Muhammad Ali

Academic

Career

Verbatim text

Group Members: Abdullah Shakeel (i19 - 1717) Ahmed Wadood (i19 - 1858) Muhammad Ali (i19 - 1882) Technology Used: Python, Tensorflow, Keras, OpenCV Flask, Visual Studio Supervisor Name: Mr. Shoaib Saleem Dr. Uzair Iqbal Cardiac Diagnostic Framework In this R&D project, a deep learning based Cardiac Diagnostic framework is being developed which would assist healthcare professionals. Main goal of the project is automated diagnosis of an individual's cardiac strength using cardiac MRI. There are two modules of this app - first module uses image segmentation model UNET to detect the region of interest i.e left ventricle from the cardiac MRI of patient - second module calculates blood pumped out of the heart within each heartbeat i.e ejection fraction. An ejection fraction below 40 percent means your heart isn't pumping enough blood and may be failing. A low ejection fraction number can be an indicator of heart failure. This manual procedure when performed by a cardiologist takes about 20 minutes, our goal is to automate this process and assist cardiologists. The research part of the project successfully covers the experimental gaps of previous state of the art work done in this field. Features of web application include: - MRI image preprocessing using opencv and automated segmentation of our region of interest. - Automated calculation of ejection fraction volume and suggestion of suitable treatments. A research paper titled “A UNET based Cardiac Diagnostic Framework for Automatic Left Ventricle Segmentation and Ejection Fraction Estimation” will be submitted to medical conferences.

AI enrichment