An AI-powered application that creates personalized nutrition plans for individuals with diabetes. The system integrates multiple health-related factors (medical analyses, weight, height, age, dietary habits, health conditions) along with socioeconomic factors (income, location, education, literacy levels) to ensure plans are accessible and effective for diverse populations, including those in rural or underserved communities.
Python, OpenAI, Streamlit, Pinecone vector database
The AI Doctor is a virtual healthcare assistant application that provides a medical consultation experience. Built with Python, Streamlit and powered by OpenAI's GPT models, it provides personalized data-driven medical guidance through an interactive interface.
LLM, Python, OpenAI, Streamlit
A web application that translates complex medical documents into plain English using AI, helping patients better understand their lab result.
Python FastAPI, OpenAI AP, PyMuPDF
In the rapidly advancing field of neuroscience, we heavily rely on brain MRI scans to diagnose diseases, assess injuries, and deepen our research. A key technique in our work is brain masking, where we isolate the brain tissue from surrounding structures such as the skull and skin. This isolation is crucial, as it allows us to focus our analysis exclusively on the brain, enabling more precise studies. Despite its critical role, brain masking has traditionally been a manual process. This method is slow and carries a high risk of human error, creating a significant bottleneck in our workflows. Furthermore, the existing automated tools available were unreliable, often failing because they relied on overly simplistic methods of geometric detection or outdated algorithms. Faced with these challenges, we turned to a more innovative solution: Convolutional Neural Networks (CNNs), a technology that has been successful in various industries for quick and accurate image segmentation.
Python, Computer Vision, Deep Learning
Luna is an open-source C/C++ software package for manipulating and analyzing polysomnographic recordings, with a focus on the sleep EEG.
C++
Moonlight is an interactive web viewer for EDF signal and annotation data, specifically designed for polysomnographic data.
Shiny
Protein structure determines protein function is a fundamental problem in molecular biology. We can use Computational Geometry to study the geometry, shape, physicochemical texture of binding surfaces to understand how they work and how functional roles of proteins can be predicted.
C++
This is a Next.js project bootstrapped with create-next-app
JavaScript, TypeScript