Teknofest 2024 AI in Healthcare Competition Code

This repository contains the code for the models developed by our 5-member team for the Teknofest AI in Healthcare Competition. It includes a dual-module system that detects calcifications and masses from mammography images and performs BI-RADS classification from report texts.

About the Project

Core Modules

1. Image Processing Module — On mammography images:

   • Calcification detection
   • Mass detection
   • Anomaly localization

2. Natural Language Processing (NLP) Module — On report texts:

   • BI-RADS categorization
   • Clinical term recognition
   • Multi-label classification

Technical Specifications

• Platform: Python 3.8+
• ML Framework: TensorFlow/Keras
• NLP Tools: BERT-based models
• Image Processing: CNN architectures (YOLO, RCNN), PyTorch

Installation

1. Clone the repository:

   git clone https://github.com/Kizil-AI/Teknofest2024

Note: Due to dataset licensing restrictions, the competition dataset is not included in the repository. To use the project with your own dataset, please update the data paths accordingly.

Contribution Guide

This README aims to make the project clear for both developers and interested users. If you have any questions or suggestions for contributions, please feel free to get in touch.

Important Notes

• Performance Metrics:

  • Image Models:

    • YOLO: 0.51 F1-score
    • RCNN: 0.535 F1-score
    • Ensemble: 0.617 F1-score
    • BI-RADS Classification: 0.91 F1-score

  • NLP Model: 0.92 F1-score

---