Automating Medical Image Analysis Workflows Using Machine Learning

Speaker: Mahendra Khened

Venue : HSB 209

Date : 30th August, 2019

Time : 5.30 p.m.

Abstract

Machine learning is currently one of the hot topics of research and most of the industries are turning their efforts towards automating most of their work-flow processes using machine learning. Medical images are acquired for the diagnosis of various diseases in the body. Most often the analysis of medical images involves detection and identifying the tissue structures as normal or abnormal and estimating certain clinical metrics. In a clinical setting, medical images are analyzed manually by trained medical professionals. Some of these interpretation tasks are time-consuming and are sometimes prone to error due to various reasons ranging from lack of availability of trained medical experts and human fatigue. In order to address some of these issues, our research focuses on developing pipelines for certain clinical work-flow applications acting as an auxiliary aid for the medical professionals by automatically analyzing and giving a preliminary diagnosis report. We explore some of the popular machine learning techniques like Convolution Neural Networks for our research work and showcase some of our clinical applications in Cardiac Image analysis field.

About the speaker

Mahendra Khened completed his B.E in Electronics and Communication Engineering from BMS College of Engineering, Bangalore. Currently, he is a research scholar working with Dr.Ganapathy Krishnamurthy at Medical Imaging and Reconstruction Laboratory, Department of Engineering Design, IIT Madras. His research focuses on medical image analysis problems using computer vision, machine learning, and deep neural networks. He has worked on various medical image segmentation problems, texture analysis for disease classification and quantification of response to treatment in clinical practice.