BCI: Breast Cancer Immunohistochemical Image Generation through Pyramid Pix2pix

SHEPHERD: Few shot learning for phenotype-driven diagnosis of patients with rare genetic diseases

Code for "Graph-Evolving Meta-Learning for Low-Resource Medical Dialogue Generation". [AAAI 2021]

Intelligent Python service with FastAPI for real-time heart disease predictions using machine learning. Features AI-assisted consultations, user authentication, analysis history, RESTful API, and comprehensive error handling. Secure and scalable solution for healthcare applications.

[ACL 2025 Findings] "Worse than Random? An Embarrassingly Simple Probing Evaluation of Large Multimodal Models in Medical VQA"

A Deep Learning Based approach for diagnosis of Schizophrenia using EEG brain recordings

A Vietnamese dataset of over 12 thousands questions about common disease symptoms. Perfect for researchers and developers building Vietnamese healthcare chatbots or disease prediction models.

This project leverages advanced AI agents from crewAI to assist doctors in diagnosing medical conditions and recommending treatment plans based on patient-reported symptoms and medical history. The solution uses Streamlit for the user interface and crewai library to define and manage AI agents and tasks.

Early detection of Autism Spectrum Disorder (ASD) is crucial for children's development, yet the diagnostic procedure remains challenging. EyeTism employs machine learning on eye tracking data from both high-functioning ASD and typically developing children (TD) to create a diagnostic tool based on their distinct visual attention patterns.

An AI-powered deep learning system using VGG16 transfer learning to classify brain tumors (glioma, meningioma, pituitary, no tumor) from MRI scans. Built with TensorFlow, deployed on Render with Flask.

Machine learning diabetes risk predictor using KNN classifier. Built as part of Intel AI for Youth program with scikit-learn, pandas, and data visualization for early medical screening.

Ethnic bias analysis in medical imaging AI: Demonstrating that explainable-by-design models achieve 80% bias reduction across 5 ethnic groups (50k images)

A MATLAB-based GUI tool that predicts diabetes risk using machine learning (Random Forest) on the Pima Indian dataset.

A full-stack SaaS platform for hosting and monetizing medical AI models. It features a complete credit-based payment system (Razorpay), JWT/OAuth2 authentication, a full admin dashboard, and an LLM-powered assistant. The platform is live with its first two models for clinical diagnostics.

Using TensorFlow Object Detection API to detect blood cells

Clustering Analysis of all available research data on the Iowa Gambling Task(list of sources in readme) using R. The Scripts produce the output for the most common archetypes among the dataset of one researcher using PCA.

Early detection of Autism Spectrum Disorder (ASD) is crucial for children's development, yet the diagnostic procedure remains challenging. EyeTism employs machine learning on eye tracking data from both high-functioning ASD and typically developing children (TD) to create a diagnostic tool based on their distinct visual attention patterns.

Code for my Master Thesis project on "Prompting Techniques for Natural Language Generation in the Medical Domain" at the University of Bologna

MediCheck is a web-based application that predicts possible diseases based on user-inputted symptoms using machine learning.