You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This repository presents a dual-disease diagnostic system using Gradient Boosting Machines (XGBoost & LightGBM) on clinical blood parameters. The system demonstrates 100% accuracy for both disease predictions by employing target realignment, class imbalance correction, and feature importance analysis. The pipeline is intended as a decision-support.
Offline-first PWA for ASHA/ANM workers: explainable diabetes risk, photo-based anemia screening, and a safety guardrail — with a village outbreak radar. Works with no network, in 10 Indian languages. Screening aid, not a diagnosis.
Notebooks used for my Minor Project for palm-image anemia detection with ConvNeXt baseline and ensemble experiments, which was then later used in the Major Project
Revolutionary AI-powered anemia detection platform combining browser-based ML (98.75% accuracy) with Gemini AI. Features intelligent temporal analysis comparing past vs current blood reports to track health progression, enabling early intervention for 3 billion at-risk individuals globally.
Medical image classification system on the AneRBC dataset using custom CNNs, transfer learning (MobileNetV2, ResNet18, DenseNet121), and Explainable AI (Grad-CAM) for anemia diagnosis.