KO | EN

GitLite

Minimalist Open Source Repository Search Engine

gitlite — search

#dsh #dsh-plugin #typescript #developer-tools #claude-code #ai-agent #codex #deepseek-harness #ai-agents #cli #react #mcp

HealifyAI--LLM-based-Healthcare-System

★ 52 Open GitHub ↗

Leverages extensive power of multiple Machine Learning algorithms & LLM to provide in-depth answers to medical queries and predicts condition/diseases based on patient symptoms

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

<h1 align="center">HealifyAI - LLM based Healthcare System</h1>

This project aims to develop a comprehensive healthcare system to aid healthcare professionals. While also providing knowledge to patients. It uses a LLM and traditional Machine Learning (ML) to provide in-depth answers to medical health condition queries and can predict diseases based on patient symptoms.<br>
The system consists of two main Modules:
1. Disease Prediction Model
2. HealifyLLM - Q&A Language Model

## Data Collection, Cleaning, Preprocessing
<div align="center">  <strong> Healify-LLM model: </strong> </div>

* Engineered brand new LLM Corpus Dataset of size 6800 samples from scratch. Started with scraping based on [healthline.com](https://www.healthline.com/directory/topics)
* To enhance the corpus for user experience, Sample addition was done with my python script. Enabling it to provide detailed and accurate answers to a wide range of user questioning styles.  <br>

<div align="center">  <strong> Disease Model: </strong> </div>

The model was trained on a kaggle dataset from [Disease-Symptom Knowledge Database](https://people.dbmi.columbia.edu/~friedma/Projects/DiseaseSymptomKB), a database with over hundreds of patient records at the New York Presbyterian Hospital, USA. <br> 
Covering 135 Categories of important common but also rare diseases/health conditions. <br>
From a total of 400 Symptoms. <br>

Disease dataset was processed to clean the noisy symptoms, UMLScode etc. <br>
LLM dataset processing required data seperation, sample addition.<br>
The scraping can be found in `scraper` folder. <br> 
All final datasets stored in `datasets` folder. <br> 
All cleaning and processing found in `notebooks` folder.

## Model Training:
<div align="center">  <strong> Healify-LLM model: </strong> </div>

__Hyperparameters__: We used a `batch size` of 8. And `learning rate` was set dynamically using `Fast.ai`'s `learning rate finder` at every stage for Fine-Tuning. <br>

__Training Procedures:__ We used `HuggingFace` for the model and imported Fast.ai for hyperparameter tuing
* RoBERTa model has been used because the QA dataset is complex.
* Training was done using [ULMFiT Research Paper](https://arxiv.org/abs/1801.06146)'s 3-stage training policy.
* The model was fine-tuned with 6800 samples around 98% accuracy in 12 epochs. The model was tracked to avoid overfitting observing loss. The model was trained using NVIDIA T4 GPU.  With good sample ratio, as total 6800 samples for just 135 diseases. <br>

<div align="center">  <strong> Disease Model: </strong> </div>

Model was trained with `sklearn`'s ensemble random forest algorithm leveraging mutiple decision tree algorithms. This model predicts potential diseases based on the symptoms input.<br>

## Model Deployment
A `Gradio App` was coded to deploy the LLM model in HuggingFace. This live huggingfaces API is later integrated in the Back-End. The Gradio implementation can be found in `deployment_hf` folder and online [here](https://huggingface.co/spaces/tanvir-ishraq/healifyLLM-classifier)

## Live Website Deployment
Deployed a Flask App, built to provide user interface to users. Check the `flask-deployment` github branch. The website is live [here](https://healifyai-llm.onrender.com)

## Future work and limitations
The combination of these two components allows for a robust interactive healthcare system that can assist both patients and healthcare professionals in diagnosing diseases, finding relevant medical information and diseases relation potentially. The system is designed to be user-friendly, with an intuitive interface that makes it easy for anyone to use.
Please note that while this system can provide valuable insights and information, it is not intended to replace professional medical advice. Always consult with a healthcare professional for medical concerns. <br>

I do have plan to incorporate a more powerful model for a more seamless interactive experience. Currently I'm occupied with working with a startup and their project. In my free time, I do hope to execute it.

<!--
Disease Prediction Model: This component uses traditional ML algorithm to predict potential diseases based on the symptoms input by the user. Covering a total of 135 categories of common and as well as rare yet important health conditions, diseases, psychology disorders such as diabetes, dehydration, depression, bipolar disorder, HIV, breast cancer, stroke, pneumonia, flu, asthma, obesity and so on. The model is trained on a large dataset of hundreds to thousands of patient records (denoted by frequency in dataset) to ensure reliable predictions based on NY Hospital based [Disease-Symptom Knowledge Database](https://people.dbmi.columbia.edu/~friedma/Projects/DiseaseSymptomKB/).

QA Language Model: This component uses a Language Model (LLM) to answer medical queries from users. The LLM is trained on my from-scratch scraped then enhanced corpus dataset of medical queries & professional solutions, enabling it to provide detailed and accurate answers to a wide range of medical questions. Sample addition was done to enhance the dataset for user experience. Covering urgent topics of diagnosis, treatment, prevention, causes, risks, complications, details of symptoms, disease description.

The combination of these two components allows for a robust interactive healthcare system that can assist both patients and healthcare professionals in diagnosing diseases, finding relevant medical information and diseases relation potentially. The system is designed to be user-friendly, with an intuitive interface that makes it easy for anyone to use. -->

HealifyAI--LLM-based-Healthcare-System

Explore Similar Repositories

Was this content helpful?

Recent Feedback

Download README