ThyroidScan

A medical tool for classifying and segmenting thyroid diseases based on medical images.

Project Overview

Thyroid Scan is an intelligent diagnostic system that uses machine learning to analyze thyroid medical images. The project is designed to assist medical professionals in accurate and rapid classification of pathologies.

Technology Stack

• Programming Language: Python 3.8+
• Deep Learning Framework: PyTorch
• Machine Learning: scikit-learn
• Optimization: Optuna
• Medical Image Processing: MONAI
• Data Analysis: NumPy, Pandas
• Visualization: Matplotlib, Seaborn

Installation & Setup

Prerequisites

• Python 3.8 or higher
• pip (Python package manager)
• CUDA-compatible GPU (recommended for training)
• 8GB+ RAM (16GB+ recommended)

Clone Repository

git clone https://github.com/ArtemKushnir/ThyroidScan.git
cd ThyroidScan

Install Dependencies

Core dependencies

pip install -r requirements.txt

Development dependencies

pip install -r requirements.dev.txt

Quick Start

Basic training example

import src.training_module.main_pipeline.training_system

from src.data_loaders.ddti_loader import DDTILoader
from src.data_loaders.bus_bra_loader import BUSLoader

DATA_PATH = "path to BUS-BRA dataset"
bus_loader = BUSLoader(DATA_PATH)

XML_PATH = "path to DDTI dataset xml folder"
IMAGE_PATH = "path to DDTI dataset image folder"
ddti_loader = DDTILoader(XML_PATH, IMAGE_PATH)

config = training_system.Config(
    loader=ddti_loader,
    models=["svm"],
    experiment_dir="experiments",
    target_metric="f1",
    tune_params=False,
    is_update=False,
)

ml_system = training_system.MLSystemFacade(config)

# Use-case 1: train single model
model_name = ml_system.train_single_model("svm")

# Use-case 2: train all models
models = ml_system.train_all_models()

# Use-case 3: model comparison
exp_id = ml_system.run_model_comparison_experiment("experiment_name", save_plots=True)