Browsing by Author "Raman, Sundaresan"

Now showing 1 - 20 of 22

Advances in the diagnosis of herpes simplex stromal necrotising keratitis: A feasibility study on deep learning approach
(India Journal of Ophthalmology, 2022-09) Raman, Sundaresan
Infectious keratitis, especially viral keratitis (VK), in resource-limited settings, can be a challenge to diagnose and carries a high risk of misdiagnosis contributing to significant ocular morbidity. We aimed to employ and study the application of artificial intelligence-based deep learning (DL) algorithms to diagnose VK.
Advances in the diagnosis of herpes simplex stromal necrotising keratitis: A feasibility study on deep learning approach
(Wolters Kluwer, 2022-09) Raman, Sundaresan
Infectious keratitis, especially viral keratitis (VK), in resource-limited settings, can be a challenge to diagnose and carries a high risk of misdiagnosis contributing to significant ocular morbidity. We aimed to employ and study the application of artificial intelligence-based deep learning (DL) algorithms to diagnose VK.
Analysis of spatial variation of nuclear morphology in tissue microenvironments
(IEEE, 2010) Raman, Sundaresan
We present a study of the spatial variation of nuclear morphology of stromal and cancer-associated fibroblasts in the mouse mammary gland. The work is part of a framework being developed for the analysis of the tumor microenvironment in breast cancer. Recent research has uncovered the role of stromal cells in promoting tumor growth and progression. In specific, studies have indicated that stromal fibroblasts - formerly considered to be passive entities in the extra-cellular matrix - play an active role in the progression of tumor in mammary tissue. We have focused on the analysis of the nuclear morphology of fibroblasts, which several studies have shown to be a critical phenotype in cancer. An essential component of our approach is that the nuclear morphology is studied within the 3D spatial context of the tissue, thus enabling us to pose questions about how the locus of a cell relates to its morphology, and possibly to its function. In order to make quantitative comparisons between nuclear populations, we build statistical shape models of cell populations and infer differences between the populations through these models. We present our observation on both normal and tumor tissues from the mouse mammary gland.
Analysis of Transfer and Residual Learning for Detecting Plant Diseases Using Images of Leaves
(Springer, 2018-09) Raman, Sundaresan
The study of plant diseases is critical for alleviating the problem of food security all over the world. The most critical step in mitigating this problem is the correct and appropriate timely identification of the disease. The first step in identification of a disease is visual inspection. The massive scale of this problem and lack of professionals create a need for a automated accurate visual inspection technique. Recent advances in the field of computer vision, primarily through techniques such as use of convolutional neural networks and deep learning have generated impressive results in the field of image classification and object recognition. In this paper, we address the problem of detecting plant diseases using images of leaves using different state-of-the-art approaches. We use the Plant Village dataset comprising of 86,198 images of 25 crops across 57 classes (healthy and specific diseases). The images are of high quality and have been taken manually under appropriate lighting conditions. On this dataset, our model is able to attain a significantly high average accuracy of 99.374% using transfer learning on state-of-the-art models trained on the ILSVRC 2012 dataset having 1.2 million images across 1000 classes
Analyzing Surface Defects in Apples Using Gabor Features
(IEEE, 2016) Raman, Sundaresan
This paper describes different approaches for detection and identification of diseases in apples using computer vision. Our proposed algorithms analyze surface appearance of apple for defects using image features, viz. color and texture. For segmentation of Region Of Interest (ROI), K-means clustering is performed over the image pixels based on their intensity values. For creation of feature vector, combinations of Gabor Wavelets with different feature descriptors were explored. Comparative study has been carried out between Haralick features, Local Binary Patterns, and kernel PCA, to observe their performance over Gabor features. Classification is achieved via Support Vector Machines and K-Nearest Neighbors. For the task of disease detection, accuracy recorded was greater than 96.9% for Gabor+LBP approach and in range of 89.8% to 96.25% for Gabor+Haralick approach. Gabor+kernel PCA recorded lowest accuracy of 90%. For disease identification, combination of Gabor+LBP outperformed other combinations, recording highest accuracy ranging from 85.93% to 95.31%.
Capturing variations in nuclear phenotypes
(Elsevier, 2019-09) Raman, Sundaresan
Relating genotypes with phenotypes is important to understand diseases like cancer, but extremely challenging, given the underlying biological variability and levels of phenotypes. 3D quantitative tools are increasingly used to provide robust inferences pertaining to variations across collections of cells. We especially focus on the changes wrought to the nucleus of specific genotypes. Fibroblasts in the tumor microenvironment of mammary epithelial tissue serve as our model system and provide the context, although our methods are applicable to a broader range of biological systems. Using an image based approach, we analyze in 3D and compare phenotypes at nuclear level using estimates of texture, morphology and spatial context based on confocal images. Our data demonstrates that deletion of TP53 in stromal fibroblasts results in reorganization of chromatin content across the nucleus, especially the nuclear periphery, while simultaneously reducing nuclear size and making it more spindly. No such shape change was observed for PTEN-deleted genotype, although there were some differences in distribution of chromatin and an increase in the local nuclear density. The relative changes in phenotypes are in line with the larger role that the TP53 plays in tumor initiation and progression.These findings play an important role in uncovering the relationships of those genes with the subcellular phenotypes, as well as formulating new hypotheses, especially pertaining to the relative impact of genes in specific pathways. More importantly, they demonstrate the efficacy of methodology of analyzing a large number of cellular phenotypes.
Comparison of Two Ultra-Widefield Cameras With High Image Resolution and Wider View for Identifying Diabetic Retinopathy Lesions
(Association for Research in Vision and Ophthalmology, 2021-10) Raman, Sundaresan
To compare the effectiveness of the Optos P200dTx and Zeiss Clarus 500 fundus cameras in detecting diabetic retinopathy (DR) lesions.
Comparison of various fractal analysis methods for retinal images
(Elsevier, 2021-01) Raman, Sundaresan
Retinal vessels are known to behave like a fractal, waherein a part of a geometrical pattern resembles the whole. Although the box counting method has been used most commonly, currently there exists no "best method" for fractal analysis on retinal vessels. In the present study we compared the different methods of fractal analysis of retinal images. This study included 43 normal retinal images from public databases (STARE & DRIVE) and 40 retinal images (20 normal and 20 diseased) collected from an epidemiological study database (Sankara Nethralaya diabetic retinopathy epidemiology and molecular genetics study; SNDREAMS). In our study we calculated and compared the values of fractal dimensions by Box counting method, Hausdorff Fractal Dimension (HFD), Modified Hausdorff Fractal Dimension (MHFD) and Fourier Fractal Dimension (FFD). The coefficient of variation(CV) was the least with HFD methods in different databases (DRIVE & STARE: −0.088, SNDREAMS Normal retinal images: −0.117, SNDREAMS Diseased retinal images: −0.103). Our study showed that HFD method was the best method to calculate the fractal dimensions of normal and diseased retinal images.
Deep3DSCan: Deep residual network and morphological descriptor based framework forlung cancer classification and 3D segmentation
(IET, 2020-04) Raman, Sundaresan; Chamola, Vinay; Narang, Pratik
With the increasing incidence rate of lung cancer patients, early diagnosis could help in reducing the mortality rate. However, accurate recognition of cancerous lesions is immensely challenging owing to factors such as low contrast variation, heterogeneity and visual similarity between benign and malignant nodules. Deep learning techniques have been very effective in performing natural image segmentation with robustness to previously unseen situations, reasonable scale invariance and the ability to detect even minute differences. However, they usually fail to learn domain-specific features due to the limited amount of available data and domain agnostic nature of these techniques. This work presents an ensemble framework Deep3DSCan for lung cancer segmentation and classification. The deep 3D segmentation network generates the 3D volume of interest from computed tomography scans of patients. The deep features and handcrafted descriptors are extracted using a fine-tuned residual network and morphological techniques, respectively. Finally, the fused features are used for cancer classification. The experiments were conducted on the publicly available LUNA16 dataset. For the segmentation, the authors achieved an accuracy of 0.927, significant improvement over the template matching technique, which had achieved an accuracy of 0.927. For the detection, previous state-of-the-art is 0.866, while ours is 0.883.
Early Blight Identification in Tomato Leaves using Deep Learning
(IEEE, 2020) Raman, Sundaresan
Tomatoes are one of the major horticulture crops in the world. Early Blight is one of the most widespread tomato diseases in India, often causing a significant reduction in produce. Agricultural produce of tomatoes is of utmost importance, making it necessary for timely recognition of Early Blight. Using self-collected images, we first explore classification of Early Blight in diseased leaves using ResNet and Xception networks, achieving a classification accuracy of 99.952%. However, significant focus has already been dedicated to disease classification in crops. Additionally, the lack of spatial information for affected leaves persuades us to move towards an object detection approach, utilizing variants based on the YOLO framework. We illustrate results with a twin focus on accuracy and real-time inference. Through our work, we aim to assist the development of a mobile application for disease identification.
Hybrid computational intelligence algorithms and their applications to detect food quality
(Springer, 2019-04) Raman, Sundaresan
Food security is a major problem faced today. With primitive storage facilities, especially in developing countries, it often leads to extensive losses. This work aims to develop algorithms based on vision data to assess the food quality and deploy them in food storage facilities to detect early signs of spoilage. This paper presents various segmentation techniques for finding spoilt food. Novel optimization techniques have been developed and implemented to improve K-means clustering and multilevel thresholding. A hybrid of moth flame optimization (MFO) and gravitational search algorithm (GSA) has been developed. Also, in another hybrid, particle swarm optimization (PSO) was also incorporated along with MFO and GSA. Both the hybrids performed better than the individual algorithms and the MFO–GSA–PSO hybrid performed better than the MFO–GSA hybrid on the benchmark functions. Segmented images using optimized K-means were used for feature extraction using local binary patterns (LBP). Multiclass support vector machine was used for classification which gave an accuracy of 81% for features from segmented images obtained using MFO–GSA hybrid and 83.33% for that using MFO–GSA–PSO hybrid. Results of simple linear iterative clustering superpixels for segmentation have also been discussed. The segmented clusters are then used to judge the rottenness of the food. Classification using LBP and Haralick features of the segmented image obtained using graphs over superpixels gave an accuracy of 81.7% and 78% respectively
An ImageJ macro tool for OCTA-based quantitative analysis of Myopic Choroidal neovascularization
(PLOS One, 2023-04) Raman, Sundaresan
Myopic Choroidal neovascularization (mCNV) is one of the most common vision-threatening com- plications of pathological myopia among many retinal diseases. Optical Coherence Tomography Angiography (OCTA) is an emerging newer non-invasive imaging technique and is recently being included in the investigation and treatment of mCNV. However, there exists no standard tool for time-efficient and dependable analysis of OCTA images of mCNV. In this study, we propose a customizable ImageJ macro that automates the OCTA image processing and lets users measure nine mCNV biomarkers. We developed a three-stage image processing pipeline to process the OCTA images using the macro. The images were first manually delineated, and then denoised using a Gaussian Filter. This was followed by the application of the Frangi filter and Local Adaptive thresholding. Finally, skeletonized images were obtained using the Mexican Hat filter. Nine vascular biomarkers including Junction Density, Vessel Diameter, and Fractal Dimension were then computed from the skeletonized images. The macro was tested on a 26 OCTA image dataset for all biomarkers. Two trends emerged in the computed biomarker values. First, the lesion-size dependent parameters (mCNV Area (mm2) Mean = 0.65, SD = 0.46) showed high variation, whereas normalized parameters (Junction Density(n/mm): Mean = 10.24, SD = 0.63) were uniform throughout the dataset. The computed values were consistent with manual measurements within existing literature. The results illustrate our ImageJ macro to be a convenient alternative for manual OCTA image processing, including provisions for batch processing and parameter customization, providing a systematic, reliable analysis of mCNV.
Layers for effective volume rendering
(ACM Digital Library, 2008) Raman, Sundaresan
A multi-layer volume rendering framework is presented. The final image is obtained by compositing a number of renderings, each being represented as a separate layer. This layer-centric framework provides a rich set of 2D operators and interactions, allowing both greater freedom and a more intuitive 2D-based user interaction. We extend the concept of compositing which is traditionally thought of as pertaining to the Porter and Duff compositing operators to a more general and flexible set of functions. In addition to developing new functional compositing operators, the user can control each individual layer's attributes, such as the opacity. They can also easily add or remove a layer from the composition set, change their order in the composition, and export and import the layers in a format readily utilized in a 2D paint package. This broad space of composition functions allows for a wide variety of effects and we present several in the context of volume rendering, including two-level volume rendering, masking, and magnification. We also discuss the integration of a 3D volume rendering engine with our 2.5D layer compositing engine.
A low power consumption mobile based IoT framework for real-time classification and segmentation for apple disease
(Elsevier, 2022-10) Raman, Sundaresan; Chamola, Vinay
Untreated diseases in plants not only lead to monetary losses but can have adverse implications when consumed. Disease diagnosis requires early detection and analysis of the disease. Apple horticulture has been a significant agriculture industry around the world and is affected by three most prominent domains of disease in apple namely: Blotch, Scab and Rot. In this paper, we provide a real-time mechanism for simultaneous classification and segmentation of the disease which significantly improves the speed of prediction. We have introduced atrous skip connections with UNet (with ResNet as backbone) furthering the performance. Experimental results on our proposed framework, achieves an accuracy of 94.29% to classify the disease and a dice score of 90.01% for segmentation of the diseased part. We also have developed a mobile application to demonstrate the objectives and to facilitate a user-friendly interface for using the proposed framework.
LWCNN: a lightweight convolutional neural network for agricultural crop protection
(ACM Digital Library, 2022-07) Raman, Sundaresan; Chamola, Vinay
Automatic identification of plant diseases is critical for agricultural crop protection so as to enhance the crop yield. The recent advances in deep learning and image processing gives hope for the development of efficient algorithms to address this issue. In this manuscript, we make use of these schemes to develop a Light-Weight Convolutional Neural Network (LWCNN) for identifying diseases in the leaves and ears of pearl millets. Although many models exist in the literature, the total number of parameters employed by our model is far fewer, by an order of thousand as compared to many other light-weight networks such as MobileNet(v2), EfficientNet, NASNet etc. Hence our scheme can be employed and run directly on devices with much lesser compute power. It is noteworthy that despite using few parameters, the proposed model achieves an accuracy of 97.4% in detecting the existence of the downy mildew disease in pearl millets, and takes the least time for both training and testing as compared to other models. To eliminate most of the pre-processing steps and to make our system suitable for on-field detection, we explore three single stage object detectors namely SSD, YOLOv3 and RetinaNet which localize and classify multiple instances of healthy and diseased leaves and ears in the image. We present a comparative analysis of the models and our experiments indicate that SSD is most suitable outperforming the other two models by a significant margin.
Machine Learning-Based Diagnosis and Ranking of Risk Factors for Diabetic Retinopathy in Population-Based Studies from South India
(MDPI, 2023-06) Raman, Sundaresan
This paper discusses the importance of investigating DR using machine learning and a computational method to rank DR risk factors by importance using different machine learning models. The dataset was collected from four large population-based studies conducted in India between 2001 and 2010 on the prevalence of DR and its risk factors. We deployed different machine learning models on the dataset to rank the importance of the variables (risk factors). The study uses a t-test and Shapely additive explanations (SHAP) to rank the risk factors. Then, it uses five machine learning models (K-Nearest Neighbor, Decision Tree, Support Vector Machines, Logistic Regression, and Naive Bayes) to identify the unimportant risk factors based on the area under the curve criterion to predict DR. To determine the overall significance of risk variables, a weighted average of each classifier’s importance is used. The ranking of risk variables is provided to machine learning models. To construct a model for DR prediction, the combination of risk factors with the highest AUC is chosen. The results show that the risk factors glycosylated hemoglobin and systolic blood pressure were present in the top three risk factors for DR in all five machine learning models when the t-test was used for ranking. Furthermore, the risk factors, namely, systolic blood pressure and history of hypertension, were present in the top five risk factors for DR in all the machine learning models when SHAP was used for ranking. Finally, when an ensemble of the five machine learning models was employed, independently with both the t-test and SHAP, systolic blood pressure and diabetes mellitus duration were present in the top four risk factors for diabetic retinopathy. Decision Tree and K-Nearest Neighbor resulted in the highest AUCs of 0.79 (t-test) and 0.77 (SHAP). Moreover, K-Nearest Neighbor predicted DR with 82.6% (t-test) and 78.3% (SHAP) accuracy.
The Need for Artificial Intelligence Based Risk Factor Analysis for Age-Related Macular Degeneration: A Review
(MDPI, 2022-10) Raman, Sundaresan
In epidemiology, a risk factor is a variable associated with increased disease risk. Understanding the role of risk factors is significant for developing a strategy to improve global health. There is strong evidence that risk factors like smoking, alcohol consumption, previous cataract surgery, age, high-density lipoprotein (HDL) cholesterol, BMI, female gender, and focal hyper-pigmentation are independently associated with age-related macular degeneration (AMD). Currently, in the literature, statistical techniques like logistic regression, multivariable logistic regression, etc., are being used to identify AMD risk factors by employing numerical/categorical data. However, artificial intelligence (AI) techniques have not been used so far in the literature for identifying risk factors for AMD. On the other hand, artificial intelligence (AI) based tools can anticipate when a person is at risk of developing chronic diseases like cancer, dementia, asthma, etc., in providing personalized care. AI-based techniques can employ numerical/categorical and/or image data thus resulting in multimodal data analysis, which provides the need for AI-based tools to be used for risk factor analysis in ophthalmology. This review summarizes the statistical techniques used to identify various risk factors and the higher benefits that AI techniques provide for AMD-related disease prediction. Additional studies are required to review different techniques for risk factor identification for other ophthalmic diseases like glaucoma, diabetic macular edema, retinopathy of prematurity, cataract, and diabetic retinopathy.
Quality Isosurface Mesh Generation Using an Extended Marching Cubes Lookup Table
(Wiley, 2008-09) Raman, Sundaresan
The Marching Cubes Algorithm may return degenerate, zero area isosurface triangles, and often returns isosurface triangles with small areas, edges or angles. We show how to avoid both problems using an extended Marching Cubes lookup table. As opposed to the conventional Marching Cubes lookup table, the extended lookup table differentiates scalar values equal to the isovalue from scalar values greater than the isovalue. The lookup table has 38= 6561 entries, based on three possible labels, ‘−’ or ‘=’ or ‘+’, of each cube vertex. We present an algorithm based on this lookup table which returns an isosurface close to the Marching Cubes isosurface, but without any degenerate triangles or any small areas, edges or angles.
Relationship of fractal analysis in retinal microvascularity with demographic and diagnostic parameters
(Elsevier, 2022-01) Raman, Sundaresan
Problems and diseases with eye are common in diabetic patients. Early diagnosis and detection of various diseases like retinopathy, neuropathy and nephropathy is crucial in diabetic patients. Certain demographic and diagnostic parameters play a significant role in predicting diseases related to diabetes. Development of a novel diagnostic method which helps to predict the disease by establishing a significant correlation with the demographic and diagnostic parameters is of prime importance. This study proposes a new methodology in which retinal fractals are obtained for the images and the derived retinal fractals are analysed to aid in disease prediction. This study comprises of images from patients with retinopathy, non retinopathy, neuropathy, nephropathy and hypertension. The proposed research is carried out in two aspects: 1) to correlate the retinal fractals of retinopathy and non retinopathy images with certain demographic and diagnostic parameters and interpret its significance, and 2) to exhibit a relationship between the retinal fractals and various diseases/addictive habit to facilitate the prediction of the disease/addictive habit. Hausdorff fractal dimension (HFD) was applied and higher fractal dimension was obtained for healthy cases. Then using Statistical Package for the Social Sciences (SPSS) various statistical parameters and significance were calculated to analyse the relationship. Analysis results showed that fractal value helped in distinguishing between the retinopathy and non retinopathy conditions. It also helped in diagnosing the presence and absence of hypertension. Correlation analysis between certain demographic parameters and fractal value showed a positive correlation whereas few exhibited negative correlation.
Relationship of fractal analysis in retinal microvascularity with demographic and diagnostic parameters
(Elsevier, 2022-01) Raman, Sundaresan
Problems and diseases with eye are common in diabetic patients. Early diagnosis and detection of various diseases like retinopathy, neuropathy and nephropathy is crucial in diabetic patients. Certain demographic and diagnostic parameters play a significant role in predicting diseases related to diabetes. Development of a novel diagnostic method which helps to predict the disease by establishing a significant correlation with the demographic and diagnostic parameters is of prime importance. This study proposes a new methodology in which retinal fractals are obtained for the images and the derived retinal fractals are analysed to aid in disease prediction. This study comprises of images from patients with retinopathy, non retinopathy, neuropathy, nephropathy and hypertension. The proposed research is carried out in two aspects: 1) to correlate the retinal fractals of retinopathy and non retinopathy images with certain demographic and diagnostic parameters and interpret its significance, and 2) to exhibit a relationship between the retinal fractals and various diseases/addictive habit to facilitate the prediction of the disease/addictive habit. Hausdorff fractal dimension (HFD) was applied and higher fractal dimension was obtained for healthy cases. Then using Statistical Package for the Social Sciences (SPSS) various statistical parameters and significance were calculated to analyse the relationship. Analysis results showed that fractal value helped in distinguishing between the retinopathy and non retinopathy conditions. It also helped in diagnosing the presence and absence of hypertension. Correlation analysis between certain demographic parameters and fractal value showed a positive correlation whereas few exhibited negative correlation.