kidney-cancer

This repository contains the code for an AI-Based Universal Lesion Segmentation application based on my master's thesis. It includes a deep learning model for CT scan segmentation, a web interface built with Vite and React, and a Flask server for backend support.

End-to-end deep learning system for kidney CT scan classification (Normal/Cyst/Tumor/Stone) with EfficientNetB4, Grad-CAM explainability, FastAPI backend, Streamlit UI, and complete MLOps pipeline using MLflow, DVC, Docker, and GitHub Actions.

A curated list of authoritative datasets, resources, and top-conference papers on kidney cancer (RCC) — from a computer science perspective. Toward curing kidney cancer with computation.

Random Forest classifier for kidney cancer subtype discrimination (KIRC vs KIRP) using TCGA miRNA expression — includes preprocessing, DEA, ROC evaluation (R)

ClusterTrack software allowed for the first time comprehensive measurements of microtubule behavior, including during stages of depolymerization, for which there is no known molecular marker

ClinicalxCRAFT (app) — KIRC cohort console with a live LLM agent that answers plain-language questions by writing SQL over a bundled dataset. FastAPI + static frontend; no CRAFT/OAuth.

ClinicalxCRAFT (viewer) — clone-and-run static dashboard for exploring the TCGA-KIRC kidney-cancer cohort. Open the HTML, no setup required.