Personalized Recommender System for Virus Research and Diagnosis Laboratory Network

DST File No: DST/TSG/ICT/2013/37 Project Title: Design and Development of Computer Aided Diagnosis of Eye Diseases Diabetic Retinopathy & Glaucoma) and Watermarking of Medical Images for Tele-Ophthalmology.

An Adaptive Threshold Based Image Processing Technique for Improved Glaucoma Detection & Classification

In this proposed work, the discriminatory parameters of glaucoma infection, such as cup to disc ratio (CDR), neuro retinal rim (NRR) area and blood vessels in different regions of the optic disc has been used as features and fed as inputs to learning algorithms for glaucoma diagnosis. These features which have discriminatory changes with the occurrence of glaucoma are strategically used for training the classifiers to improve the accuracy of identification. The segmentation of optic disc and cup based on adaptive threshold of the pixel intensities lying in the optic nerve head region. Unlike existing methods the proposed algorithm is based on an adaptive threshold that uses local features from the fundus image for segmentation of optic cup and optic disc making it invariant to the quality of the image and noise content which may find wider acceptability. The experimental results indicate that such features are more significant in comparison to the statistical or textural features as considered in existing works. The proposed work achieves an accuracy of 94.11% with a sensitivity of 100%. A comparison of the proposed work with the existing methods indicates that the proposed approach has improved accuracy of classification glaucoma from a digital fundus which may be considered clinically significant.

Research paper from this work:

Ashish Issac, M. ParthaSarathi & Malay Kishore Dutta, “An Adaptive Threshold Based Image Processing Technique for Improved Glaucoma Detection & Classification” Computer Methods and Programs in Biomedicine, Elsevier Publishers, 122(2):229-44. doi: 10.1016/j.cmpb.2015.08.002, Nov. 2015.

Image Processing Based Automatic Diagnosis of Glaucoma using Wavelet Features of Segmented Optic Disc from Fundus Image

This work presents an automatic image processing based method for glaucoma diagnosis from the digital fundus image. In this paper wavelet feature extraction has been followed by optimized genetic feature selection combined with several learning algorithms and various parameter settings. Unlike the existing research works where the features are considered from the complete fundus or a sub image of the fundus, this work is based on feature extraction from the segmented and blood vessel removed optic disc to improve the accuracy of identification. The experimental results presented in this paper indicate that the wavelet features of the segmented optic disc image are clinically more significant in comparison to features of the whole or sub fundus image in the detection of glaucoma from fundus image. Accuracy of glaucoma identification achieved in this work is 94.7 % and a comparison with existing methods of glaucoma detection from fundus image indicates that the proposed approach has improved accuracy of classification.

Research paper from this work:

Anushikha Singh, Malay Kishore Dutta, M. ParthaSarathi, Vaclav Uher & Radim Burget, “Image Processing Based Automatic Diagnosis of Glaucoma using Wavelet Features of Segmented Optic Disc from Fundus Image” Computer Methods and Programs in Biomedicine, Elsevier Publishers, 2015, DOI: 10.1016/j.cmpb.2015.10.010.

Blood Vessel Inpainting based Technique for Efficient Localization and Segmentation of Optic Disc in Digital Fundus Images

This work presents a novel framework for fast and fully automatic detection of OD and its accurate segmentation in digital fundus images. The methodology involves optic disc centre localization followed by removal of vascular structure by accurate inpainting of blood vessels in the optic disc region. An adaptive threshold based Region Growing technique is then employed for reliable segmentation of fundus images. The proposed technique achieved significant results when tested on standard test databases like MESSIDOR and DRIVE with average overlapping ratio of 89% and 87% respectively. Validation experiments were done on a labeled dataset containing healthy and pathological images obtained from a local eye hospital achieving an appreciable 91% average OD segmentation accuracy.

Research paper from this work:

M.Parthasarathi, Malay Kishore Dutta, Anushikha Singh and Carlos Travieso, “Blood Vessel Inpainting based Technique for Efficient Localization and Segmentation of Optic Disc in Digital Fundus Images” Biomedical Signal Processing and Control, Elsevier Publishers, DOI : 10.1016/j.bspc.2015.10.012.

An Efficient Image processing based Technique for Comprehensive Detection and Grading of Non Proliferative Diabetic Retinopathy from Fundus Images

This paper proposes an automatic image processing method which can be an effective diagnostic tool to detect and grade the severity of diabetic retinopathy. This computer vision based algorithm imitates the logic and medical sense used by ophthalmologist in detecting the abnormality and its location in the image for grading the severity of the disease. The proposed algorithm aims at providing an all-inclusive diagnostic solution of efficiently detecting the symptoms of Diabetic Retinopathy and accurately grading the stage of the disease. The detection is based on finding abnormalities like exudates and red lesions and the grading is based on the location of these abnormalities in the image referred with distance from the macula. Accordingly the entire image has been divided into regions and occurrence of abnormalities in these regions indicates the severity of the disease. The methodology stresses on two major factors, the accuracy of the results and the computation time. It is found that both these design parameters are nicely achieved in the proposed method. The experimental results indicate that overall accuracy of 80% can be achieved in the proposed method and also significantly reduce the computational complexity in this region based approach.

Research paper from this work:

Malay Kishore Dutta, M.Parthasarathi, Shaumik Ganguly, Shaunak Ganguly and Kshitij Srivastava, “An Efficient Image Processing Based Technique for Comprehensive Detection and Grading of Non Proliferative Diabetic Retinopathy from Fundus Images” Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization, Taylor & Francis Publishers, DOI: 10.1080/21681163.2015.1051187, July 2015.

Unique Identification Code for Medical Fundus Images Using Blood Vessel Pattern for Tele-ophthalmology Applications

Identification of fundus images during transmission and storage in database for tele-ophthalmology applications is an important issue in modern era. The proposed work presents a novel accurate method for generation of unique identification code for identification of fundus images for tele-ophthalmology applications and storage in databases. Unlike existing methods of steganography and watermarking, this method does not tamper the medical image as nothing is embedded in this approach and there is no loss of medical information. Strategic combination of unique blood vessel pattern and patient ID is considered for generation of unique identification code for the digital fundus images. Segmented blood vessel pattern near the optic disc is strategically combined with patient ID for generation of a unique identification code for the image. The proposed method of medical image identification is tested on the publically available DRIVE and MESSIDOR database of fundus image and results are encouraging. Conclusions: Experimental results indicate the uniqueness of identification code and lossless recovery of patient identity from unique identification code for integrity verification of fundus images.

Research paper from this work:

Anushikha Singh & Malay Kishore Dutta, “Unique Identification Code for Medical Fundus Images Using Blood Vessel Pattern for Tele-ophthalmology Applications” Computer Methods and Programs in Biomedicine, Elsevier – Accepted for Publication. – DOI: 10.1016/j.cmpb.2016.07.011, Thomson Reuter Impact Factor –1.897.

Wavelet Based Reversible Watermarking System for Integrity Control and Authentication in Tele-Ophthalmological Applications.

In numerous applications, such as in the areas of law enforcement and medical imaging systems, together with perceptual transparency, it is required to reverse the modified media back to the original without any deformation, after the secret data are regained for some legal deliberations. This work proposes a completely reversible watermarking system with enhanced imperceptibility, robustness, and capacity for colour medical images which can effectively check the unlawful utilisation of the medical images with no affect to medical information and its visual quality. The proposed algorithm depends on transforming non-overlapping blocks of the host image using wavelet transform and is completely blind. The proposed technique attains high values of peak signal to noise ratio (PSNR) of watermarked image and high values of normalised correlation (NC) of the extracted watermark and recovered original image. State-of-the-art comparison shows that the proposed scheme has advantages over the existing techniques. This scheme may be proficient in providing a broad range of applications that aims at security and privacy protection in the medical field. The scheme can also be helpful for the purpose of identification of fundus images for storage in distributed medical databases and tele-ophthalmological applications.

Research paper from this work:

Abhilasha Singh, Malay Kishore Dutta, "Wavelet Based Reversible Watermarking System for Integrity Control and Authentication in Tele-Ophthalmological Applications", International Journal of Electronic Security and Digital Forensics, Inderscience Publishers, UK. Vol. 8, No. 4, 2016, pp- 366-391.pp-392-411

An adaptive algorithm for automatic detection of exudates based on localized properties of retinal fundus images

This work presents an algorithm for detecting exudates, abnormality in eyes, which provide a major step towards correctly identifying diabetic retinopathy (DR). The proposed algorithm is invariant to the illumination of the image and works well on poor contrast images with high reflection noise. The proposed algorithm correctly rejects the reflections and artifacts during segmentation of exudates from the fundus image despite the colour, intensity and contrast of reflections being almost similar to that of exudates. The algorithm is also computationally fast, efficient and detects exudates with a high precision. Optic disc (OD) localization and segmentation is performed using mathematical morphological filtering techniques like the averaging filter of a specially determined size which is an important step in the rejection of false positives in the process of detection of exudates. Exudates are located by first generating candidate regions using variance and median filters followed by morphological reconstruction techniques. Unlike existing methods, the exudates are detected by using an adaptive threshold that is based on the local properties of the image. The strategic selection of the local properties for deciding the threshold of the image makes this a novel approach that is highly accurate for detection of exudates. The proposed method was tested on a publically available labelled database and a local database containing 245 images overall. The method achieved a sensitivity of 96.765% and a positive predictive value of 93.514%. The computation time for OD localization is on average 2.467sec and the computation time for exudates detection is 16.814 sec, which makes the method suitable for real time applications.

Research paper from this work:

Katha Chanda, Ashish Issac, Malay Kishore Dutta, “An adaptive algorithm for automatic detection of exudates based on localized properties of retinal fundus images” – Journal of Biomedical Optics, SPIE Publishers, Submitted and Under Review.

Automatic System from Hierarchical detection and grading of Diabetic Retinopathy

Diabetic Retinopathy (DR) is an ocular disease found in diabetic patients that can lead to acquired blindness. Early detection and treatment can limit and reverse the progression of DR. The clinical manifestations of DR include bright lesions such as hard exudates and cotton wools spots and red lesions such as hemorrhages and micro-aneurysms in the retina. Computer aided DR detection involves the classification of these bright and red lesions from the fundus images of the eye. In most DR patients red lesions are indicative of the initial onset of DR while bright lesions appear at a later stage. This disparity in the onset time of the bright and red lesions is captured in the proposed hierarchical DR grading system. First, bright and red lesions are detected in fundus images to separate the normal images from the images affected by DR. Next, based on the number of bright and red lesions, the images affected by DR are further categorized with mild, moderate and severe DR grades. Experimental analyses show that the proposed system achieves classification sensitivity/specificity for bright and red lesions as 97%/89% and 94.2%/84.5%, respectively. Classification of normal images from DR images achieves an average sensitivity/specificity of 93.90%/76.49%. The proposed system achieves an overall DR grading sensitivity of 98.48%, which makes it very useful for screening applications.

Research paper from this work:

Namita Sengar, Malay Kishore Dutta, and Sohini Roychowdhury, “Automated System for Detection and Hierarchal Grading of Diabetic Retinopathy” Biomedical Signal Processing and Control, Elsevier Publishers, Submitted and under review.

Optic Disc and Cup Segmentation in Fundus Images

This work presents an image processing technique for segmentation of optic disc and cup based on adaptive thresholding using features from the image. This algorithm uses the features obtained from the image, such as mean and standard deviation, to remove information from the red and green channel of a fundus image and obtain an image which contains only the optic nerve head region in both the channels. The optic disc is segmented from the red channel and optic cup from the green channel respectively. The threshold is determined from the smoothed histogram of the preprocessed image. The results of the proposed algorithm are compared with the images that are marked by doctors. The accuracy of the algorithm is good and is computationally very fast. The proposed method can be used for screening purpose.

Research paper from this work:

Ashish Issac, Parthasarthi Mangipudi and Malay Kishore Dutta, “An Adaptive Threshold Based Algorithm for Optic Disc and Cup Segmentation in Fundus Images” – International Conference on Signal Processing and Integrated Networks, 2015, pp-143-147, Proceedings published by IEEE Xplore, New York, USA.

Glaucoma Detection using Adaptive Threshold based Technique in Fundus Image

The conventional methods to detect glaucoma like Optical Coherence Tomography (OCT) and Heidelberg Retinal Tomography (HRT) are expensive and need specialized manpower. A digital fundus image can be used to identify glaucoma. This work describes an efficient method to analyze a computer-aided fundus image which can act as a diagnostic tool for detection of glaucoma. A technique based on histogram of the image is used to study some statistical features of the image such as mean and standard deviation. A relationship between them is established to find a threshold value for segmenting optic disc and optic cup. An adaptive threshold based method which is independent of image quality and invariant to noise is used to segment the optic disc, optic cup and the cup-to-disc ratio CDR which is used to screen glaucoma. The experimental results obtained are compared with those of the ophthalmologist and are found to have a high accuracy. Also in addition the proposed method is efficient having a low computational cost.

Research paper from this work:

Ayushi Agarwal, Shradha Gulia, Somal Chaudhary, Malay Kishore Dutta, Radim Burget and Kamil Riha, “Automatic Glaucoma Detection using Adaptive Threshold based Technique in Fundus Image” 38th IEEE International Conference on Telecommunications and Signal Processing (TSP-2015), Prague, Europe, pp. 416-420, IEEE Xplore Digital Library, New York USA.

Detect Glaucoma in Retinal Fundus Images using Cup-Disk and Rim-Disk Ratio

In this work an adaptive threshold based method which is independent of image quality and invariant to noise is used to segment optic disk, optic cup, Neuroretinal rim and cup to disk ratio is calculated to screen glaucoma. Another ocular parameter, rim to disk ratio is also considered which in combination with CDR gives more reliability in determining glaucoma and makes the system more robust. Further an SVM classifier has been used to categorize the images as glaucomatic or non glaucomatic. The experimental results obtained are compared with those of ophthalmologist and are found to have high accuracy of 90%. Also in addition, the proposed method is faster having low computational cost.

Research paper from this work:

Ayushi Agarwal, Shradha Gulia, Somal Chaudhary, Malay Kishore Dutta, Carlos M. Travieso & Jesús B. Alonso Hernández, “A Novel Approach to Detect Glaucoma in Retinal Fundus Images using Cup-Disk and Rim-Disk Ratio” 4th IEEE International Conference and Workshop on Bioinspired Intelligence (IWOBI 2015) San Sebastián, SPAIN - June 10-12, 2015, pp. 139-144, IEEE Xplore Digital Library, New York, USA.

Grading for Non-Proliferative Diabetic Retinopathy using Region Based Detection

This work proposes an image processing algorithm to grade the severity of Non Proliferative Diabetic Retinopathy. For this disease the most important parameter to classify the stage of the disease is the proximity of abnormalities from the centre of Macula. The proposed algorithm in this work provides an efficient grading technique by segmenting the fundus image into specific regions of interest and avoids redundancy in computation. Instead of detecting abnormalities for the whole fundus image, the proposed method emphasizes on the segmented regions for the abnormalities, thereby reducing the computation time significantly. Furthermore, this approach provides a simple and direct method to measure the severity of the disease. This region based segmentation also has the advantage of a mesh lesser computational load making this process suitable for real time applications. The accuracy of this region based segmentation method is more than 80% when tested in a database.

Research paper from this work:

Shaumik Ganguly, Kshitij Srivastava, Shaunak Ganguly, Malay Kishore Dutta, M.Parthasarathi, Radim Burget, Jan Masek, “An Efficient Grading Algorithm for Non-Proliferative Diabetic Retinopathy using Region Based Detection” 38th IEEE International Conference on Telecommunications and Signal Processing (TSP-2015) July 2015, Prague, Europe, pp-743-747.

Exudates Detection in Digital Fundus Image

This work uses an image processing algorithm to accurately detect the presence of exudates in Fundus images. For Diabetic Retinopathy, presence of exudates in the fundus image marks the beginning of vision loss and hence detecting the exudates accurately and efficiently is of prime concern. The proposed algorithm in this work is a strategic method which removes false detection because of noise generated for different reasons. Using a strategic combination of two independent approaches based on threshold and edge detection helps in eliminating all possible types of noises leading to false exudates that may have crept in. Hence, this method of detecting the exudates has an advantage of increased accuracy. Experimental results indicate that this method has a clear advantage of accuracy in terms of exudates detection in the digital Fundus image without compromising the computational time.

Research paper from this work:

Kshitij Srivastava, Shaunak Ganguly, Shaumik Ganguly, Malay Kishore Dutta, M.Parthasarathi, Radim Burget, Jiri Prinosil, “Exudates Detection in Digital Fundus Image Using Edge Based Method & Strategic Thresholding” 38th IEEE International Conference on Telecommunications and Signal Processing (TSP-2015) July 2015, Prague, Europe, pp-748-752.

Extraction of Retinal Vasculature by using morphology in Fundus Images

In this work algorithm is proposed for detection of vessels present in a fundus image of an eye. Blood vessels extraction and removal are used to detect the other artifacts like lesions, the fovea and optic nerve. This algorithm used the combination of different morphological operators which make this method less complex and also computationally efficient. Two different channels of an image green and L respectively are utilized to get the final vessel structure. This method also gives the region of interest for macula which may make macula detection easy. The proposed algorithm is tested on DRIVE data set of fundus image of an eye. The result gives good detection of vessel structure and the proposed method is computationally efficient.

Research paper from this work:

Namita Sengar, Malay Kishore Dutta, Parthasarathi Mangipudi and Radim Burget, “Extraction of Retinal Vasculature by using morphology in Fundus Images” – International Conference on Signal Processing and Integrated Networks, 2015, pp-139-142, Proceedings published by IEEE Xplore, New York, USA.

Method for Detection of Macula in Retinal Images

This work demonstrates an automatic proficient intensity based approach for automatic detection and extraction of macula from the retinal fundus images in the field of teleopthalmology. Detection of the macula from a retinal image is an indispensable step for developing automated screening system for ophthalmic pathologies. For eye diagnosis, digital fundus images have become significant, thus opening up the possibility of applying digital image processing techniques in ocular fundus images to facilitate and improve diagnosis. The paper proposes a method for automatic and efficient detection and extraction of macula, based on the reference position of the optical disc from the fundus images. Further, suitable morphological steps are performed on the region of interest obtained with the help of centroid of optical disc to extract the exact location of macula. The proposed algorithm is computationally simple, efficient and can be used to assist ophthalmologists in as a diagnostic tool for screening of various eye related ailments. The experimental results from this indicate that this method of detection of macula is highly accurate and efficient.

Research paper from this work:

Arpit Bansal, Aashwin Vats, Akshita Jain, Malay Kishore Dutta, Radim Burget and Jiri Prinosil, “An Efficient Automatic Intensity Based Method for Detection of Macula in Retinal Images” 38th IEEE International Conference on Telecommunications and Signal Processing (TSP-2015), Prague, Europe, pp. 507-510, IEEE Xplore Digital Library, New York USA.

Detection of Diabetic Macular Edema in Retinal Images Using a Region Based Method

Diabetic macular edema (DME) is one of the severe complications of Diabetic Retinopathy. In this work macula centre is detected which is independent of optic disc location. Grading of DME is done by dividing the image of retina in different regions according to the international standard. Disease severity is accessed using scaling of bright lesions in macular regions. In this method search region for detection of macula is adaptive to the size of image. Independency from optic disc detection to detect the macula is an efficient method because it is unaffected by wrong detection of optic disc position under the presence of noises and reflections. The proposed method is tested on 100 images of MESSIDOR database and has achieved good accuracy 80 to 90 % for different cases.

Research paper from this work:

Namita Sengar, Malay Kishore Dutta, Radim Burget & Lukas Povoda, “Detection of Diabetic Macular Edema in Retinal Images Using a Region Based Method”, 38th IEEE International Conference on Telecommunications and Signal Processing (TSP-2015), Prague, Europe, pp. 262-265, IEEE Xplore Digital Library, New York USA.

Exudates Detection in Fundus Image using Intensity Thresholding and Morphology

In this work for detection of exudates, two independent approaches based on intensity thresholding and morphological processing are strategically combined to detect any small exudates present while removing all possible types of false positives. This strategic combination removes the noise sources from blood vessels and reflections during image capture making the detection of exudates accurate. Experimental results indicate that the proposed method has good accuracy in exudates detection without compromising the computational time and hence can be considered for screening purpose of DR.

Research paper from this work:

Anushikha Singh, Namita Sengar, Malay Kishore Dutta & Kamil Riha, “Automatic Exudates Detection in fundus image using Intensity Thresholding and Morphology” -7th IEEE International Congress on Ultra Modern Telecommunications and Control Systems, Brno, Czech Republic2015, pp. 330-334, IEEE Xplore Digital Library, New York, USA.

Automatic Extraction of Blood Vessels and Veins using Laplace Operator in Fundus Image

This work utilizes an approach of preprocessing of image by using adaptive histogram equalization by CLAHE algorithm of green channel of fundus retinal image. Subsequently, using Laplace operator as key point of proposed algorithm and subsequently is applied the operation erosion processed image and removed small segments from image to enhance extraction of blood vessels from fundus image. The proposed technique analyzes detection and evaluates precision of the method on dataset from public fundus image libraries DRIVE, and HRF and compare with reference training results provided by these libraries.

Research paper from this work:

Jiri Minar, Marek Pinkava, Kamil Riha, Malay Kishore Dutta, Namita Sengar, “Automatic Extraction of Blood Vessels and Veins using Laplace Operator in Fundus Image”, International Conference on Green Computing and Internet of Things (ICGCIoT 2015), Accepted, proceedings will be published by IEEE Xplore Digital Library, New York, USA.

Hierarchical Classification and grading of Diabetic Macular Edema using Texture Features

Diabetic macula edema (DME) is an eye pathology, a complication of diabetic retinopathy, which is caused due to the presence of exudates around the fovea. In this paper, an automated method for robust classification and grading of DME is presented. The algorithm proposed presents a computerized method of processing the images in the database, extracting texture features in both spatial domain and wavelet domain from sub-regions with a specified radius around the macula. Unlike other well-known approaches of machine learning classifiers, we propose a method that processes the specific sub-regions of interests instead of the whole image which makes it computationally efficient. Grading of the disease into 3 stages namely normal, moderate and severe diabetic macular edema based on severity is done in a hierarchal manner.

Research paper from this work:

Shrey Magotra, Aditya Kunwar, Namita Sengar, Partha Mangipudi, Malay Kishore Dutta, “Hierarchical Classification and Grading of Diabetic Macular Edema using Texture Features” IEEE International Conference on Image Information Processing 2015, pp-185-189, IEEE Xplore Digital Library, USA.

Digital Identification Tags for Medical Fundus Images for Tele-Ophthalmology Applications

This work proposes a method of inserting a digital pattern having patient identity in the medical image without tampering the medical information of the image. To attain imperceptible insertion of the digital pattern a frequency domain approach is used in the mid frequency band of the discrete cosine transform. The original medical image and the stego-image is compared and analyzed for all features and also tested for retaining of all features and medical information. Blood vessels have been extracted from the original and stego image and it has been established from experimental results that the features remains unaltered. Texture features also has been analyzed and experimental results indicates that the variation in the texture features is minimal and do not affect the medical information. The correlation of the features extracted is above 0.99 indicating the insertion of the digital pattern did not cause any loss of medical information in the image.

Research paper from this work:

Malay Kishore Dutta, Anushikha Singh, Abhilasha Singh, Radim Burget, Jiri Prinosil, “Digital Identification Tags for Medical Fundus Images for Tele-Ophthalmology Applications” 38th IEEE International Conference on Telecommunications and Signal Processing (TSP-2015) July 2015, Prague, Europe, pp-781-784, IEEE Xplore Digital Library, USA.

Classification of Glaucoma Based on Texture Features Using Neural Networks

This work is an attempt to study and analyze the texture features of the Fundus image and its variations when the Fundus image is infected with glaucoma. The texture features extracted are localized around the optic cup which gives clear results for the purpose of distinct identification and classification. The classification method proposed is use of neural network classifier with the help of texture feature extraction of the localized area of the optic cup of the fundus images. The classification method gives high level of accuracy based on the different test-train ratios. The experimental results are encouraging indicating an accuracy of above 90% accuracy in classification.

Research paper from this work:

Deepti Yadav, M.Parthasarathi & Malay Kishore Dutta, “Classification of Glaucoma Based on Texture Features Using Neural Networks” Seventh IEEE International Conference in Contemporary Computing, August 2014, pp. 109 - 112, IEEE Xplore, Digital Library, New York USA.

Glaucoma Detection by Segmenting the Super Pixels from Fundus Colour Retinal Images

This work proposes an automated image processing approach for detection of glaucoma which may be a diagnostic tool to help ophthalmologist in mass screening of glaucoma suspects. The proposed approach is based on the segmentation of optic disk and the optic cup and computing the cup-to-disc ratio. For segmentation of optic cup and optic disk, a double threshold method is used, one for removing blood vessels and background and second threshold for segmenting the super intensity pixels contained by the optic disk and optic cup. Further, Hough Transform is used to calculate the radius of optic disk and optic cup. The vertical cup to disk ratio is used as a parameter for identification of glaucoma symptoms in the fundus image. The results of the proposed method indicate that the approach is effective in glaucoma detection with better accuracy over existing methods.

Research paper from this work:

Malay Kishore Dutta, Amit Kumar Mourya, Anushikha Singh, M Parthasarathi, RadimBurget & Kamil Riha, “Glaucoma Detection by Segmenting the Super Pixels from Fundus Colour Retinal Images” International Conference on Medical Imaging, m- Health & Emerging Communication System (MEDCOM 2014), Nov 2014, pp. 86-90, IEEE Xplore, Digital Library, New York USA.

An Efficient Automatic Method of Optic Disc Segmentation using Region Growing Technique in Retinal Images

This work proposes a method of automatic optic disk segmentation based on region growing technique with automatic seed selection. In this method centre of optic disk is considered as a seed to apply region growing technique to segment the optic disk from the preprocessed retinal image. Automatic detection of centre of optic disk is done by double windowing method. The algorithm uses image processing techniques like contrast adjustment, morphological operations & filtering to process the retinal image and to remove the blood vessels from the retinal image. The performance of optic disk segmentation by proposed method is compared with Optic disk segmentation by ophthalmologists and results are found convincing and efficient. The experimental results indicate this method of segmentation of the OD has good accuracy and also is computationally cheap.

Research paper from this work:

Anushikha Singh, Malay Kishore Dutta & M.Parthasarathi Radim Burget & Kamil Riha, “An Efficient Automatic Method of Optic Disc Segmentation using Region Growing Technique in Retinal Images” IEEE International Conference on Contemporary Computing and Informatics”, 2014, pp. 480-484.

Dynamic Thresholding Technique for Detection of Hemorrhages in Retinal Images

This work proposes a dynamic thresholding based image processing technique for the detection of hemorrhages in retinal images. The algorithm uses the information about color and size of hemorrhages as a tool for classifying hemorrhages from other dark lesions present in the retinal images. The algorithm uses the concepts of contrast enhancement, background estimation and intensity variation at edges that is gradient magnitude information supported by some morphological operations. The algorithm follows a simple approach of step by step removal of unwanted features from targeted images using concepts of thresholding and morphology without compromising with accuracy and time of execution. The experimental results indicate that hemorrhages are detected with good accuracy in the retinal images.

Research paper from this work:

Akhilesh Sharma, Malay Kishore Dutta, Anushikha Singh, M.Parthasarathi & Carlos M. Travieso, “Dynamic Thresholding Technique for Detection of Hemorrhages in Retinal Images”, Seventh IEEE International Conference in Contemporary Computing, August 2014, pp. 113 - 116, IEEE Xplore Digital Library, New York, USA.

An Adaptive Threshold Based Algorithm for Detection of Red Lesions of Diabetic Retinopathy in a Fundus Image

The work proposes an algorithm for detection of Red Lesions present in a fundus image of an eye. Red Lesions include Micro-aneurysms and Hemorrhages, which are the symptoms of Diabetic Retinopathy, a widespread eye disease which affects almost every diabetic patient at some point of the patients life. The paper presents an adaptive method to detect the red lesions present in an image. The proposed method will estimate the upper threshold and the lower threshold of the red lesions for the given fundus image individually based on local image information. The significance of the adaptive nature of this proposed algorithm is that fundus images acquired from different cameras may vary in quality and resolution. As a result the intensity of red lesions may vary from image to image. Since, the intensity of red lesions is similar to that of the blood vessels for a specific image, therefore this similarity has been utilized to develop an accurate, adaptive algorithm for the detection of red lesions, wherein every fundus image is processed with a different intensity threshold value resulting more accurate detection.

Research paper from this work:

Shaunak Ganguly, Shaumik Ganguly, Kshitij Srivastava, Malay Kishore Dutta, M.Parthasarathi, Radim Burget & Kamil Riha, “An Adaptive Threshold Based Algorithm for Detection of Red Lesions of Diabetic Retinopathy in a Fundus Image”, International Conference on Medical Imaging, m- Health & Emerging Communication System (MEDCOM 2014), Nov 2014, pp. 91-94, IEEE Xplore, Digital Library, New York USA.

Imperceptible Digital Watermarking in Medical Retinal Images for Tele-Medicine

This work proposes a method of inserting a digital pattern having patient identity information in the medical retinal image without changing the perceptual property and without causing any loss of medical information of this image. To achieve this insertion of the digital signature identity of the patient is done in the singular value based decomposition (SVD) domain of the image. After the insertion of this digital signature is done in the retinal image a detailed comparative study and analysis is done between original image and marked image to test if all the medical information and features of the image is retained. The important feature of the medical image like blood arteries, macula and optic disc has been segmented from the original and stego-image. The experimental results indicate that the original and the stego image have similar perceptual properties and no medical information is lost in the process of digital watermarking. Features of the original and stego image has been analyzed and experimental results indicates that the variation in the features is minimal and do not affect the medical information which has been validated by professional ophthalmologists. The correlation of the features extracted is above 0.99 indicating the insertion of the digital pattern did not cause any loss of medical information in the image.

Research paper from this work:

Malay Kishore Dutta, Anushikha Singh, M.Parthasarathi & Carlos M. Travieso, “Imperceptible Digital Watermarking in Medical Retinal Images for Tele-Medicine Applications” IEEE International Conference on Contemporary Computing and Informatics”, 2014. pp. 517-521.

Fast localization of the optic disc in fundus images using region-based segmentation

Optic disc (OD) is considered as one of the primary features in retinal fundus images. Detection of the OD is important for the identification and severity assessment of various ophthalmic pathologies. OD localization is the first step towards accurate OD segmentation process. In this paper an adaptive region-based image segmentation method is presented for automated localization of the OD. The proposed method is tested on two publically available datasets of MESSIDOR and DRIVE. For these two datasets, the OD was successfully localized in 90 images out of 100 (90% success) and 38 images out of 40 (95 % success), respectively, with computation time of approximately 1.6 seconds per image. Experimental results indicate that the proposed method is successful in fast and robust OD localization and therefore this method can be useful for real-time automated ophthalmic pathology detection systems.

Research paper from this work:

Namita Sengar, Malay Kishore Dutta, M. Parthasarathi, Sohini Roychowdhury and Radim Burget, "Fast localization of the optic disc in fundus images using region-based segmentation," 2016 3rd International Conference on Signal Processing and Integrated Networks (SPIN), Noida, 2016, pp. 529-532. doi: 10.1109/SPIN.2016.7566752, IEEE Xplore, Digital Library, New York, USA.

Automatic detection of red lesions in Diabetic Retinopathy using Shape based extraction technique in fundus image

The paper proposes an automatic image processing algorithm based on shape features for the detection of red lesions. In Diabetic Retinopathy Micro-aneurysms and hemorrhages comes under the category of red lesions, which is the most common eye disease caused in diabetic patients and also leads to blindness. This paper describes an effective methodology to study any computer-aided fundus image that can be utilized as a tool for diagnosis and detection of red lesions. A Shape based extraction technique using three parameters i.e. Perimeter Area and Eccentricity is used to segment out the red lesions from rest of the image. Since the algorithm considers shape features for detection of red lesion which makes it efficient and independent of image quality. The results that are experimentally obtained by applying this algorithm has been compared with those of the ophthalmologist and the comparison of these results have been highly accurate. In addition to the accuracy of the obtained results, the proposed method is fast and having very low computational time.

Research Paper from this work:

Aman Pandey, Ritu Chandra, Malay Kishore Dutta, Radim Burget, Vaclav Uher and Jiri Minar, "Automatic detection of red lesions in Diabetic Retinopathy using Shape based extraction technique in fundus image," 2016 39th International Conference on Telecommunications and Signal Processing (TSP), Vienna, Austria, 2016, pp. 538-542. doi: 10.1109/TSP.2016.7760938, IEEE Xplore, Digital Library, New York, USA.

An SVD based zero watermarking scheme for authentication of medical images for tele-medicine applications

The proposed work presents a zero watermarking method for to solve the issue of medical image security for telemedicine, tele-radiology & tele-opthalmology applications. This method provides medical image security for tele-medicine application without tempering the medical image and no loss of clinical information. Local features in the Singular value decomposition (SVD) domain are used to generate a digital binary code (Master Share) for each fundus image. This master share is strategically combined with encrypted patient ID resulting into a secret share. At the diagnosis centre the patient ID can be accurately recovered by the authorized person only on access of the generated Secret Share. The proposed zero watermarking method is tested on the publically available DRIVE dataset of fundus image and results achieved are encouraging in the direction of medical image identification and verification. The proposed work can be used in telemedicine applications where perfect and loss-less identification is required for medical images as this has direct relevance to human life.

Research Paper from this work:

Anushikha Singh, Namrata Raghuvanshi, Malay Kishore Dutta, Radim Burget and Jan Masek, "An SVD based zero watermarking scheme for authentication of medical images for tele-medicine applications," 2016 39th International Conference on Telecommunications and Signal Processing (TSP), Vienna, Austria, 2016, pp. 511-514. doi: 10.1109/TSP.2016.7760932, IEEE Xplore, Digital Library, New York, USA.

Use of Morphological Techniques and Active Contour Fitting techniques to segment Optic disc correctly

Optic disc segmentation is a crucial step in diagnosis of various ocular diseases like Glaucoma and Diabetic Retinopathy. This work proposes a technique for automatic detection of optic disc from the fundus images using edge based and active contour fitting method. The proposed work has used image processing techniques such as smoothing filters for removal of blood vessels, morphological operations to correctly segment the optic disc and reject the false positives, active contour snake based model for smoothing of optic disc boundary. The results of optic disc segmentation obtained from the proposed work are compared with the ground truth marked by the ophthalmologists. The results are convincing and segmentation results show that the method has good accuracy. An average overlapping score of more than 90% is obtained for the fundus images under test.

Research paper from this work:

Ashi Agarwal, Ashish Issac, Anushikha Singh, Malay Kishore Dutta “Automatic Imaging Method for Optic Disc Segmentation using Morphological Techniques and Active Contour Fitting” 9th IEEE International Conference on Contemporary Computing, IC3 2016, Proceedings will be published by IEEE Xplore Digital Library.

Method for detection of bright lesions from contrast normalized fundus images

Exudates are one of the abnormalities present in the eye which can lead to vision loss. Fundus images may consist of artifacts which occur during image acquisition and hamper the accuracy of detection of exudates. There is a need to develop an image processing based techniques for automated and correct segmentation of exudates from fundus images. This work demonstrates an automatic computer vision algorithm for efficient identification of the exudates from fundus images by strategic fusion of techniques i.e. contrast normalization, top-hat transformation and average filtering. The proposed technique correctly detects exudates from the fundus images and rejects the artifacts and reflections. The average computation time for exudates segmentation from fundus images is 11 seconds. The proposed method is computationally efficient and robust and can be used for real time applications.

Research paper from this work:

Ashish Issac, Rishabh Madan, Malay Kishore Dutta, Carlos M. Travieso, “Automated detection of bright lesions from contrast normalized fundus images” 9th IEEE International Conference on Contemporary Computing, (IC3 2016), Proceedings will be published by IEEE Xplore Digital Library.

Method for Exudates Segmentation using Image Normalization and Histogram Analysis

Exudates are one of the abnormalities present on the retina which are used for identification of diseases like Diabetic Retinopathy and Macular Edema. There arises a need for automated and correct segmentation of exudates from digital fundus images. This work proposes an automated computer vision technique for efficient exudates segmentation from fundus images. The proposed method segments the exudates using an adaptive intensity based threshold which is selected by strategically combining first order statistical parameters and local thresholding based method. The proposed technique correctly detects exudates from the fundus images with an average computation time of 9 seconds. The proposed method is computationally fast and can be used in image processing based applications for diagnosis of ocular diseases.

Research paper from this work:

Ashmita Gupta, Ashish Issac, Namita Sengar, Malay Kishore Dutta, "An Efficient Automated Method for Exudates Segmentation using Image Normalization and Histogram Analysis", 9th IEEE International Conference on Contemporary Computing, IC3 2016, Proceedings will be published by IEEE Xplore Digital Library.

Detection of Macula in Fundus Images

The Macula is an important part of the retina of human eye which is responsible for sharp central vision. Accurate and automatic detection of macula from fundus images is an essential step to develop automated screening tool for ocular pathologies. The proposed work presents an imaging method for detection of macula in fundus images automatically. The proposed method includes a strategic windowing based approach for accurate detection of macula. Instead of searching macula from whole fundus image, a search region is considered with the help of optic disc and then macula is detected from that search region using double windowing based method. Normal and affected fundus images from a local eye hospital were used to test the performance of proposed method and achieved encouraging results. The proposed method of macula detection is accurate, computationally cheap and hence can be helpful in real time automated screening of various eye diseases.

Research paper from this work:

Anushikha Singh, Namita Sengar, Ashish Issac, Malay Kishore Dutta, “An Automated Imaging Algorithm for Detection of Macula in Fundus Images” 9th IEEE International Conference on Contemporary Computing, IC3 2016, Proceedings will be published by IEEE Xplore Digital Library.

Segmentation of Blood Vasculature from Retinal Images

In this work algorithm is proposed for vessel extraction present in a fundus image of an eye. Blood vessels detection and removal are used to find the other features or abnormalities like red lesions, optic nerve and fovea. The proposed method utilized the strategic combination of green and L channel to get the final vessel structure which increases the accuracy. The proposed algorithm used the combination of morphological operators and intensity based thresholding which make this method less complex and also computationally efficient. The proposed algorithm is tested on DRIVE data set of fundus image of an eye. The result shows the better comprehensive performance of vessel extraction and the proposed method is also computationally efficient.

Research paper from this work:

Varun Gupta, Namita Sengar, Malay Kishore Dutta, “Automated Segmentation of Blood Vasculature from Retinal Images” 2nd IEEE International Conference on Communication Control and Intelligent System (CCIS-2016), Proceedings will be published by IEEE Xplore Digital Library.

Optic Disc Detection from Non-uniform Illuminated Digital Fundus Images

Accurate segmentation of optic disc from fundus images is an essential step to develop automatic screening device for eye pathologies. Automatic and correct segmentation of optic disc from non-uniformly illuminated or abnormal/affected fundus images is still a challenging issue. The proposed work presents a computer vision based method for optic disc segmentation from variety of normal, affected and blurred/non uniform illuminated fundus images. The methodology involves separation of super pixels from fundus images followed by removal of false positives like reflections, exudates, choroid vessels using analysis of geometrical features for correct optic disc segmentation. The proposed method was tested on both normal and abnormal/affected fundus images obtained from local eye hospital and achieved 90.78% overlapping ratio. The proposed OD segmentation method is robust and computationally cheap which makes it applicable for real time.

Research paper from this work:

Ashish Issac, Namita Sengar, Anushikha Singh, M. Partha Sarathi, Malay Kishore Dutta, Carlos M. Travieso, “Automated Computer Vision Method for Optic Disc Detection from Non-uniform Illuminated Digital Fundus Images” 2nd IEEE International Conference on Communication Control and Intelligent System (CCIS-2016), Proceedings will be published by IEEE Xplore Digital Library.

Method for Macula Localization from Fundus Images

Localization of macula from fundus image plays an important role to design an automated screening tool for detection of retinal diseases. The similar color and texture of red lesions act as a bottleneck in accurate localization of macula in the fundus image. This work presents a computer vision algorithm for automated and efficient localization of macula from low contrast and diabetic retinopathy affected fundus images. A statistical based model is used to detect macula in a specified region of fundus image which is designed using the geometric features of optic disc. The performance of the proposed algorithm of macula detection was tested on 200 normal/affected fundus images and presents an accuracy of 92.5%. The computational efficiency and accurate localization of macula makes the proposed method competent enough to be used as a part of an automated screening tool for detection of retinal diseases.

Research paper from this work:

Ashish Issac, Namita Sengar, Anushikha Singh, Malay Kishore Dutta, Jiri Prinosil, Kamil Riha, “An Efficient Imaging Technique for Automated Macula Localization from Fundus Images”, 8th IEEE International Congress on Ultra Modern Telecommunications and Control Systems and Workshops (ICUMT), Lisbon, Portugal, 2016, pp. 387-391. doi: 10.1109/ICUMT.2016.7765390, IEEE Xplore, Digital Library, New York, USA.

An Efficient and Lossless Cryptosystem for security in Tele-Ophthalmology Applications

Tele-ophthalmology has gained a lot of popularity as it involves retinal fundus images which can be analyzed for identification of severe diseases like diabetic retinopathy and glaucoma. With this increasing popularity, requirement for medical data confidentiality and privacy has also increased during transmission or storage. To meet this challenge, this paper proposes an efficient and lossless cryptosystem based upon chaotic theory for encryption of medical fundus images. In the proposed encryption scheme a strategic combination of scrambling and substitution architecture is proposed which complements each other. The proposed scheme of encryption for fundus images is challenging as these images are 3-D color image and cannot be compressed as compression may not be able to retain all relevant medical information. For performance analysis, the proposed algorithm has been evaluated for perceptual and cryptographic security. The experimental results indicate that the proposed method is lossless and resistant against attacks making the proposed scheme suitable for real time applications.

Research Paper from this work:

Garima Mehta, Malay Kishore Dutta, Pyung Soo Kim “An Efficient and Lossless Cryptosystem for security in Tele-Ophthalmology Applications Using Chaotic Theory”, International Journal of E-Health and Medical Communications, Vol. 7, December 2016, pp. 28-47.

Automatic extraction of blood vessels and veins using adaptive filters in Fundus image

The paper proposes a novel method for extraction of blood vessels and veins from medical image of human eye - retinal fundus images that can be used in ophthalmology for detecting various eyes' diseases such glaucoma, diabetic retinopathy or macula oedema. The method utilizes an approach of preprocessing of image by using adaptive histogram equalization by CLAHE algorithm of green channel of fundus retinal image. Subsequently, using adaptive filters and image convolution with filter mask as key point of proposed algorithm and subsequently is applied the operation erosion processed image and removed small segments from image to enhance extraction of blood vessels from fundus image. The proposed technique analyzes detection and evaluates precision of the method on dataset from public fundus image libraries DRIVE, and HRF and compare with reference training results provided by these libraries.

Research Paper from this work:

Jiri Minar, Marek Pinkava, Kamil Riha, Malay Kishore Dutta, Anushikha Singh and H. Tong, "Automatic extraction of blood vessels and veins using adaptive filters in Fundus image," 2016 39th International Conference on Telecommunications and Signal Processing (TSP), Vienna, Austria, 2016, pp. 546-549, IEEE Xplore, Digital Library, New York, USA.

Automated Imaging Algorithm for Detection of Macula in Fundus Images

This work presents an imaging method for detection of macula in fundus images automatically. This method includes a strategic windowing based approach for accurate detection of macula. Instead of searching macula from whole fundus image, a search region is considered with the help of optic disc and then macula is detected from that search region using double windowing based method. This method of macula detection is accurate, computationally cheap and hence can be helpful in real time automated screening of various eye diseases.

Research paper from this work:

Anushikha Singh, Namita Sengar, Ashish Issac, Malay Kishore Dutta, “An Automated Imaging Algorithm for Detection of Macula in Fundus Images” 9th IEEE International Conference on Contemporary Computing, IC3 2016, pp. 1-4, IEEE Xplore Digital Library, New York USA.

Automated classification of exudates from digital fundus images

In the following research paper, a robust method has been proposed to segment hard exudates from digital, color fundus images using anisotropic diffusion and adaptive thresholding followed by a support vector machine for classification. The geometrical, shape and orientation features have been used to correctly classify the segmented objects as exudates or false pixels. The proposed technique has a high specificity and eliminates false positives correctly when applied across a wide range of images. The exudates segmented have a high degree of accuracy and no false positives are generated in case of non-diseased images. The proposed method has been tested on a total 189 images of the DIARETDB1 and MESSIDOR database and achieves an accuracy of 92.13% and 90% respectively. The proposed method can be used in the development for some computer aided technology for ocular diseases detection from fundus images.

Research Paper from this work:

Ravitej Singh Rekhi, Ashish Issac, Malay Kishore Dutta, Carlos M. Travieso, “Automated classification of exudates from digital fundus images” 5th IEEE International Work Conference on Bioinspired Intelligence (IWOBI 2017), Funchal, Portugal, 2017, pp. 1-6.

Automated Detection of Suspected Glaucoma in Digital Fundus Images

In this work an algorithm is proposed to detect suspected glaucoma by using the presence or absence of hemorrhages in a particular region, near the optic disc, in fundus image. Unlike existing methods, which only uses cup to disc ratio as a deciding parameter to detect glaucoma, this method helps to diagnose the case of suspected glaucoma efficiently. The optic disc and hemorrhages are segmented in a particular region automatically by using adaptive thresholding and some geometrical features. This algorithm achieves accuracy of 93.57% on digital fundus images for detection of suspected glaucoma.

Research paper from this work:

Namita Sengar, Malay Kishore Dutta, Radim Burget, Martin Rajnoha, “Automated Detection of Suspected Glaucoma in Digital Fundus Images”- 40th IEEE International Conference on Telecommunications and Signal Processing, Barcelona, Spain- Accepted for Publication

Automated detection and grading of Diabetic Macular Edema from digital colour fundus images

A robust method is proposed in this work to segment exudates from fundus images using a support vector machine for classification. This is followed by detection of macula using morphological features and comparing the segmented exudates against specially created regions around the macula. The algorithm further grades the image according to how severe the disease is. The proposed algorithm has been tested on 89 images of DIARETDB1 database and gives an accuracy of 92.11% in detecting severe case of DME and 90%accuracy for the same when tested on 100 images of MESSIDOR database. The proposed method can be used to directly give an idea about the severity of Diabetic Macular Edema in the image ad give a reliable and cost-effective diagnosis in real time.

Research Paper from this work:

Ravitej Singh Rekhi, Ashish Issac, Malay Kishore Dutta, “Automated detection and grading of Diabetic Macular Edema from digital colour fundus images”, 4th IEEE Uttar Pradesh Section International Conference (UPCON) on Electrical, Computer and Electronics, October 26-28, 2017 – GLA University, Mathura, India, Accepted for Publication, Proceedings will be published by IEEE Xplore, New York, USA.

Wavelet Based Robust Watermarking Scheme for Copyright Enforcement and Integrity Control in Tele-Ophthalmology applications.

Tele-medicine permits medical images to be transmitted among health care centers through the insecure open networks for clinical investigation and for improved healthcare stipulations. There is a growing need for copyright enforcement technologies in these networked multimedia systems. This work proposed an advanced and modified method to watermark information into a biomedical retinal cover image which aims to provide a robust and adaptive system for medical image protection against authentication and copyright infringement issues in tele-ophthalmological applications. Watermark is embedded in low frequency band using Singular values. Both methods strategically combined increase the level of security of proposed method. In short, a as a new SVD-DWT watermarking algorithm which is robust against various attacks is presented.

Research Paper from this work:

Abhilasha Singh, Malay Kishore Dutta, Jiri Prinosil, Kamil Riha, “Wavelet Based Robust Watermarking Scheme for Copyright Enforcement and Integrity Control in Tele-Ophthalmology” – 8th International Congress on Ultra-Modern Telecommunications and Control Systems, Lisbon, Portugal, European Union, pp. 408-413, 2017, IEEE Xplore Digital Library, New York USA.

Glaucoma Detection using novel blood vessel tracking and bend point detection

Correct segmentation of objects like optic disc and cup results in greater accuracy of Glaucoma detection from a fundus image. In this work, optic disc has been segmented using intensity based thresholds and the noise has been removed using a geometrical feature based framework to result in correct optic disc segmentation. The blood vessels on the optic disc have been segmented using a novel tracking method in which an elliptical window is placed across the cross-section of blood vessels and an inverted Gaussian profile is obtained from the intensity plot and the combination of profile centroid, extreme points and angle difference between two consecutive Gaussian profiles are used to propagate in the vessel direction. During tracking process, the vessel bends are identified and stored if the propagation direction changes sharply. Such bends in the vessels are detected and joined to obtain optic cup boundary. Finally, vertical CDR is considered to determine the severity of Glaucoma from the fundus image.

Research Paper from this work:

Soorya M., Ashish Issac, Malay Kishore Dutta, “An image processing algorithm for automated and robust Glaucoma Diagnosis from fundus images using novel blood vessel tracking and bend point detection”, Submitted and under review.

Deep learning based classifier for automated diagnosis of glaucoma from fundus image

This work developed deep learning based classifier for automated diagnosis of glaucoma from fundus image to enable treatment at its early stages. This method involves deep learning using convolutional neural network (CNN). A pre-trained CNN, VGG-16 is exploited for classification of image as either glaucomatous or normal, using transfer learning, where the pre-trained network is used to extract robust and suitable features. The developed classifier consists of a small dense neural network on top of the VGG-16 convolutional base. Dropout, data augmentation and regularization strategies are adopted to improvise the quality of classifier. This method can be used for glaucoma screening diagnostic tool.

Research Paper from this work:

Shubham Mittal, M ParthaSarathi, Malay Kishore Dutta, “Transfer Learning for Glaucoma Detection using Convolutional Neural Network”, Submitted to Computer Methods and Programs in Biomedicine, Elsevier Publishers, under review.

Imperceptible Watermarking for Security of Fundus Images in Tele-Ophthalmology Applications and Computer Aided Diagnosis of Retinal Diseases

Accurate identification of medical images and patient verification is an essential requirement to prevent error in medical diagnosis. In this work, an imperceptible watermarking system is developed to address the security issue of medical fundus images for tele-ophthalmology applications and computer aided automated diagnosis of retinal diseases. Patient identity is embedded in fundus image in singular value decomposition domain with adaptive quantization parameter to maintain perceptual transparency for variety of fundus images like healthy fundus or disease affected image. Insertion of watermark in fundus image does not disturb the image processing based automatic diagnosis of retinal objects & pathologies which ensure no uncompromised computer based diagnosis associated with fundus image. Patient ID is correctly recovered from watermarked fundus image for integrity verification of fundus image at the diagnosis centre. Correct recovery of patient ID from watermarked fundus image makes the proposed watermarking system applicable for authentication of fundus images for computer aided diagnosis and Tele-ophthalmology applications

Research Paper from this work:

Anushikha Singh & Malay Kishore Dutta, “Imperceptible Watermarking for Security of Fundus Images in Tele-Ophthalmology Applications and Computer Aided Diagnosis of Retinal Diseases” – Biomedical Signal Processing and Control, Elsevier – Submitted and under review.

Zero-Watermarking Scheme on fundus images for Tele-Ophthalmological Applications

This work proposed a zero-watermarking system which strategically uses singular value coefficients in Wavelet domain for generation of unique identification code which suits to the specific requirements of fundus images and optimizes the conflicting requirements of traditional watermarking. The technique aims at unique identification, authentication and integrity verification of medical images. This technique does not tamper images as no data is being embedded into the image. The proposed system is robust against image processing attacks and suitable for secure exchange of medical images and for storage in large distributed medical databases.

Research Paper from this work:

Abhilasha Singh, Malay Kishore Dutta, "A Robust Zero-Watermarking Scheme for Tele-Ophthalmological Applications", IETE Journal of Research, Taylor and Francis publishers, Submitted and under review.

Wavelet-SVD based zero Watermarking scheme for security of fundus image.

This work presented a Wavelet transform–Singular Value Decomposition based robust zero-watermarking technique for medical images to address the privacy and security issues. The proposed method conserves the reliability of the cover image without bringing any artifacts and without any change in the medical information of the image. The performance of the scheme is assessed with tele-ophthalmological images. The proposed watermarking scheme is robust against various image processing attacks and also suitable for safe exchange of medical images among remote medical practitioners.

Research Paper from this work:

Abhilasha Singh, Malay Kishore Dutta, “Wavelet-SVD based Zero Watermarking Scheme for Tele-Ophthalmological Applications” Next Generation Computing Technologies (NGCT-2017)- Submitted and under review.

An Integrity Control System for Fundus Images

This work proposed a watermarking algorithm which utilizes the advantages of Reversible, Region of Non-Interest and Zero Watermarking and collectively conforms to all the unique requirements of medical image watermarking viz. complete restoration of original cover image after watermark extraction, vital parts being not tampered at all and lossless watermarking respectively. The proposed system is fragile enough to ensure the integrity of medical image. Proposed algorithm is also suitable for the purpose of unique identification of medical images during transmission and storage in large distributed databases along with enhanced security and confidentiality with applications in tele-ophthalmological applications.

Research Paper from this work:

Abhilasha Singh & Malay Kishore Dutta, “An Integrity Control System for Retinal Images Based on Watermarking” – Biomedical Signal Processing and Control, Elsevier – Submitted and under review.