AI/ML Projects
Detection of Myocardial Infarction in 2D Echocardiograms and Cardiac Magnetic Resonance Images Using Deep Learning
This project detects Myocardial Infarction (heart attack) from echocardiograms (echo) or magnetic resonance imaging (MRI). The method contains three main parts:
1. Image Segmentation using Attention U-Net
2. Quantification of Wall Motion and Myocardial Thickening
3. Binary Classification using Modified Siamese Neural Network
The project combines image segmentation, feature extraction from computer vision, and binary classification to detect heart attacks from echo or MRI data. It's a comprehensive approach that leverages both structural information (from the masks) and functional information (from wall motion and myocardial thickening) to make predictions.
​
The GitHub code is available here.
English Subtitle Generator
This project uses OpenAI's Whisper Model, a Large Language Model, to generate English subtitles (SRT file) for any video. You may view the GitHub code here.
To Do App
This is a simple application that manages your To Do List using React + Vite + Express + SQLite3. This code is mostly generated by an AI coding assistant and demonstrates my ability to leverage LLMs for my coding work. You may view the GitHub code here.
Workflow Automation
This project focuses on workflow automation by leveraging conversational AI through no-code/low-code platforms like n8n. By integrating Large Language Models (LLMs) and AI Agents into workflows, a seamless, efficient, and intelligent automation process is created. You may see a sample work here.
