : Shilla Hotel, Jeju island, Korea

Organized by : Optical Society of Korea / Biomedical Optics Division

Sponsored by : Korea Optical Industry AssociationAdvanced Photonics Research Institute, GISTGraduate-program of Medical System Engineering(GMSE), GISTChungbuk BIT Research-Oriented University ConsortiumIntelligent System Research Division, KISTSeongKeong PhotonicsHANA engineering Co., Ltd.Huentek Co., Ltd.Fovice Co.Ainnotech Inc.

www.apbp2009.org www.apbp2009.org

Welcome to APBP 2009

On behalf of the Organizing committee and the Technical Program Committee, it is our great pleasure to welcome you

to the 4th Asian and Pacific-Rim Symposium on Biophotonic (APBP 2009) at the fantastic Jeju Island, Korea. The APBP

conference, started at 2001 in Japan, has been held every two years (except for the 3rd APBP in Cairns, Australia).

Our field of Biophotonics has gained a huge growth over the last couple of years and we believe that the APBP 2009 has

supported great places for tracking that progress especially in Asian and Pacific-Rim region. This year, we have over 150

contributed papers on various topics including Optical Coherence Tomography, Optical microscopy & nanoscopy, Optical

spectroscopy, Photonic therapeutics, Molecular/cellular imaging & manipulation, Diffuse optics for biomedical applications,

Nano-biophotonics, Optical sensing & measurement, and Optical devices & systems for bioscience. Especially we are

excited to invite two distinguished plenary speakers, Prof. Luke Lee from UC Berkeley and Prof. David Boas from Harvard

University.

We would like to thank all the people who worked so hard to make the APBP 2009 a success. The executive committee

members, international steering committee members, organizing committee members, advisory committee members,

and technical program members have worked so hard behind scenes to keep everything on schedule.

We are excited about this year’s conference and we hope you will all feel that your time here was well-spent. We encourage

every participant to enjoy the relaxed and open atmosphere and go home with warm memories about this conference

and Korea.

Welcome to Jeju, Korea!

April 24, 2009

Co-chair Beop-Min Kim

Associate ProfessorDept. of Biomedical Engineering,

College of Health Science,Korea University,

Korea

Co-chairByeong Ha LeeProfessorDept. of Information and Communications, Gwangju Institute of Scienceand Technology, Korea

Beop-Min Kim Byeong Ha Lee

www.apbp2009.org

Committee

Advisory committeeMembers : Dong-Seob Ko (Mokwon University, Korea)

Kie-Bong Nahm (Hallym University, Korea)Seung-Han Park (Yonsei University, Korea)Kwang-Sup Soh (Seoul National University, Korea)Gilwon Yoon (Seoul National University of Technology, Korea)

Organizing committee - Chair: Beop-Min Kim (Korea University, Korea)Members : Chang-Seok Kim (Pusan National University, Korea)

Donghyun Kim (Yonsei University, Korea)Eun Seong Lee (Korea Research Institute of Standards and Science, Korea)Eun Seo Choi (Chosun University, Korea)Ho Lee (KyungPook National University, Korea)Jaebum Choo (Hanyang University, Korea)Jee Hyun Choi (Korea Institute of Science and Technology, Korea)Jeehyun Kim (KyungPook National University, Korea)Jong Chul Ye (Korea Advanced Institute of Science and Technology, Korea)Minah Suh (Sungkyunkwan University, Korea)Seung Hee Han (The Catholic University of Korea, Korea)Sungchul Hohng (Seoul National University, Korea)Tae Jung Eom (Advanced Photonic Research Institute, Korea)

Program Committee - Chair: Byeong Ha Lee (Gwangju Institute of Science and Technology, Korea)Members : Arthur Chiou (National Yang-Ming University, Taiwan)

Chang-Seok Kim (Pusan National University, Korea)Eric P. Y. Chiou (University of California, USA)Eun Seo Choi (Chosun University, Korea)Ho Lee (KyungPook National University, Korea)Jaebum Choo (Hanyang University, Korea)Jeehyun Kim (KyungPook National University, Korea)Ki-Hun Jeong (Korea Institute of Science and Technology, Korea)Kin F. Chan (Fourier Biotechnologies, USA)Koichi Shimizu (Hokkaido University, Japan)Lai Kwan Chau (National Chung Cheng University, China)Minah Suh (Sungkyunkwan University, Korea)Min Gu (Swinburne University of Technology, Australia)Satoshi Kawata (Osaka University, Japan)Seung-Han Park (Yonsei University, Korea)Sungchul Hohng (Seoul National University, Korea)Taner Akkin (University of Minnesota, USA)Yoshiaki Yosuno (University of Tsukuba, Japan)

International Steering Committee Members : Bernard Choi (Beckman Laser Institute, University of California, Irvine, USA)

C. C. Yang (National Taiwan University, Taiwan)Eiji Okada (Keio University, Japan)Gert von Bally (University of. Münster, Germany)Halina Rubinsztein-Dunlop (University of Queensland Brisbane, Australia)Hans - A. Bachor (Australian National University, Australia)Koichi Shimizu (Hokkaido University, Japan)Masamitsu Haruna (Osaka University, Japan)Mamoru Tamura (Hokkaido University, Japan)Min Gu (Swinburne University of Technology, Australia)Shunichi Sato (National Defence Medical College, Japan)T. K. Alex (Director, ISRO Satellite Centre (ISAC), India)Yukio Yamada (University of Electro-Communications, Japan)Zhizhan Xu (Chinese Academy of Sciences, Shanghai, China) APBP20

09

www.apbp2009.org

Address : 3093-3, Saekdal-dong, Seogwipo-si, Jeju-do, 697-808, Korea

Telephone : +82-64-735-5114

FlightsAccessible by air from Seoul, Busan, Daegu, Gwangju, Ulsan, Cheongju, Jinju, Gunsan, Mokpo, and Yeosu.

It takes between 40 minutes and 1 hour 10 minutes to get to this island.

Airport Limousine Service Route

Airport - Crown Plaza Hotel - Yeomiji Botanical Garden - Hana Hotel - Hyatt Hotel – The Shilla Jeju

RegistrationPlease make online registration and fill in the registration form on APBP2009 website: www.apbp2009.org

The registration fee includes 2 lunch and 2 dinner servings at the Shilla Jeju Hotel.

If you apply for OSK membership, you can receive the OSK discount benefit. The OSK membership application is available on site.

Location and Transportation

Destination Distance km (miles) Ride Time (Fare) Transportation

Airport 37 (23)50 min (￦3,900) Limousine Bus (No.600)

35 min(￦30,000) Taxi

Jeju Port 37 (23) 45 min (￦35,000) Taxi

Pre-registration Fee Regular Student

OSK Member ￦500,000 ￦250,000

Non-OSK Member ￦700,000 ￦350,000

On-site registration Fee Regular Student

OSK Member ￦600,000 ￦300,000

Non-OSK Member ￦800,000 ￦450,000

APBP Technical ProgramWEDNESDAY, May 27

10:00-12:00 Registration

12:30-13:00

Opening Ceremony (Lotus Hall 1,2,3)Opening Address : Prof. Beop-Min Kim (Co-chair of APBP2009)Congratulatory Address : Prof. Bum Ku Rhee (President of Optical Society of Korea)Welcome Address : Prof. Yongjin Shin (Agency Leader of Gwangju Regional Innovation Agency)

13:00-13:45 Plenary Talk I Satellite Nanoscope and OASIS’ for Molecular MedicineProf. Luke Lee (UC Berkeley, USA), Session Chair : Prof. Ki-Hun Jeong

13:45-14:30 Plenary Talk II Optical Imaging of the Cerebral Metabolic Rate of Oxygen Satellite Prof. David Boas (Harvard Medical School, USA), Session Chair : Prof. Dong Hyun Kim

14:30-15:00 Coffee BreakRoom Lotus 1 Room Lotus 2 Room Lotus 3

Biomedical Diagnostics Using Nanoprobe-Based Optical Spectroscopy(I)Session Chair : Prof. Kaoru Sakatani & Prof. Joon Myoung Song

Diffuse Optics for Biomedical Applications(I)Session Chair : Prof. Minah Suh

Novel Optical Sensing(I)Session Chair : Prof. Jeehyun Kim

15:00-15:30SPE-01 [Invited] Plasmonic nano-bio-sensors, Lai-Kwan Chau et. al. (National Chung Cheng University, Taiwan)

DOT-01 [Invited] Time-resolved spec-troscopy for quantitative characterization of optical properties in tissue-simplified methodologies, Goro Nishimura (Hokkaido University, Japan)

SEN-01 [Invited] Depth-resolved optical detection of action potential related tran-sient deformations and dye signals, Taner Akkin (Minnesota Univ., USA)

15:30-16:00

SPE-02 [Invited] DNA conjugate polymers and nanoparticles for rapid and reliable gene sensing, Mizuo Maeda (The Research of Physical and Chemical Research, Japan)

DOT-02 [Invited] Quantitative imaging of neurohemodynamics using NIRS-SPM,Jong Chul Ye et. al. (KAIST, Korea)

SEN-02 [Invited] Scientific and engineer-ing applications of full-field swept-source optical coherence tomography, Dalip Singh Mehta (Indian Institute of Technology, India)

16:00-16:15SPE-03 [Invited] Application of bio-functionalized metal nanoprobes for medical diagnosis, Jaebum Choo (Hanyang University, Korea)

DOT-03 Extraction of weakly scattered light for transillumination imaging, Kazuto Takagi et. al. (Hokkaido University, Japan)

SEN-03 [Invited] Phase-sensitive surface plasmon resonance biosensors- multi-analyte arrays and wide dynamic range, Aaron Ho (Chinese University of Hong Kong, China)16:15-16:30

DOT-04 Experimental evaluation of spatial resolution of near infrared topography by using a head phantom, Hirokazu Kakuta et. al. (Keio University, Japan)

16:30-16:45 Coffee BreakBiomedical Imaging Using Optical Spectroscopy(II)Session Chair : Dr. Mizuo Maeda & Prof. Jaebum Choo

Diffuse Optics for Biomedical Applications(II)Session Chair : Prof. Jong Chul Ye & Prof. Goro Nishimura

NIR Sensing(II)Session Chair : Prof. Taner Akkin

16:45-17:15

SPE-04 [Invited] Differential hemodyanmic re-sponse between brain and muscle tissue during novel cyanide antidote cobinamide treatments as observed by near infrared spectroscopy, Jae G Kim et. al. (Univ. of California at Irvine, USA)

DOT-05 [Invited] Transillumination imaging with scattering suppression technique, Koichi Shimizu (Hokkaido University, Japan)

SEN-04 [Invited] MEMS mirror applica-tions in biomedical engineering, Daniel McCormick (AdvancedMEMS, USA)

17:15-17:45SPE-05 [Invited] Quantitative cellular imaging-based anticancer drug screening, Joon Myoung Song (Seoul National University, Korea)

DOT-06 [Invited] Diffuse Optical Tomography and Spectroscopy of Disease in Brain and Breast, Regine Choe et. al. (Univ. of Pennsylvania, USA)

SEN-05 [Invited] In vivo photoacoustic im-aging and its clinical application, Chulhong Kim et. al. (Washington University, USA)

17:45-18:15

SPE-06 [Invited] Instrumentation of multi-plex coherent anti-Stokes Raman scattering micro-spectroscopy with femtosecond broadband pump beam and its biomedical application, Eun Seong Lee et. al. (KRISS, Korea)

DOT-07 Effects of spatial variation of skull and cerebrospinal fluid layers on optical mapping of brain activities, Shuping Wang et. al. (Univ. of Electro-Communications, Japan)

SEN-06 [Invited] Dual wavelength polarim-etry toward noninvasive glucose sensing, Gerald Cote et. al. (Texas A&M University, USA)DOT-08 Ultrasound-modulated optical

tomography, Chulhong Kim (Washington University, USA)

18:30-20:30 Banquet(Room Lotus Foyer)

THURSDAY, May 28Room Lotus 1 Room Lotus 2 Room Lotus 3

Molecular/Cellular Imaging(I)Session Chair : Prof. Yuji C. Sasaki & Prof. Tae-Young Yoon

Novel Technology-Based Optical Coherence Tomography(I)Session Chair : Prof. Eun Seo Choi & Prof. Yoshiaki Yasuno

Plasmonic and Biophotonics Sensing(III)Session Chair : Dr. Chulhong Kim

09:00-09:15 MOL-01 [Invited] Direct visualization of dynamic biomolecular processes by high-speed AFM, Toshio Ando (Kanazawa University, Japan)

OCT-01 [Invited] Fourier domain optical imaging and endoscopic applications, Zhi-hua Ding et. al. (Zhejiang University, China)

SEN-07 Evaluation of crosstalk resistant noise canceller for extraction of brain sig-nal in near-infrared spectroscopy, Wataru Matsui (Keio University, Japan)

09:15-09:30SEN-08 Hamiltonian based metric for bio-medical image evaluations, Chung-Jen Oh (Hsiuping Institute of Technology, Taiwan)

09:30-09:45MOL-02 [Invited] Super-accurate dynami-cal single molecular observations using high-energy probes, Yuji C. Sasaki (The University of Tokyo, Japan)

OCT-02 [Invited] High speed and high res-olution endoscopic optical coherence and multiphoton tomography, Zhongping Chen (University of California at Irvine, USA)

SEN-09 In situ fluorescence detection system of 3D microfluidic cell culture devices, Jong-ryul Choi et. al. (Yonsei University, Korea)

09:45-10:00

SEN-10 Solution for fraction estimation of small dense LDL by dynamic light scatter-ing, Suchin Trirongjitmoah et. al. (Hokkaido University, Japan)

www.apbp2009.org

17:00-19:00, Thursday, May 28APBP Poster presentation Molecular/Cellular Imaging and Manipulation[MOL-P1] ALEX three-color FRET combined with optical twee-zers, Sanghwa Lee (Korea)[MOL-P2] Single-molecule FRET combined with magnetic twee-zers, Heesoo Uhm (Korea)

Diffuse Optics for Biomedical Applications[DOT-P1] Preclinical study of extremities metabolic detection with a venous occlusion test based on diffuse optical tomogra-phy, Yo-Wei Lin, et al. (Taiwan)[DOT-P2] Effects of the anatomy of the head on optical mapping of brain activities, Ryo Togashi, et al. (Japan)[DOT-P3] Reduction of Poisson noise from time-resolved data for diffuse optical tomography, Shinpei Okawa, et al. (Japan)[DOT-P4] Optimisation of algorithm for blood-flow measurement by laser speckle method, Haruka Nakayama, et al. (Japan)[DOT-P5] Effect of the optical properties of tissue, Yosuke Taka-hashi, et al. (Japan) [DOT-P6] Phantom experiments of fluorescence diffuse optical tomography, Yano Akiraet, et al. (Japan)[DOT-P7] Effect of the movement of probes on the images of dif-fuse optical tomography, Syoko Matsuhashi, et al. (Japan)[DOT-P8] Realistic skin and lip models for simulation of diffuse reflectance spectra, Wakana Fujita, et al. (Japan)[DOT-P9] Optical density spectroscopy of oral cancer ex-vivo using a spatially mapping fiber-optic probe for early diagnosis, Youngjin Oh, et al. (Korea)

Optical Sensing and Measurement[SEN-P1] Experimental study on estimation parameters in bio-speckle blood flow imaging, Yoshihisa Aizu, et al. (Japan)[SEN-P2] A background-compensated reflection-type pulse oximeter using a gain-enhanced gated avalanche photodiode, Tsuyoshi Miyata, et al. (Japan)[SEN-P3] Characterization of wavelength swept laser based on fiber F-P tunable filter, Byoung Chang Lee, et al. (Korea) [SEN-P4] Measurement of charge transfer through DNA on gold electrodes, Hojeong Ryu, et al. (Korea)[SEN-P5] Sensitivity-enhanced plasmonic detection of DNA hy-bridization, Seyoung Moon, et al. (Korea)[SEN-P6] Recognition of incremental changes in corneal ring (arcus-senilis), AMT Nasution, et al. (Indonesia)[SEN-P7] Investigation of the effect of target localization on the sensitivity enhancement of nanowire-mediated surface plasmon resonance biosensors, Seong Min Jang, et al. (Korea)

Optical Device and Systems for Bioscience[DEV-P1] New valve and splitter designs for microfluidic bio-chips containing proteins, Samuel I En Lin (Taiwan)[DEV-P2] Phase sensitive surface plasmon resonance biosensors using periodic nanowire structures, Kyungjae Ma, et al. (China)[DEV-P3] Measurement of vascular functions in mice brain using two-photon laser scanning microscopy, Jinho Kim, et al. (Korea)[DEV-P4] Frequency swept source at 1500 nm with a simple wavelength selection filter for biomedical imaging, Mansik Jeon, et al. (Korea)[DEV-P5] Optical microscanner with vertical comb electrodes for endoscopic OCT, Min-Ho Jun, et al. (Korea)[DEV-P6] Transmittance of UV light through TiO2 thin film de-posited by ion-assisted electron-beam evaporator, Seon Hoon Kim, et al. (Korea)[DEV-P7] Novel wavelength-swept FDML laser source using broadband Raman amplifier, Eun Joo Jung, et al. (Korea)[DEV-P8] Reflectivity tunable Sagnac loop device for the linear cavity of FBG laser system, Tae Ho Lee, et al. (Korea)[DEV-P9] Quantitative q-dot signal detection by using novel compact laser source, Hyung Man Lee, et al. (Korea)[DEV-P10] Recognition of incremental changes in corneal ring (arcus senilis) using hybrid N-feature neural network (HNFNN), Aulia M T Nasution, et al. (Indonesia)

Optical Coherence Tomography[OCT-P1] Endoscopic common-path OCT with flexible optical fiber probe, Jae Seok Park, et al. (Korea)[OCT-P2] Development of high-resolution high-speed spectral domain OCT for volumetric imaging, Cheol Song, et al. (Korea) [OCT-P3] Absolute wavelength calibration for spectral-domain OCT using fiber Bragg gratings, Tae Joong Eom, et al. (Korea)[OCT-P4] Spectral-domain OCT having compact handheld probe for dermatology applications, Eun Jung Min, et al. (Korea)[OCT-P5] Refocusing technique with phase shifting digital holography based on integrating bucket method, Gi Hyeon Min, et al. (Korea)[OCT-P6] Application of OCT for colorectal cancer diagnosis, Chih-Wei Lu, et al. (Korea)[OCT-P7] Dynamic analysis of internal mental sweating by OCT, Motomu Tanigawa, et al. (Japan) [OCT-P8] Trade-off among axial resolution, depth range, and sensitivity for high-resolution spectral domain OCT at 1.3 µm, Sang-Won Lee, et al. (Korea)[OCT-P9] Real-time label-free live cancerous cell imaging by using full-field OCT, Woo June Choi, et al. (Korea)[OCT-P10] Opto-mechanical design and simulation of 2D MEMS lens scanners for OCT based endoscopy, Hyeon-Cheol Park, et al. (Korea)

Optical Microscopy and Nanoscopy[MIC-P1] Development of reflection and fluorescence hybrid in-vivo confocal microscope, MyoungKi Ahn, et al. (Korea)[MIC-P2] Excitation energy migration processes in cyclic por-phyrin arrays probed by single molecule spectroscopy, Jaesung Yang, et al. (Korea)[MIC-P3] Shear-force imaging of human hair cuticle damaged by an acid solution, Kyoung-Duck Park, et al. (Korea)[MIC-P4] Femtosecond transient absorption microscope system for nano-imaging, Sang-Youp Yim, et al. (Korea)[MIC-P5] High-resolution temporal and spatial PL measure-ment of semiconductor quantum structures at room tempera-ture, Hong-Gyu Ahn, et al. (Korea)[MIC-P6] Fluorescence dynamics of directly meso−meso linked porphyrin rings probed by single molecule spectroscopy, Hyejin Yoo, et al. (Korea) [MIC-P7] Measurement of THG signal in fused silica using third-harmonic generation microscopy, Lee EungJang, et al. (Korea)[MIC-P8] Nanostructure-based surface plasmon resonance im-aging, Dong Jun Kim, et al. (Korea)[MIC-P9] Implementation of resolution-enhanced integral im-aging microscope, Yong-Tae Lim, et al. (Korea)[MIC-P10] Two-photon photoluminescence imaging of single gold nanorod using polarization selective supercontinuum light, Wei Tao, et al. (Australia)

Optical Spectroscopy for Biomedical Research[SPE-P1] Differential sensitivity of near-infrared spectroscopy to oxygenation changes in different vascular compartments, T. Eriguchi, et al. (Japan)[SPE-P2] Changes in cerebral blood oxygenation after suba-rachnoid hemorrhage evaluated by near-infrared time-resolved spectroscopy, T. Hoshino, et al. (Japan)[SPE-P3] Influence of the depth of brain activation on wavelength dependence of optical path length, N. Sakashita, et al. (Japan)[SPE-P4] Monitoring of microvascular reaction in skin tissue by use of multispectral reflectance images, I. Nishidate, et al. (Japan)[SPE-P5] Investigating the relationships between the backscat-tering spectrum and nuclear size, Guo-Shan, et al. (Taiwan)[SPE-P6] The role of NaCl in triple helix DNA formation, Ja Eun Lee, et al. (Korea)[SPE-P7] Highly sensitive SERS detection of duplex DNAs in a microdroplet channel, C. Lim, B. Han, et al. (U.K.)[SPE-P8] Highly sensitive immunoassay of cancer markers using SERS of hollow gold nanospheres, H. Chon, et al. (Korea)[SPE-P9] Biological application of pre-patterned glass microarray chip us-ing hydrophobic/hydrophilic surface modification, M. Lee, et al. (Korea)

www.apbp2009.org www.apbp2009.org

Recognition of incremental changes in corneal ring (arcus-senilis)

using the hybrid N-feature neural network (HNFNN)

AMT Nasution, WW Kusuma

Dept. of Engineering Physics, Institut Teknologi Sepuluh Nopember

Kampus ITS Sukolilo, Surabaya 60111, Indonesia

Correspondence: anasution@ep.its.ac.id

Abstract: The Arcus Senilis formation have a strong correlation to blood’s lipid level, and has potential diagnostics values.

In this paper efforts to develop inexpensive system, which capable to accurately recognize incremental changes of

the ring, is reported. The hybrid N-feature neural network (HNFNN) is utilized for the recognition process.

1. Introduction Arcus senilis (corneal ring), a circular ring

normally present to be deposited around the corneal periphery of older people. This ring is also even found in younger ones with higher blood’s lipid level. In medical terms it is known as arcus senilis, or corneal arcus (ring) in general. Many researchers have supposed that the ring have diagnostic values, i.e. as predictor for coronary artery diseases [1,2] as well as lipid stratification [3].

Previous studies mentioned that the formation of

arcus senilis is due to infiltration of cholestrol and

cholesterol esters [4]. This lipid deposition is

believed to have possible relationship with the

systemic disease. Reviews on this matter can be

found in [5,6].

The ring formation process is governed by

interrelation of many variables [1,7], among others

are age, sex, race, and pathogenetic factor, and the

relationship among these factors are still poorly

understood and requires further studies [7].

An accurate diagnostic tool would be required

for these studies. In this paper it is reported an effort

to develop a diagnostic system which is required to

have a good accuracy, simple, as well as

inexpensive to realize. The system works based on

pictures of interior segment of eye, which taken by

using pocket digital camera. A hybrid N-features

neural network (HNFNN) algorithm is

implemented for this purpose. This algorithm has

already shown a good performance for facial

recognition, as reported by Haddadnia [8] with

high recognition rates of 99,5 % - 100 % . The goal

of the developed system is to be capable to

recognize the incremental changes in arcus-senilis.

2. Materials and Method In this Section several important fundamental principles underlying the works will be briefly mentioned. A. Arcus Senilis (Corneal Arcus) and its charac-

teristics

Arcus senilis, also known as corneal arcus lipoides, is a lipid rich annular ring that is formed on the corneal periphery. It is normally found in older people, but also common to younger ones with hyperlipidemia . Visually it has white color – tend to be yellowish opacity. The typical width of the ring is in the order of 1 – 1.5 mm.

The formation of the ring is frequently associated to the abnormality in the serum lipid level. It is also associated with alcohol intake, diabetes mellitus, smoking, blood pressure, body mass index (BMI), obesity, and age [1 and references therein]. Lipid deposition are mostly comes through perfusion from the limbic vasculatures, which are located close to the peripheral cornea. Initially the deposition will forms arcs at the inferior, and the spreads to posterior poles of the cornea. These arcs will finally join to form a complete annulus with a width ca 1 mm. Typical form of the ring can be seen in Figure-1.

Figure-1: Arcus senilis (indicated by red arrow)

The proposed system is developed using the following steps, as shown in Figure-2.

Figure-2: Flowchart of the recognition process

B. Image Acquisition and Preprocessing

Pre-processing stage is done prior the recognition using HNFNN algorithm. This step is utilized to separate the corneal ring image from its backgrounds. A canny segmentation method combined with Circular Hough Transform (CHT) is implemented for achieving such purpose. The segmented corneal ring image is grey-scaled and saved in pgm file format for further processing.

HNFNN Input

images Pre

processing Features

Extraction

Classification

using RBF

neural network

Voting:

Majority Rule

C. Feature Extraction

In the HNFNN method, instead of using single

feature extractor, several feature extractors are used

in hybrid manner. The more extractors used the

more image characteristics can be gained from an

image, which ultimately yield better recognition.

The Principal Component Analysis (PCA), Exact

Lagendre Moment (ELM), Geometric Moment

(GM), Pseudo Zernike Moment (PZM), Discrete

Cosine Transform (DCT) are the feature extractors

that usually used in HNFNN. In the work done, only

the first three of the extractors will be utilized.

D. Classification and Voting

The identification is accomplished by using

Radial Basis Function (RBF) neural network with

inputs of results from the three feature-extractors

used. Decision is then taken by using majority vote,

based on the highest scores given by the classifiers.

All of the algorithms mentioned above is

implemented under Matlab 7.0.4

3. Results and Analysis

The developed recognition system is prior tested

using dummy corneal ring data with different ring

width, i.e. 0.1 mm – 2 mm. It is then tested using

the real corneal ring images, which taken using

pocket digital camera.

The dummy corenal ring is actually taken from

real picture of eye, with substitute of dummy

cornea together with the variably changes ring. It

can be shown in Figure-3.

Figure-3: Picture of dummy corneal ring

Ten pictures are used for training and the other ten

are used for testing purposes. The sample results of

HNFNN system’s recognition rate under test using

the dummy corneal rings (1 mm and 2 mm) is

shown Table-1.

Table 1. Sample results of HNFNN recognition

rate for dummy corneal picture different ring width

Width Order-n

*)

1 2 3 4 5

1 mm 90 % 100 % 100 % 100 % 100 %

2 mm 100 % 100 % 100 % 100 % 100 % *)order = order of the GM and ELM extractor function

Similar procedures is then implemented for real

corenal ring pictures. Using 15 pictures for traing

and another 15 for testing, results can be shown in

Table-2.

Table 2. Sample results of HNFNN system

recognition rate for real corneal pictures with arcus

senilis

Status Order-n

1 2 3 4 5

Arcus

senilis 53,3 % 86,7 % 93,3 % 93,3 % 93,3 %

Ongoing works is now devoted to improve the

accuracy of recogntition by improving the

extraction and classification algorithms, as well as

to set-up the calibration procedure of the recording

system.

4. Conclusion

The works concludes that:

• it is possible to recognize the incremental

changes in corneal arcus width, as

demonstrated by dummy corneal ring images.

• Order of extractor functions (GM and ELM)

will influence the obtained accuracy

• Setting up calibration procedure of image

recording system as well as improvements in

extraction and classification algorithms are

believed can improve the detection accuracy.

5. References [1] A Fernandéz, A Sorokin, P D Thompson,

“Corneal arcus as coronary artery disease risk factor”, Atherosclerosis, 193, 2007, 235–240.

[2] A Fernandéz et al., “Relation of Corneal Arcus to Cardiovascular Disease (from the Framingham Heart Study Data Set)”, The American Journal of Cardiology, Vol. 103 (1), January 2009, p. 64-66.

[3] Chua, B.E., Mitchel, P., et.al., “Corneal arcus and Hyperlipidemia: Findings from an Older Population”, American Journal of Opthalmology, vol. 137 (2), Feb. 2004, 363 – 365.

[4] Feldman, G.I., “The Lipids in Pathology od the Eye”, The Journal of American Oil Chemists’ Society, vol 44, Nov 1967, 615-622.

[5] Barchiesi, B.J., Eckel, R.H., Ellis, P.P., “The cornea and disorders of lipid metabolism”. Surv. Ophthalmol. 36, 1991, 1–22.

[6] Winder, A.F., 1994. “Disorders of Lipid and lipoprotein metabolism”. In: Garner, A., Klintworth, G.K. (Eds.), Pathobiology of Ocular Disease: A Dynamic Approach, 2

nd

Edition. Marcel Dekker, New York, pp. 1099–1121.

[7] Crispin, S., “Ocular lipid deposition and hyperlipoproteinaemia”, Progress in Retinal and Eye Research, vol 21, 2002, 169 – 224.

[8] Haddadnia J., Ahmadi, M., “N-feature neural network human face recognition”, Image and Vision Computing, vol 22, 2004, 1071–1082.

6. Acknowledgement The authors would express the highest gratitude for Dr. Tini Endang (Clinical Pathology Dept.) and staffs of the General Hospital Saiful Anwar – Malang (East Java) for the invaluable supports and discussions during the research.