Research Interests:
Deep neural networks suffer from Catastrophic Forgetting (CF) on old tasks when they are trained to learn new tasks sequentially, since the parameters of the model will change to optimize on the new class. The problem of alleviating CF is... more
Deep neural networks suffer from Catastrophic Forgetting (CF) on old tasks when they are trained to learn new tasks sequentially, since the parameters of the model will change to optimize on the new class. The problem of alleviating CF is of interest to Computer aided diagnostic (CAD) systems community to facilitate class incremental learning (IL): learn new classes as and when new data/annotations are made available and old data is no longer accessible. However, IL has not been explored much in CAD development. We propose a novel approach that ensures that a model remembers the causal factor behind the decisions on the old classes, while incrementally learning new classes. We introduce a common auxiliary task during the course of incremental training, whose hidden representations are shared across all the classification heads. Since the hidden representation is no longer task-specific, it leads to a significant reduction in CF. We demonstrate our approach by incrementally learning 5 different tasks on Chest-Xrays and compare the results with the state-of-the-art regularization methods. Our approach performs consistently well in reducing CF in all the tasks with almost zero CF in most of the cases unlike standard regularisation-based approaches.
Research Interests:
Segmentation and analysis of sub-cortical structures is of interest in diagnosing some neurological diseases. Segmentation is a challenging task because of brain tissue ambiguity and data scarcity. Deep learning (DL) solutions are widely... more
Segmentation and analysis of sub-cortical structures is of interest in diagnosing some neurological diseases. Segmentation is a challenging task because of brain tissue ambiguity and data scarcity. Deep learning (DL) solutions are widely used for this purpose by considering the problem as a semantic segmentation of the brain. In general, DL approaches exhibit a bias towards larger structures when training is done on the whole brain. We propose a method to address this problem wherein a pre-training step is used to learn tissue characteristics and a rough ROI extraction step aids focusing on local context. We use a Residual U-net for demonstrating the proposed method. Experiments on the IBSR and MICCAI datasets show that our proposed solution leads to an improvement in segmentation performance in general with medium and small size structures benefiting significantly. The performance with the proposed method is also marginally better than a more complex, state of art sub-cortical structure segmentation method. A strength of the proposed method is that it can also be applied as a modification to any existing segmentation solution.
Research Interests:
Deep learning based methods have shown great promise in achieving accurate automatic detection of Coronavirus Disease (covid) - 19 from Chest X-Ray (cxr) images.However, incorporating explainability in these solutions remains relatively... more
Deep learning based methods have shown great promise in achieving accurate automatic detection of Coronavirus Disease (covid) - 19 from Chest X-Ray (cxr) images.However, incorporating explainability in these solutions remains relatively less explored. We present a hierarchical classification approach for separating normal, non-covid pneumonia (ncp) and covid cases using cxr images. We demonstrate that the proposed method achieves clinically consistent explainations. We achieve this using a novel multi-scale attention architecture called Multi-scale Attention Residual Learning (marl) and a new loss function based on conicity for training the proposed architecture. The proposed classification strategy has two stages. The first stage uses a model derived from DenseNet to separate pneumonia cases from normal cases while the second stage uses the marl architecture to discriminate between covid and ncp cases. With a five-fold cross validation the proposed method achieves 93%, 96.28%, and 84.51% accuracy respectively over three large, public datasets for normal vs. ncp vs. covid classification. This is competitive to the state-of-the-art methods. We also provide explanations in the form of GradCAM attributions, which are well aligned with expert annotations. The attributions are also seen to clearly indicate that marl deems the peripheral regions of the lungs to be more important in the case of covid cases while central regions are seen as more important in ncp cases. This observation matches the criteria described by radiologists in clinical literature, thereby attesting to the utility of the derived explanations.
Research Interests:
Disease diagnosis often requires segmentation of structures from a given image followed by shape analysis. Shape analysis entails quantifying the variability in a shape by constructing a template for a given population. We propose an... more
Disease diagnosis often requires segmentation of structures from a given image followed by shape analysis. Shape analysis entails quantifying the variability in a shape by constructing a template for a given population. We propose an orientation-invariant representation using varifolds for the shape elements in a given shape population and present a novel diffeomorphic Log-demons based template creation pipeline. The proposed method generates a good quality template at a significantly less computation time compared to state of the art method.
Research Interests:
Optic nerve head (ONH) segmentation problem is of interest for automated glaucoma assessment. Although various segmentation methods have been proposed in the recent past, it is difficult to evaluate and compare the performance of... more
Optic nerve head (ONH) segmentation problem is of interest for automated glaucoma assessment. Although various segmentation methods have been proposed in the recent past, it is difficult to evaluate and compare the performance of individual methods due to a lack of a benchmark dataset. The assessment involves segmentation of optic disk and cup region within the ONH. In this paper, we present a comprehensive dataset of retinal images of both normal and glaucomatous eyes with manual segmentations from multiple human experts. The dataset also provides expert opinion on an image representing a normal or glaucomatous eye and on the presence of notching in an image. Several state of the art methods are assessed against this dataset using cup to disc diameter ratio (CDR), area and boundary-based evaluation measures. These are presented to aid benchmarking of new methods. A supervised, notch detection method based on the segmentation results is also proposed and its assessment results are i...
Research Interests:
Anatomical variabilities seen in longitudinal data or inter-subject data is usually described by the underlying deformation, captured by non-rigid registration of these images. Stationary Velocity Field (SVF) based non-rigid registration... more
Anatomical variabilities seen in longitudinal data or inter-subject data is usually described by the underlying deformation, captured by non-rigid registration of these images. Stationary Velocity Field (SVF) based non-rigid registration algorithms are widely used for registration. SVF based methods form a metric-free framework which captures a finite dimensional submanifold of deformations embedded in the infinite dimensional smooth manifold of diffeomorphisms. However, these methods cover only a limited degree of deformations. In this paper, we address this limitation and define an approximate metric space for the manifold of diffeomorphisms $\mathcal{G}$. We propose a method to break down the large deformation into finite compositions of small deformations. This results in a broken geodesic path on $\mathcal{G}$ and its length now forms an approximate registration metric. We illustrate the method using a simple, intensity-based, log-demon implementation. Validation results of the...
Research Interests:
Context: A brain magnetic resonanace imaging (MRI) atlas plays an important role in many neuroimage analysis tasks as it provides an atlas with a standard coordinate system which is needed for spatial normalization of a brain MRI.... more
Context: A brain magnetic resonanace imaging (MRI) atlas plays an important role in many neuroimage analysis tasks as it provides an atlas with a standard coordinate system which is needed for spatial normalization of a brain MRI. Ideally, this atlas should be as near to the average brain of the population being studied as possible. Aims: The aim of this study is to construct and validate the Indian brain MRI atlas of young Indian population and the corresponding structure probability maps. Settings and Design: This was a population-specific atlas generation and validation process. Materials and Methods: 100 young healthy adults (M/F = 50/50), aged 21–30 years, were recruited for the study. Three different 1.5-T scanners were used for image acquisition. The atlas and structure maps were created using nonrigid groupwise registration and label-transfer techniques. Comparison and Validation: The generated atlas was compared against other atlases to study the population-specific trends....
Research Interests:
Melanoma is a leading cause of deaths due to skin cancer deaths and hence, early and effective diagnosis of melanoma is of interest. Current approaches for automated diagnosis of melanoma either use pattern recognition or analytical... more
Melanoma is a leading cause of deaths due to skin cancer deaths and hence, early and effective diagnosis of melanoma is of interest. Current approaches for automated diagnosis of melanoma either use pattern recognition or analytical recognition like ABCDE (asymmetry, border, color, diameter and evolving) criterion. In practice however, a differential approach wherein outliers (ugly duckling) are detected and used to evaluate nevi/lesions. Incorporation of differential recognition in Computer Aided Diagnosis (CAD) systems has not been explored but can be beneficial as it can provide a clinical justification for the derived decision. We present a method for identifying and quantifying ugly ducklings by performing Intra-Patient Comparative Analysis (IPCA) of neighboring nevi. This is then incorporated in a CAD system design for melanoma detection. This design ensures flexibility to handle cases where IPCA is not possible. Our experiments on a public dataset show that the outlier inform...
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Person following with a mobile robot using a modified optical flow by
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Registration of multimodal retinal images such as fundus and Optical Coherence Tomography (OCT) images is important as the two structural imaging modalities provide complementary views of the retina. This enables a more accurate... more
Registration of multimodal retinal images such as fundus and Optical Coherence Tomography (OCT) images is important as the two structural imaging modalities provide complementary views of the retina. This enables a more accurate assessment of the health of the retina. However, registration is a challenging task because fundus image (2D) is obtained via optical projection whereas the OCT image (3D) is derived via optical coherence and is very noisy. Furthermore, the field of view of imaging possible in the two modalities is very different resulting in low overlap (5–20%) between the obtained images. Existing methods for this task rely on either key-point (junction/corner) detection or accurate segmentation of vessels which is difficult due to noise. We propose a registration algorithm for finding efficient landmarks under noisy conditions. The method requires neither accurate structure segmentation nor key-point detection. The Modality Independent Neighborhood Descriptor (MIND) featu...
Research Interests:
Normative aging trends of the brain can serve as an important reference in the assessment of neurological structural disorders. Such models are typically developed from longitudinal brain image data – follow-up data of the same subject... more
Normative aging trends of the brain can serve as an important reference in the assessment of neurological structural disorders. Such models are typically developed from longitudinal brain image data – follow-up data of the same subject over different time points. In practice, obtaining such longitudinal data is difficult. We propose a method to develop an aging model for a given population, in the absence of longitudinal data, by using images from different subjects at different time points, the so-called cross-sectional data. We define an aging model as a diffeomorphic deformation on a structural template derived from the data and propose a method that develops topology preserving aging model close to natural aging. The proposed model is successfully validated on two public cross-sectional datasets which provide templates constructed from different sets of subjects at different age points.
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Segmentation of sub-cortical structures from MRI scans is of interest in many neurological diagnosis. Since this is a laborious task machine learning and specifically deep learning (DL) methods have become explored. The structural... more
Segmentation of sub-cortical structures from MRI scans is of interest in many neurological diagnosis. Since this is a laborious task machine learning and specifically deep learning (DL) methods have become explored. The structural complexity of the brain demands a large, high quality segmentation dataset to develop good DL-based solutions for sub-cortical structure segmentation. Towards this, we are releasing a set of 114, 1.5 Tesla, T1 MRI scans with manual delineations for 14 sub-cortical structures. The scans in the dataset were acquired from healthy young (21-30 years) subjects ( 58 male and 56 female) and all the structures are manually delineated by experienced radiology experts. Segmentation experiments have been conducted with this dataset and results demonstrate that accurate results can be obtained with deep-learning methods. Our sub-cortical structure segmentation dataset, Indian Brain Segmentation Dataset (IBSD) is made openly available at <https://doi.org/10.5281/zen...
Objective: Diabetic retinopathy is the leading cause of blindness in urban populations. Early diagnosis through regular screening and timely treatment has been shown to prevent visual loss and blindness. It is very difficult to cater to... more
Objective: Diabetic retinopathy is the leading cause of blindness in urban populations. Early diagnosis through regular screening and timely treatment has been shown to prevent visual loss and blindness. It is very difficult to cater to this vast set of diabetes patients, primarily because of high costs in reaching out to patients and a scarcity of skilled personnel. Telescreening offers a cost-effective solution to reach out to patients but is still inadequate due to an insufficient number of experts who serve the diabetes population. Developments toward fundus image analysis have shown promise in addressing the scarcity of skilled personnel for large-scale screening. This article aims at addressing the underlying issues in traditional telescreening to develop a solution that leverages the developments carried out in fundus image analysis. Method: We propose a novel Web-based telescreening solution (called DrishtiCare) integrating various value-added fundus image analysis component...
Research Interests:
The aim of a local descriptor or a feature descriptor is to efficiently represent the region detected by an interest point operator in a compact format for use in various applications related to matching. The common design principle... more
The aim of a local descriptor or a feature descriptor is to efficiently represent the region detected by an interest point operator in a compact format for use in various applications related to matching. The common design principle behind most of the mainstream descriptors like SIFT, GLOH, Shape context etc is to capture the spatial distribution of features using histograms computed over a grid around interest points. Histograms provide compact representation but typically loose the spatial distribution information. In this paper, we propose to use projection-based representation to improve a descriptor’s capacity to capture spatial distribution information while retaining the invariance required. Based on this proposal, two descriptors based on the CS-LBP are introduced. The descriptors have been evaluated against known descriptors on a standard dataset and found to outperform, in most cases, the existing descriptors. The obtained results demonstrate that proposed approach has the...
Abstract—We propose a method for synthetic zooming of tomographic images by applying super resolution technique on reconstructed data via a union of rotated lattices (URL). The proposed method consists of two steps: (i) sinogram data is... more
Abstract—We propose a method for synthetic zooming of tomographic images by applying super resolution technique on reconstructed data via a union of rotated lattices (URL). The proposed method consists of two steps: (i) sinogram data is filtered and backprojected on to two lattices, which are rotated versions of each other and (ii) the samples from the two lattices are interpolated to generate the upsampled image. Square and hexagonal lattices have been investigated for URL. Results of subjective and objective evaluations of the proposed method on analytic phantoms are presented and compared with direct upsampling of data reconstructed on a single square lattice and upsampled image generated by union of low resolution shifted images (USL). The proposed method shows qualitative and quantitative improvement over direct upsampling but when compared with USL, generated upsampled images are of comparable quality. Index Terms—Super-resolution, Tomographic images, Combination of rotated la...
A major challenge that deep learning systems face is the Catastrophic Forgetting (CF) phenomenon that is observed when fine-tuning is used to try and adapt a system to a new task or a sequence of datasets with different distributions. CF... more
A major challenge that deep learning systems face is the Catastrophic Forgetting (CF) phenomenon that is observed when fine-tuning is used to try and adapt a system to a new task or a sequence of datasets with different distributions. CF refers to the significant degradation in performance on the old task/dataset. In this paper, a novel approach is proposed to address CF in computer aided diagnosis (CAD) system design in the medical domain. CAD systems often need to handle a sequence of datasets collected over time from different sites with different imaging parameters/populations. The solution we propose is to move samples from all the datasets closer to a common manifold via a reformer at the front end of a CAD system. The utility of this approach is demonstrated on two common tasks, namely segmentation and classification, using publicly available datasets. Results of extensive experiments show that manifold learning can yield about 74% improvement on an average in the reduction of CF over the baseline fine-tuning process and the state-of-the-art regularization based methods. The results also indicate that a Reformer when used in conjunction with the state-of-the-art regularization methods, has the potential to yield further improvement in CF reduction.
Research Interests:
Research Interests: Computer Science and IEEE
The most widely followed procedure for diagnosis and prognosis of dementia is structural neuroimaging of hippocampus by means of MR. Hippocampus segmentation is of wide interest as it enables quantitative assessment of the structure. In... more
The most widely followed procedure for diagnosis and prognosis of dementia is structural neuroimaging of hippocampus by means of MR. Hippocampus segmentation is of wide interest as it enables quantitative assessment of the structure. In this paper, we propose an algorithm for hippocampus segmentation that is unsupervised and image driven. It is based on a hybrid approach which combines a coarse segmentation and surface evolution. A coarse solution is derived using region growing which is further refined using a modified version of the physics based water flow model (Liu and Nixon, 2007). The proposed method has been tested on a publicly available dataset. The performance of this method is assessed using Dice coefficient against the ground truth provided for 25 volume images. It is consistent across volumes and the average Dice values are comparable to a multi-atlas based method reported on a subset of the same dataset.
Research Interests:
Intra-retinal layer segmentation of Optical Coherence Tomography images is critical in the assessment of ocular diseases. Existing Energy minimization based methods employ handcrafted cost terms to define their energy and are not robust... more
Intra-retinal layer segmentation of Optical Coherence Tomography images is critical in the assessment of ocular diseases. Existing Energy minimization based methods employ handcrafted cost terms to define their energy and are not robust to the presence of abnormalities. We propose a novel, Linearly Parameterized, Conditional Random Field (LP-CRF) model whose energy is learnt from a set of training images in an end-to-end manner. The proposed LP-CRF comprises two convolution filter banks to capture the appearance of each tissue region and boundary, the relative weights of the shape priors and an additional term based on the appearance similarity of the adjacent boundary points. All the energy terms are jointly learnt using the Structured Support Vector Machine. The proposed method segments all retinal boundaries in a single step. Our method was evaluated on 107 Normal and 220 AMD B-scan images and found to outperform three publicly available OCT segmentation software. The average uns...
Research Interests:
Watermarking is being used in a wide variety of applications. Steganography, copyright protection, owner identification etc are some of them. But watermarking can also be used as means to store other kind of useful information in the... more
Watermarking is being used in a wide variety of applications. Steganography, copyright protection, owner identification etc are some of them. But watermarking can also be used as means to store other kind of useful information in the image. This work discusses the advantages of putting such information into the image. A watermarking algorithm suitable for embedding large amount of information in the image, robust of jpeg compression is also presented.
Automated Computer Aided diagnostic tools can be used for the early detection of glaucoma to prevent irreversible vision loss. In this work, we present a Multi-task Convolutional Neural Network (CNN) that jointly segments the Optic Disc... more
Automated Computer Aided diagnostic tools can be used for the early detection of glaucoma to prevent irreversible vision loss. In this work, we present a Multi-task Convolutional Neural Network (CNN) that jointly segments the Optic Disc (OD), Optic Cup (OC) and predicts the presence of glaucoma in color fundus images. The CNN utilizes a combination of image appearance features and structural features obtained from the OD-OC segmentation to obtain a robust prediction. The use of fewer network parameters and the sharing of the CNN features for multiple related tasks ensures the good generalizability of the architecture, allowing it to be trained on small training sets. The cross-testing performance of the proposed method on an independent validation set acquired using a different camera and image resolution was found to be good with an average dice score of 0.92 for OD, 0.84 for OC and AUC of 0.95 on the task of glaucoma classification illustrating its potential as a mass screening to...
Research Interests:
Pneumoconiosis is an occupational lung disease caused by the inhalation of industrial dust. Despite the increasing safety measures and better work place environments, pneumoconiosis is deemed to be the most common occupational disease in... more
Pneumoconiosis is an occupational lung disease caused by the inhalation of industrial dust. Despite the increasing safety measures and better work place environments, pneumoconiosis is deemed to be the most common occupational disease in the developing countries like India and China. Screening and assessment of this disease is done through radiological observation of chest x-rays. Several studies have shown the significant inter and intra reader observer variation in the diagnosis of this disease, showing the complexity of the task and importance of the expertise in diagnosis. The present study is aimed at understanding the perceptual and cognitive factors1 affecting the reading of chest x-rays of pneumoconiosis patients. Understanding these factors helps in developing better image acquisition systems, better training regimen for radiologists and development of better computer aided diagnostic (CAD) systems. We used an eye tracking experiment to study the various factors affecting t...
Saliency computation is widely studied in computer vision but not in medical imaging. Existing computational saliency models have been developed for general (natural) images and hence may not be suitable for medical images. This is due to... more
Saliency computation is widely studied in computer vision but not in medical imaging. Existing computational saliency models have been developed for general (natural) images and hence may not be suitable for medical images. This is due to the variety of imaging modalities and the requirement of the models to capture not only normal but also deviations from normal anatomy. We present a biologically inspired model for colour fundus images and illustrate it for the case of diabetic retinopathy. The proposed model uses spatially-varying morphological operations to enhance lesions locally and combines an ensemble of results, of such operations, to generate the saliency map. The model is validated against an average Human Gaze map of 15 experts and found to have 10% higher recall (at 100% precision) than four leading saliency models proposed for natural images. The F-score for match with manual lesion markings by 5 experts was 0.4 (as opposed to 0.532 for gaze map) for our model and very ...
Research Interests: Computer Science and Acm
Cardiac motion analysis from B-mode ultrasound sequence is a key task in assessing the health of the heart. The paper proposes a new methodology for cardiac motion analysis based on the temporal behaviour of points of interest on the... more
Cardiac motion analysis from B-mode ultrasound sequence is a key task in assessing the health of the heart. The paper proposes a new methodology for cardiac motion analysis based on the temporal behaviour of points of interest on the myocardium. We define a new signal called the Temporal Flow Graph (TFG) which depicts the movement of a point of interest over time. It is a graphical representation derived from a flow field and describes the temporal evolution of a point. We prove that TFG for an object undergoing periodic motion is also periodic. This principle can be utilized to derive both global and local information from a given sequence. We demonstrate this for detecting motion irregularities at the sequence, as well as regional levels on real and synthetic data. A coarse localisation of anatomical landmarks such as centres of left/right cavities and valve points is also demonstrated using TFGs.
The crosscutting project described here based on the revolutionary concept of Probabilistic design [17] and of Ethnomathematics [13] aims to improve literacy in developing countries and regions, with an emphasis on India as the first test... more
The crosscutting project described here based on the revolutionary concept of Probabilistic design [17] and of Ethnomathematics [13] aims to improve literacy in developing countries and regions, with an emphasis on India as the first test focussing on grades 1-5. Driving the I-Slate design are the pioneering principles trading off probability for energy, while educational pedagogy and practice drive the software design. Energy and concomitant environmental sustainability are the overarching themes that encompass all aspects of this effort. Keywords— Ethnomathematics, education, low power, probabilistic design.
Research Interests:
Rationale: Pneumoconiosis 1 , a lung disease caused by the inhalation of dust is mainly diagnosed using chest radiographs. It has been shown 12 that expertise and contralateral symmetry (CS) of the chest radiographs plays a significant... more
Rationale: Pneumoconiosis 1 , a lung disease caused by the inhalation of dust is mainly diagnosed using chest radiographs. It has been shown 12 that expertise and contralateral symmetry (CS) of the chest radiographs plays a significant role in Pneumoconiosis diagnosis. Here, we present a gaze tracking study aimed at understanding how the CS information and the expertise effect the eye movements of observers. Methods: Experimental subjects consisting of novices, medical students, residents and staff radiologists were presented with 17 double and 16 single lung images, and were asked to give profusion ratings for each lung zone. Eye movements and the time for their diagnosis were also recorded. Results: Gaze tracking analysis showed that doctors [residents & staff] move eyes more quickly (MannWhitney test: U = 20, p = .01) and over more distances (U = 24, p = .022), when compared to that of non-doctors [others]. Wilcoxon signed rank test (Z = 4.19, p < .001), revealed that the aver...
Optical Coherence Tomography (OCT) plays an important role in the analysis of retinal diseases such as Age-Related Macular Degeneration (AMD). In this paper, we present a method to construct a normative atlas for macula centric OCT... more
Optical Coherence Tomography (OCT) plays an important role in the analysis of retinal diseases such as Age-Related Macular Degeneration (AMD). In this paper, we present a method to construct a normative atlas for macula centric OCT volumes with a mean intensity template (MT) and probabilistic maps for the seven intra-retinal tissue layers. We also propose an AMD classification scheme where the deviation of the local similarity of a test volume with respect to the MT is used to characterize AMD. The probabilistic atlas was used for layer segmentation where we achieved an average dice score of 0.82 across the eight layer boundaries. On the AMD detection task, the classification accuracy and Area under the Receiver Operating Characteristic curve were 98\(\%\) and 0.996 respectively, on 170 OCT test volumes.