A5087624455- Radiomics and Machine Learning in Medical Imaging
- Brain Tumor Detection and Classification
- Advanced Neural Network Applications
- Medical Image Segmentation Techniques
- AI in cancer detection
- COVID-19 diagnosis using AI
- Domain Adaptation and Few-Shot Learning
- Artificial Intelligence in Healthcare and Education
- Medical Imaging Techniques and Applications
- Explainable Artificial Intelligence (XAI)
- Adversarial Robustness in Machine Learning
- Fetal and Pediatric Neurological Disorders
- Data Management and Algorithms
- Interconnection Networks and Systems
- Cell Image Analysis Techniques
- VLSI and FPGA Design Techniques
- Low-power high-performance VLSI design
- Topic Modeling
- Biomedical Text Mining and Ontologies
- Spam and Phishing Detection
- Machine Learning and Algorithms
- Health Systems, Economic Evaluations, Quality of Life
- Misinformation and Its Impacts
- Web Data Mining and Analysis
- Anomaly Detection Techniques and Applications
McGill University
1968-2023
Darshan Dental College and Hospital
2023
Mila - Quebec Artificial Intelligence Institute
2021-2022
International Institute of Information Technology, Hyderabad
2021
Intelligent Machines (Sweden)
2021
Delhi Technological University
2021
Jaypee Institute of Information Technology
2019-2020
Indian Institute of Technology Hyderabad
2016-2019
University of California, San Diego
2019
University of Southern California
2019
In this paper, we propose an end-to-end trainable Convolutional Neural Network (CNN) architecture called the M-net, for segmenting deep (human) brain structures from Magnetic Resonance Images (MRI). A novel scheme is used to learn combine and represent 3D context information of a given slice in 2D slice. Consequently, M-net utilizes only convolution though it operates on data, which makes memory efficient. The segmentation method evaluated two publicly available datasets compared against...
Automated segmentation of cortical and noncortical human brain structures has been hitherto approached using nonrigid registration followed by label fusion. We propose an alternative approach for this a convolutional neural network (CNN) which classifies voxel into one many structures. Four different kinds two-dimensional three-dimensional intensity patches are extracted each voxel, providing local global (context) information to the CNN. The proposed is evaluated on five publicly available...
Deep learning (DL) models have provided state-of-the-art performance in various medical imaging benchmarking challenges, including the Brain Tumor Segmentation (BraTS) challenges. However, task of focal pathology multi-compartment segmentation (e.g., tumor and lesion sub-regions) is particularly challenging, potential errors hinder translating DL into clinical workflows. Quantifying reliability model predictions form uncertainties could enable review most uncertain regions, thereby building...
Although deep networks have been shown to perform very well on a variety of medical imaging tasks, inference in the presence pathology presents several challenges common models. These impede integration learning models into real clinical workflows, where customary process cascading deterministic outputs from sequence image-based steps (e.g. registration, segmentation) generally leads an accumulation errors that impacts accuracy downstream tasks. In this paper, we propose by embedding...
A brain magnetic resonanace imaging (MRI) atlas plays an important role in many neuroimage analysis tasks as it provides with a standard coordinate system which is needed for spatial normalization of MRI. Ideally, this should be near to the average population being studied possible.The aim study construct and validate Indian MRI young corresponding structure probability maps.This was population-specific generation validation process.100 healthy adults (M/F = 50/50), aged 21-30 years, were...
The weblog is dynamic and its size growing exponentially with time in terms of navigation sessions. These stored sessions are used for Web Navigation Prediction (WNP). Each user had varied behavior on the web so their navigated With a variety large sessions, task prediction becoming challenging. There need an effective method to handle multiple labels predicting desired information. This paper analyses performance Deep Learning techniques like Multi-Layer Perceptron Long-Short Term Memory...
Although deep learning (DL) models have shown great success in many medical image analysis tasks, deployment of the resulting into real clinical contexts requires: (1) that they exhibit robustness and fairness across different sub-populations, (2) confidence DL model predictions be accurately expressed form uncertainties. Unfortunately, recent studies indeed significant biases demographic subgroups (e.g., race, sex, age) context analysis, indicating a lack models. several methods been...
Memory accounts for a considerable portion of the total power budget and area digital systems. Furthermore, it is typically performance bottleneck processing units. Therefore, critical to optimize memory with respect product power, area, delay (PAD). We propose hybrid cell assignment method based on multi-sized dual-Vth SRAM cells which improves PAD cost function by 34% compared conventional assignment. also utilize sizing minimizing data retention voltage, voltages read write operations in...
Recently, Deep Neural Networks (DNNs) have made unprecedented progress in various tasks. However, there is a timely need to accelerate the training process DNNs specifically for real-time applications that demand high performance, energy efficiency and compactness. Numerous algorithms been proposed improve accuracy, however network computationally slow. In this paper, we present scalable pipelined hardware architecture with distributed memories digital neuron implement deep neural networks....
In this paper, we develop a metric designed to assess and rank uncertainty measures for the task of brain tumour sub-tissue segmentation in BraTS 2019 sub-challenge on quantification. The is to: (1) reward where high confidence assigned correct assertions, incorrect assertions are low (2) penalize that have higher percentages under-confident assertions. Here, workings components explored based number popular evaluated dataset.
Generalization is an important attribute of machine learning models, particularly for those that are to be deployed in a medical context, where unreliable predictions can have real world consequences. While the failure models generalize across datasets typically attributed mismatch data distributions, performance gaps often consequence biases "ground-truth" label annotations. This context image segmentation pathological structures (e.g. lesions), annotation process much more subjective, and...
Segmentation of enhancing tumours or lesions from MRI is important for detecting new disease activity in many clinical contexts. However, accurate segmentation requires the inclusion medical images (e.g., T1 post contrast MRI) acquired after injecting patients with a agent Gadolinium), process no longer thought to be safe. Although number modality-agnostic networks have been developed over past few years, they met limited success context pathology segmentation. In this work, we present...
Modeling user(s) navigation sequences and predicting their preferences has been an interesting area of research. For Web Navigation Prediction (WNP) the Markov model(s) are predominantly used for analyzing discovering user patterns. One major issues with model is that it fails to predict unclassified navigations. Presence such navigations reduces prediction power model. Deep machine learning models can be address but ability deteriorates if training sessions less in number. As Navigations...
To create standards-based secure access to student's and employee's personal data, attendance records, mark sheets, expenditure library management by using RFID tags Web Service with the help of hardware kit, which synchronizes one another.This system uses ASP.net provide software interface support standard Electronic Records every individual in campus premises.Students employees at can their books, payments (canteen, bills, other fine) a single, all one, card.Because is built on services it...