Brain Tumor Segmentation using 3D U-Net (Computer Vision Project) (2022)

Model to Predict the status of a genetic biomarker important for brain cancer treatment

Deployment repository for the VUMC Picture tool, responsible for orchestrating the installation and setup of the PICTURE project's core components. This repository automates the deployment process for seamless integration across all related repositories within the PICTURE project.

Computer_vision_projects

Research on 3D brain tumor segmentation using U-Net, nnU-Net, and Swin UNETR.

A production-ready deep learning system that classifies brain MRI scans into 4 tumor categories using a hybrid architecture that fuses local CNN features with global Transformer attention — achieving 96.4% test accuracy with built-in input validation and explainability via Grad-CAM.

NeuroAssist and AI-powered brain tumor detection d clinical decision support system designed by training 3D Attention U-Net segmentation model for MRI analysis and RAG enabled LLM reasoning for structured WHO-aligned diagnosis.

This is a personal coding project that explores how various machine learning methods can be used to classify MRI images of the brain tumors. Source of MRI brain images: Kaggle (see link)

The CodeClause Internship Program is a hands-on opportunity for aspiring software developers to gain practical experience and enhance their skills in a professional environment. This GitHub repository serves as a central hub for the internship program, providing resources, code samples, and project assignments to guide interns .

This study focuses on four deep-learning models, which are Inception V3, MobileNet V2, ResNet152V2, and VGG19, aiming to enhance the accuracy of tumor Classification

Brain tumor MRI classification using EfficientNetB0 + ResNet50V2 with GradCAM++ and Spatial Consensus Confidence (SCC) trust metric. FastAPI + SQLite backend.