This repository houses machine learning models and pipelines for predicting various diseases, coupled with an integration with a Large Language Model for Diet and Food Recommendation. Each disease prediction task has its dedicated directory structure to maintain organization and modularity.

[ECCV 2024] Official Implementation of 《WSI-VQA: Interpreting Whole Slide Image by Generative Question Answering》

This is a repo for the Tanzania AI lab hackathon 2020 & the AI4Dev2020 challenge, where we as the Elixir team created the 1st AI based cancer diagnosis system, built a model comprising of Deep Convolutional Neural Network(CNN) and a web app that screens microscopic images so as to detect cancer tumors, thus increasing speed, accuracy in cancer diagnosis, and testing

Heart and Lung Sounds Dataset Recorded from a Clinical Manikin using Digital Stethoscope (HLS-CMDS)

Parkinson detection based on wave sketches and deep learning - PDS

PHD Research Focus

A comprehensive machine learning application that predicts breast cancer malignancy using cytology measurements. Features an interactive Streamlit web interface with real-time visualizations including radar charts for cell nuclei analysis. Implements logistic regression with data preprocessing pipelines for accurate benign/malignant classification.

This repository includes all notes of AI in Healthcare Specialization offered by the Standford University.

Secure, High-performance DICOM anonymization and metadata extraction for research and healthcare.

Rule-based healthcare expert system designed using Pyke and Python. The project focuses on heart failure telemonitoring, aiming to enhance patient self-care and clinical management.

SG Healthcare AI Datathon 2021 - acute kidney injury (AKI) patients requiring replacement renal therapy

RadioCare: Fighting Inefficiencies in Medical Imaging

An AI project that uses Differentiable Architecture Search (DARTS) to automatically design an optimized CNN for cervical cancer cell classification using the SIPaKMeD dataset. Compares the NAS-discovered model against a ResNet baseline across accuracy, F1-score, model size, inference time, and visualizations like confusion matrices and ROC curves.

Medical engineer & software developer. Building practical medical tools, clinical decision support systems, and AI-powered services for real-world healthcare workflows. Focus areas: emergency medicine, photodynamic therapy, mental health support, and medical microservices.

MediPriority is an AI-driven healthcare management system designed to streamline the patient care journey from pre-consultation to post-consultation. Leveraging advanced AI technologies, MediPriority aims to improve patient outcomes, enhance efficiency, and reduce the administrative burden on healthcare providers.

This project uses a TinyVGG16-based CNN to classify MRI scans for Alzheimer's Disease stages: Mild Impairment, Moderate Impairment, No Impairment, and Very Mild Impairment. It includes Jupyter notebooks for training and prediction, and a Streamlit app for easy inference. The model achieves high metrics in predicting Alzheimer's stages.

This repository shows how we can develop and save ML and DL model and use them to predict and diagnosis different diseases.

This project leverages deep learning techniques to detect and predict various dental diseases from panoramic dental X-ray images (OPG - Orthopantomogram). It uses the YOLOv8 object detection model for localizing and identifying diseased regions, enabling automated screening and assistance for dental professionals.

Predicting wether your all ok or not based off of data

This project uses OCR and machine learning to extract CBC values from reports and predict urgency levels. As of now, it supports image/pdf inputs, manual corrections, and SHAP explainability. Ideal for medical AI, healthcare OCR, and automated lab report analysis.

AI-Driven Clinical Trial Matching & Protocol Optimization System

A machine learning-based web app for detecting Parkinson's disease from voice recordings. The app extracts key voice features, applies pre-trained models, and provides real-time predictions of Parkinson's likelihood. Built using Streamlit, Librosa, and scikit-learn.

Hackathon notes

Transforms raw EEG from the DEAP dataset into Differential Entropy features. A BiLSTM captures temporal dynamics, while a Multi-View Graph Transformer models spatial and spectral relationships using self-attention. Built for accurate and interpretable EEG-based emotion recognition.

🩺🤖 PulmoVision Pro — AI-powered Pneumonia Detection from Chest X-Rays with Grad-CAM interpretability & Streamlit dashboard

A lightweight Java FHIR API server for configuration-driven HL7 FHIR based services with AI capabilities, support both native and MCP plugins for extensible healthcare APIs and workflows

[ICLR 2026] HiMAE: Hierarchical Masked Autoencoders Discover Resolution-Specific Structure in Wearable Time Series

🩺 Predict multiple diseases using advanced machine learning techniques to enhance early diagnosis and support healthcare decision-making effectively.

No description provided.

A web-based application that uses deep learning to analyze X-ray images, providing real-time medical insights.