Retinal vasculature enhancement using independent component analysis

doi:10.4236/jbise.2009.27079

Paper Menu >>

Journal Menu >>

J. Biomedical Science and Engineering, 2009, 2, 543-549

doi: 10.4236/jbise.2009.27079 Published Online November 2009 (http://www.SciRP.org/journal/jbise/

JBiSE

Published Online November 2009 in SciRes. http://www.scirp.org/journal/jbise

Retinal vasculature enhancement using independent

component analysis

Ahmad Fadzil M. Hani1*, Hanung Adi Nugroho1,2**

1Department of Electrical and Electronic Engineering, Universiti Teknologi PETRONAS, Bandar Seri Iskandar, Tronoh, Perak Darul

Ridzuan, Malaysia; 2Department of Electrical Engineering, Univeristas Gadjah Mada, Jl. Grafika 2, Kampus UGM, Jogjakarta, Indo-

nesia.

Email: *[email protected]; **[email protected]

Received 22 June 2009; revised 16 July 2009; accepted 24 July 2009.

ABSTRACT

Retinal vasculature is a network of vessels in the

retinal layer. In ophthalmology, information of reti-

nal vasculature in analyzing fundus images is impor-

tant for early detection of diseases related to the ret-

ina, e.g. diabetic retinopathy. However, in fundus

images the contrast between retinal vasculature and

the background is very low. As a result, analyzing or

visualizing tiny retinal vasculature is difficult. There-

fore, enhancement of retinal vasculature in digital

fundus image is important to provide better visuali-

zation of retinal blood vessels as well as to increase

accuracy of retinal vasculature segmentation. Fluo-

rescein angiogram overcomes this imaging problem

but it is an invasive procedure that leads to other

physiological problems. In this research work, the

low contrast problem of retinal fundus images ob-

tained from fundus camera is addressed. We develop

a fundus image model based on probability distribu-

tion function of melanin, haemoglobin and macular

pigment to represent melanin, retinal vasculature

and macular region, respectively. We determine reti-

nal pigments makeup, namely macular pigment,

melanin and haemoglobin using independent com-

ponent analysis. Independent component image due

to haemoglobin obtained is used since it exhibits

higher contrast retinal vasculature. Contrast of reti-

nal vasculature from independent component image

due to haemoglobin is compared to those from other

enhancement methods. Results show that this ap-

proach outperforms other non-invasive enhancement

methods, such as contrast stretching, histogram eq-

ualization and CLAHE and can be beneficial for

retinal vasculature segmentation. Contrast enhance-

ment factor up to 2.62 for a digital retinal fundus

image model is achieved. This improvement in con-

trast reduces the need of applying contrasting agent

on patients.

Keywords: Contrast Enhancement; Independent Comp-

onent Analysis; Medical Image Processing; Retinal Fundus

Image

1. INTRODUCTION

Analyzing retinal fundus image is important for early

detection of several diseases related to the retina, e.g.

diabetic retinopathy. In diabetic retinopathy, retinal capi-

llary occlusion occurs and accordingly causes enlarge-

ment of foveal avascular zone. Foveal avascular zone is

the fovea where there is no blood vessels and located in

the very centre of macula. Information of retinal vascu-

lature is important to accurately determine the foveal av-

ascular zone. However, digital color fundus images ob-

tained from fundus camera suffer from several problems

as can be seen from Figure 1. Figure 1(a) illustrates the

problems of very low contrast and non-uniform illumi-

nation which can be seen at the area towards the edge of

the image. Figure 1(b) shows the occurrence of noise

which consists of impulse and Gaussian noises. Detec-

tion of the foveal avascular zone is even difficult due to

very low image contrast of retinal vasculature against the

background in the macular region.

A number of enhancement methods focused in the im-

age spatial domain [2,3,4,5]. Histogram equalization wi-

th its modification is commonly used to enhance the im-

age contrast [6]. However, histogram equalization tends

to over-enhance the image and results in noisy appear-

ance of the output image. One of the adaptive methods

called contrast limited adaptive histogram equalization

(CLAHE) worked well on the enhancement of retinal

vasculature [7]. Iznita found that the contrast improve-

ment using contrast limited adaptive histogram equaliza-

tion on an image model ranges between 1.7 and 3 [8].

However, contrast limited adaptive histogram equaliza-

tion creates artefacts in the enhanced image and the se-

lection of contrast gain limit is image-dependent.

Other related works used the information of color

taken from digital color images [9,10]. Colors observed

A. F. M. H et al. / J. Biomedical Science and Engineering 2 (2009) 543-549

544

Figure 1. Digital fundus images obtained from fundus camera [1].

in the retinal image correspond to the architecture of re-

tinal layer and the optical properties of the pigments

[11,12]. Styles et al. developed a model using the con-

centrations of the five main absorbers found in the fun-

dus layers, namely retinal haemoglobin, choroidal hae-

moglobin, choroidal melanin, retinal pigment epithet-

lium melanin and macular pigment [13]. This approach

focuses more towards reconstructing rather than improv-

ing the contrast. Tsumura et al. showed that spatial dis-

tributions of melanin and haemoglobin from a skin color

image can be separated using independent component

analysis [9,14]. Nugroho et al. successfully applied a

technique based on principal component analysis and

independent component analysis to convert the RGB

skin image into a skin image that represents skin areas

due to melanin and haemoglobin only [10]. The above

efforts focus on using independent component analysis

to transform digital color image (RGB) into independent

components that correspond to the biological makeup of

the skin.

The objective of this work is to address the low con-

trast problem of retinal fundus images obtained from

fundus camera when no contrasting agent is injected. A

novel approach is presented to enhance the contrast of

retinal vasculature by determining the retinal pigments,

namely macular pigment, haemoglobin and melanin

from fundus images. Distribution of haemoglobin is ex-

tracted from a fundus image to reveal retinal vasculature,

which is a network of vessels in the retinal layer. Con-

trast of retinal vasculature obtained using this approach

is compared to those from other enhancement methods,

such as contrast stretching, histogram equalization and

contrast limited adaptive histogram equalization to test

the performance of this approach.

2. APPROACH

The approach taken in this research is as follows. First, a

model of ocular fundus based on the light interaction is

developed to describe the reflectance of the fundus. Se-

cond, a model of spectral absorbance of the retinal image

is developed to show the components composing the ob-

served colours in a digital fundus image. Third, inde-

pendent component analysis based on the spectral ab-

sorbance of the model is applied to determine retinal

pigments from fundus images. Finally, two fundus image

models are developed to test performance of the pro-

posed algorithm.

2.1. Ocular Fundus Model

Ocular fundus represents the structure of the back of the

eyes that consists of multiple layers of tissue [13]. The

ocular fundus image obtained from a fundus camera sh-

ows different intensity of reflectance. The reflectance

depends on the wavelength, the structure of fundus’ lay-

ers, the optical properties and quantities of retinal pig-

ments in the ocular fundus. The incident light from a

fundus camera can be reflected, absorbed, scattered or

transmitted by the retinal tissues.

Generally, the structure of the eye can be classified

into two main groups, namely ocular media and ocular

fundus [15]. Ocular media consists of cornea, lens and

vitreous. It is located between the ocular fundus and the

observer. The ocular fundus consists of the retina, the

retinal pigment epithelium, the choroid and the sclera.

The reflectance of the fundus can be described in the

terms of these layers [16]. Figure 2 depicts a model of

ocular fundus showing possible pathways of the re-

flected light.

2.2. Ocular Fundus Spectral Absorbance Model

The spectral absorbance image provides useful informa-

tion to identify the absorbance components [14]. In this

work, we focus on the distribution of retinal pigments,

namely haemoglobin, melanin, and macular pigment,

rather than on the fundus layers, to model spectral ab-

sorbance of the ocular fundus [17].

Basis of linear combination of the absorption coeffi-

cients of melanin, haemoglobin and macular pigment is

modelled from three absorbances μa(λ1), μa(λ2) and

Figure 2. A model of ocular fundus showing pathways of re-

flected light.

A. F. M. H et al. / J. Biomedical Science and Engineering 2 (2009) 543-549

545

μa(λ3) at three wavelengths λ1, λ2 and λ3. These wave-

lengths λ1, λ2 and λ3 represent the red (R), green (G) and

blue (B) color channels. Fundus spectral absorbance

image shows spectral characteristics of the absorbance

components in the ocular fundus. Two conditions are

assumed when analyzing fundus spectral absorbances.

First, the color observed in the fundus image is due to

the distributions of melanin, haemoglobin and macular

pigment. Second, the quantities of these components are

spatially independent of each other. The spectral absorb-

ance in the fundus image represents the linear combina-

tion of the absorption coefficients of melanin, haemo-

globin and macular pigment.

Let sx,y and vx,y designate a three-dimensional (3-D)

quantity vector and composite color vector on an image

coordinate (x, y) of a digital color image. A mixing ma-

trix A with a1, a2 and a3 represents pure color vectors of

the three components (haemoglobin, melanin, macular

pigment) per unit quantity. It is assumed that a linear

combination of mutually independent pure color vectors

with the quantities of s1x,y, s2x,y and s3x,y result in the

composite color vectors of v1x,y, v2x, y and v3x,y on the im-

age coordinate (x, y). The following equation illustrates

the transformation matrix, where T denotes the trans-

pose.

vx,y = A sx,y (1)

sx,y = [s1x,y, s2x,y, s3x,y]T (2)

The pixel value of each channel corresponds to each

element of the color vector. Figure 3 depicts the spectral

absorbances of the ocular fundus which consist of pure

spectral vectors of melanin, haemoglobin and macular

pigment.

2.3. ICA of Fundus Spectral Absorbance Image

Independent component analysis (ICA) is a technique to

determine the original signals from mixtures of several

independent sources [18,19]. The independent compo-

Figure 3. Model of spectral absorbance of the ocular fundus.

Macular pigment

Hemoglobins

Melanin

Red channel

Green channel

Blue channel

Macular pigment

Hemoglobins

Melanin

mixture

(A) separatio n

(W)

original sourcesestimated source

Ocular fundus im ag e

(observed ima g e)

Macular pigment

Hemoglobins

Melanin

Red channel

Green channel

Blue channel

Macular pigment

Hemoglobins

Melanin

mixture

(A) separatio n

(W)

original sourcesestimated source

Ocular fundus im ag e

(observed ima g e)

Figure 4. The problem of ICA in ocular fundus image.

nent analysis is modelled as

v = As (3)

with mixing matrix A and random vector v, denoting the

mixtures v1, v2, …, vn. Similarly, s random vector denotes

the elements of s1, s2, …, sn. This model shows how the

observed data vn is generated by a process of mixing the

components si. The independent components cannot be

directly observed and neither can the mixing matrix.

Only the random vector v is being observed. Mixing

matrix A and random vector s are estimated using v.

Subsequently, separating matrix W is used to find the

independent component simply by

ŝ = Wv, (4)

with ŝ is defined as estimated sources. The objective of

independent component analysis is then to get ŝ as close

as possible to s, which is determined as original sources,

by determining the optimum separating matrix W. Mu-

tually independent components are determined as ele-

ments of vector s from the mixture of vectors in the im-

age. A diagram is shown in Figure 4 to illustrate the idea

of using independent component analysis in separating

the spatial distributions of melanin, haemoglobin, and

macular pigment in the ocular fundus. Three color cha-

nnels, namely red, green and blue channels, represent

random vector v and are used to determine these inde-

pendent components [17]. By applying the independent

component analysis to the composite colour vectors in

the image, the relative quantity and pure colour vectors

of each independent component are determined with no

prior information on the quantity as well as colour vector.

In this case, the independent components represent the

retinal pigments, i.e. melanin, haemoglobin and macular

pigment. The quantities of the retinal pigments are pre-

sumed to be mutually independent for the image coordi-

nate. The separating matrix W is defined to separate

vector ŝx,y using the following equations.

ŝx,y = W vx,y (5)

ŝx,y = [ŝ1x,y, ŝ2x,y, ŝ3x,y]T (6)

The extracted independent components ŝ1x,y, ŝ2x,y and

ŝ3x,y may be similar to s1x,y, s2x,y and s3x,y, respectively.

The composite colour vector vx,y is determined based on

the logarithm transformation of the pixel intensities in

A. F. M. H et al. / J. Biomedical Science and Engineering 2 (2009) 543-549

546

the color channels of red, green and blue. Logarithmic

transformation is used to transfer reflectance spectra into

spectral absorbance since spectral absorbance image

provides useful information to identify the absorbance

components [14].

[μa(λ1),μa(λ2),μa(λ3)]=[-log(rx,y),-log(gx,y),-log(bx,y)] (7)

here, the values of rx,y, gx,y and bx,y correspond to

pixel intensity in the color channels of red, green and

blue respectively. The composite color vector is de-

noted as

vx,y = [μa(λ1), μa(λ2), μa(λ3)]T (8)

According to the model of spectral absorbance in the

ocular fundus from Figure 3, the color density vector of

the fundus can be stated as

vx,y = A sx,y + a4 (9)

where A = [a1, a2, a3] and sx,y = [s1x,y, s2x,y, s3x,y]T. Ele-

ments a1, a2 and a3 of the mixing matrix A represents

pure color vectors of the three components (haemoglo-

bin, melanin, macular pigment) per unit quantity. It is

assumed that a linear combination of mutually independ-

ent pure color vectors with the quantities of s1x,y, s2x, y

and s3x,y results in the composite color vectors of v1x,y,

v2x,y and v3x,y on the image coordinate (x, y). Additionally,

a4 is similar to noise in the ICA model. In this case, the

model is assumed to be noise-free, therefore a4 can be

neglected.

Several methods, such as fast fixed-point algorithm

(FastICA) [20], joint approximate diagonalization of

eigen-matrices (JADE) [21] and information- maximi-

zation (infomax) [22] have been proposed to solve the

problem of independent component analysis. In ICA, the

only assumption needed are: 1) the sources are statisti-

cally independent, 2) the probability densities of the

sources are non-Gaussian, 3) the mixing of the sources

into the observations is linear, and 4) the number of ob-

servations is larger than or equal to he number of sources

[19]. The FastICA algorithm with symmetrical orthogo-

nalization is used to get the estimated independent com-

ponents because of its good accuracy and high computa-

tional speed for high dimensional data [20].

2.4. Fundus Image Model

A model of fundus image is developed to test the per-

formance of independent component analysis in sepa-

rating the distribution of macular pigment, hemoglobin

and melanin. Mixture of three mutually independent

components, i.e. macular pigment, hemoglobin and

melanin is used to model a fundus image. As shown in

Table 1, the statistical intensity description of macular

pigment, haemoglobin and melanin in red, green and

blue channels are taken from the 44 test images from

FINDeRS [23]. A smaller region containing macular area

is sampled to get the probability density function of the

Table 1. Statistical intensity description of macular pigment,

haemoglobin and melanin in red (R), green (G) and blue (B)

channels.

Macular

pigment Haemoglobin Melanin

Mean R 97.14868 120.0417 156.5642

Standard

deviation R 28.39299 27.16076 23.98799

Skewness R 0.714194 0.74299 0.07827

Kurtosis R 0.522469 0.259348 0.215941

Minimum R 46.26375 73.32877 102.7158

Maximum R 174.1571 195.1525 219.6025

Mean G 48.29849 62.24773 96.08492

Standard

deviation G 13.82747 17.29835 19.95184

Skewness G 0.719616 0.376683 0.620011

Kurtosis G 1.577469 0.236853 1.59036

Minimum G 21.67789 32.06349 57.14166

Maximum G 95.15525 114.2881 164.2989

Mean B 7.971792 17.48195 35.1042

Standard

deviation B 5.483867 11.73761 18.85215

Skewness B 1.989686 1.596864 1.31945

Kurtosis B 6.239178 3.978853 2.216477

Minimum B 1.992481 2.412698 12.66622

Maximum B 31.36347 63.40678 100.4873

retinal pigments. In the macular region, retinal capillar-

ies usually show a very low contrast between retinal

blood vessels and the background. A clustering method

using k-means based on the intensity of red, green, and

blue channels of the macular pigment, haemoglobin and

melanin is performed to classify the samples due to large

value of standard deviation and intensity range of the

sample. Based on the experiment, two numbers of clus-

ters are found to be optimal to classify the samples.

In Figure 5, two fundus image models to represent

fair and dark fundus images with mixture of specified

sample intensity distribution of macular pigment, hae-

moglobin and melanin in red, green and blue channels

are shown. Using these models as the input, the inde-

pendent component analysis should be able to separate

these components into three outputs, namely macular

pigment, haemoglobin and melanin.

3. RESULTS AND DISCUSSIONS

A fundus image model is firstly tested using independ-

ent component analysis to see performance of the algo-

rithm. The inputs to the FastICA are three separate

channels (i.e. red, green and blue channels) of a color

fundus image model. As can be seen from Figure 6, the

proposed algorithm successfully separates the compo-

A. F. M. H et al. / J. Biomedical Science and Engineering 2 (2009) 543-549

547

(a). Fair image model

(b). Dark image model

Figure 5. Fundus image model.

a. Macular regionb. Melanin

c. Retinal vasculature

Figure 6. Independent component analysis of dark fundus

image model.

nents into three, namely macular pigment, haemoglobin

and melanin. These three independent components rep-

resent macular region, retinal vasculature and melanin,

respectively. In Figure 6(a), the brighter area in lower

part of the fundus image model is related to the macular

region. In Figure 6(b), the melanin is illustrated as the

brighter area in upper part of the fundus image model.

These two components can be clearly distinguished from

the other component, which is related to retinal vascula-

ture, since the retinal vasculature is almost invisible in

the appearance of these two components (i.e. macular re-

gion and melanin). Furthermore, the retinal vasculature

is clearly visualized in Figure 6(c). As a result, indepen-

dent component image due to haemoglobin obtained ex-

hibits higher contrast retinal vasculature compared to

that of the original image.

In this work, 44 retinal fundus images containing

macular region are taken from FINDeRS database to

model a retinal fundus image. The fundus image model

undergoes several enhancement methods, such as con-

trast stretching, histogram equalization, contrast limited

adaptive histogram equalization (CLAHE) to measure

contrast improvement factor of these methods and com-

pare to the proposed algorithm. A smaller region con-

taining the macular area is taken to see the enhancement

of retinal capillaries, which usually show a very low

contrast between retinal blood vessels and the back-

ground. Figure 7 shows green band of dark fundus im-

age model undergoing several enhancement methods, i.e.

contrast stretching, histogram equalization and CLAHE.

Qualitatively, haemoglobin related ICA shows better

enhancement because no artefacts is produced in the

process. Nevertheless, the other three enhancement me-

thods tend to increase the noise presence in the image as

well as to produce artefacts.

From the fundus image model, the green band image

shows the average contrast intensity of 24.60 and 16.80

for fair and dark image model, respectively. Using these

values as a reference, the proposed algorithm using ICA

a. Contrast stretchingb. Histogram equalization

c. Contrast limited AHEd. ICA (haemoglobin)

Figure 7. Dark fundus image model undergoes several en-

hancement methods.

with contrast enhancement factor of 1.47 and 2.62 for

A. F. M. H et al. / J. Biomedical Science and Engineering 2 (2009) 543-549

548

fair and dark fundus image models shows a better im-

provement for both fair and dark fundus image models

than that of the other three enhancement methods. Fur-

thermore, as is shown in Figure 8, CLAHE with en-

hancement factor of 1.37 and 1.98 for fair and dark fun-

dus image model, respectively, is still better than that of

contrast stretching and histogram equalization. However,

compared to that of CLAHE, the proposed algorithm

produces no artefacts in the process.

Here an example of retinal image showing macular

region is taken to see enhancement of retinal vasculature

using the proposed algorithm. In a preliminary work

using the above algorithm, it is found that non-uniform

illumination in fundus images resulted in false detection

of the retinal pigments [17]. This is because the algo-

rithm responds to the spectral reflectance or absorbance

of the retinal pigments in the image. Therefore, homo-

morphic filtering is performed prior to independent com-

ponent analysis to reduce the problem of non-uniform

illumination. Homomorphic filtering is used to reduce

illumination which varies slowly in space and at the

same time [24].

Figure 9 shows an original color fundus image un-

dergoing homomorphic filtering and its independent

components estimated by the FastICA algorithm. The

components represent the distribution of the pigments,

namely macular pigment, haemoglobin and melanin. The

brighter area in the centre of the first independent com-

ponent (Figure 9(b)) represents the distribution of

macular pigment. The second independent component

(Figure 9(c)) shows the distribution of haemoglobin. It

is indicated by the enhancement of retinal vasculature.

The third independent component (Figure 9(d)) shows

brighter area related to the distribution of melanin. This

result is consistent with the location of melanin, which is

fairly distributed in the retinal pigment epithelium and

the choroid. Based on the assumption that the image is

noise-free, independent component analysis is able to

determine the retinal pigments. Moreover, as shown in

Figure 10, a green band image undergoing CLAHE is

0.5

1.5

2.5

Contrast

Stretching

Histogr am

Equalization

CLAHE ICA

Fair

Dark

Figure 8. Contrast enhancement factor of retinal vasculature in

fundus image model.

a. Fundus image after

homomorphic filtering

b. First component

c. Second componentd. Third component

a. Fundus image after

homomorphic filtering

b. First component

c. Second componentd. Third component

Figure 9. Independent component analysis of a retinal image

containing macular region.

Figure 10. Comparison of contrast enhancement of retinal

vasculature between CLAHE and ICA.

compared to the haemoglobin-related component image

after the intensity is being inverted to demonstrate that

contrast enhancement is also achieved. In this work,

CLAHE is also performed on the same images undergo-

ing the proposed algorithm to compare the contrast im-

provement between these two methods. Having meas-

ured the contrast improvement factor on the fundus im-

age model, the proposed algorithm consistently shows

better visualization and enhancement compared to that

of the CLAHE, which is commonly used as pre-proce-

ssing for segmentation of retinal vasculature in fundus

images. This improvement can be beneficial to improve

the accuracy of retinal vasculature segmentation and re-

duce the need for injecting contrasting agent to the pa-

tients.

4. CONCLUSIONS

Analyzing retinal fundus images is usually difficult as

they are of very low contrast. Low contrast between

blood vessels and the background makes it difficult to

A. F. M. H et al. / J. Biomedical Science and Engineering 2 (2009) 543-549

549

accurately determine retinal vasculature. Retinal vascu-

lature can be used to determine existence of pathology,

macular area and foveal avascular zone. Typical contrast

enhancement methods usually create artefacts or intro-

duce noise. Even though fluorescein angiography pro-

duces better contrast enhancement, it is not preferable

due to its invasive nature of injecting contrasting agent.

JBiSE

In this work, the developed method based on the spec-

tral absorbance model and independent component ana-

lysis enables us to determine the retinal pigments, name-

ly haemoglobin, melanin and macular pigment. A fundus

image model has been developed to test the performance

of the proposed algorithm. As a result, retinal vascula-

ture, macular pigment and melanin distribution can be

determined from digital fundus image. Results show that

this approach outperforms other non-invasive enhance-

ment methods, such as contrast stretching, histogram

equalization and CLAHE and can be beneficial for ves-

sel segmentation. The algorithm produces no artefacts in

the process. Using the haemoglobin component, the con-

trast between retinal blood vessels and the background

can be enhanced with contrast enhancement factor up to

2.62 for a model of fundus image. This improvement in

contrast reduces the need of applying contrasting agent

on patients.

REFERENCES

[1] Staal, J., Abramoff, M. D., Niemeijer, M., Viergever, M.

A., and Ginneken van B., (2004) Ridge-based vessel

segmentation in color images of the retina, Medical Im-

aging, IEEE Transactions, 23, 501–509.

[2] Niemeijer, M., Staal, J., Ginneken v. B., Loog M., and

Abramoff M. D., (2004) Comparative study of retinal

vessel segmentation methods on a new publicly available

database, Medical Imaging: Image Processing, Milan,

648–656.

[3] Shimahara, T., Okatani, T., and Deguchi, K., (2004) Con-

trast enhancement of fundus images using regional his-

tograms for medical diagnosis, SICE Annual Conference,

1, 650–653.

[4] Sinthanayothin, C., Kongbunkiat, V., Phoojaruenchana-

chai, S., and Singalavanija, A., (2003) Automated screen-

ing system for diabetic retinopathy, Image and Signal

Processing and Analysis, ISPA, Proceedings of the 3rd

International Symposium on, 915.

[5] Wu, D., Zhang, M., Liu, J. C., and Bauman, W., (2006)

On the adaptive detection of blood vessels in retinal im-

ages, Biomedical Engineering, IEEE Transactions on, 53,

341–343.

[6] Yang, S., Oh, J. H., and Park, Y., (2003) Contrast en-

hancement using histogram equalization with bin under-

flow and bin overflow, in Image Processing, ICIP, Pro-

ceedings International Conference on, 1, I–881–4.

[7] Fadzil, M. H. A., Izhar, L. I., Venkatachalam, P. A., and

Karunakar, T. V. N., (2007) Extraction and reconstruction

of retinal vasculature, Journal of Medical Engineering

Technology, 31, 435–442.

[8] Iznita, L., (2006) Analysis of retinal vasculature and

foveal avascular zone for grading of diabetic retinopathy,

in M.Sc, Thesis Electrical and Electronics Engineering

Programme Bandar Seri Iskandar, Malaysia, Universiti

Teknologi PETRONAS.

[9] Tsumura, N., Haneishi, H., and Miyake, Y., (1999) Inde-

pendent-component analysis of skin color image, J. Opt.

Soc. Am. A, 16, 2169.

[10] Nugroho, H., Fadzil, M. H. A., Yap, V. V., Norashikin, S.,

and Suraiya, H. H., (2007) Determination of Skin Repi-

gmentation Progression, Engineering in Medicine and

Biology Society, EMBS Annual International Conference

of the IEEE, 3442.

[11] Preece, S. J. and Claridge, E., (2002) Physics Based

Medical Image Understanding of the Colouration of the

Ocular Fundus with Application to Detection of Diabetic

Retinopathy, Department of Computer Science, Univer-

sity of Birmingham, Edgbaston, Birmingham Technical

Report CSR–04–08.

[12] Orihuela-Espina, F., Claridge, E., and Preece, S. J., (2003)

Histological parametric maps of the human ocular fundus:

preliminary results, Medical Image Understanding and

Analysis, 133–136.

[13] Styles, I. B., Calcagni, A., Claridge, E., Orihuela-Espina,

F., and Gibson, J. M., (2006) Quantitative analysis of

multi-spectral fundus images, Medical Image Analysis:

Special Issue on Functional Imaging and Modelling of

the Heart (FIMH 2005), 10, 578–597.

[14] Tsumura, N., Haneishi, H., and Miyake, Y., (2000) Inde-

pendent Component Analysis of Spectral Absorbance

Image in Human Skin, Optical Review, 7, 479–482.

[15] Knighton, R. W., (1995) Quantitative reflectometry of the

ocular fundus, Engineering in Medicine and Biology

Magazine, IEEE, 14, 43–51.

[16] Berendschot, T. T. J. M., DeLint, P. J., and Norren, D.,

(2003) Fundus reflectance-historical and present ideas,

Progress in Retinal and Eye Research, 22, 171–200.

[17] Ahmad Fadzil, M. H., Nugroho, H. A., Venkatachalam, P.

A., Nugroho, H., and Izhar, L. I., (2008) Determination

of Retinal Pigments from Fundus Images using Inde-

pendent Component Analysis, in Biomed 4th Kuala

Lumpur International Conference on Biomedical Engi-

neering, Kuala Lumpur, 555–558.

[18] Comon, P., (1994) Independent component analysis, A

new concept?, Signal Processing, 36, 287.

[19] Hyvarinen, A. and Oja, E., (2000) Independent compo-

nent analysis: algorithms and applications, Neural Net-

works, 13, 411.

[20] Hyvarinen, E., (1999) Fast and robust fixed-point algo-

rithms for independent component analysis, Neural Net-

works, IEEE Transactions on, 10, 626–634.

[21] Cardoso, J. F. and Souloumiac, A., (1999) Blind beam-

forming for non Gaussian signals, IEE Proceedings-F,

140, 362–370.

[22] Bell, A. J. and Sejnowski, T. J., (1995) An informa-

tion-maximisation approach to blind separation and blind

deconvolution, Neural Computation, 7, 1129–1159.

[23] Ahmad Fadzil, M. H., Nugroho, H. A., Nugroho, H., and

Izhar, L. I., (2009) Fundus Image Database for Non In-

vasive Diabetic Retinopathy Monitoring and Grading

System (FINDeRS), Universiti Teknologi PETRONAS.

[24] Adelmann, H. G., (1998) Butterworth equations for ho-

momorphic filtering of images, Computers in Biology

and Medicine, 28, 169–181.