🧠 MINDSETS - Dementia Differential Diagnosis

📌 Overview

This project processes MRI scans for segmentation, radiomics feature extraction, and classification into four classes: Control, Mild Cognitive Impairment (MCI), Alzheimer's Disease (AD), and Vascular Dementia (VaD). The pipeline integrates SynthSeg segmentation and a trained classification model.

🏗️ Architecture

The pipeline consists of three main steps:

1. 🧠 Segmentation: MRI scans are segmented using the SynthSeg model.
2. 📊 Feature Extraction: Radiomics features are extracted from segmented brain structures.
3. 🤖 Classification: Extracted features are fed into a pre-trained classifier to predict the disease category.

🛠 Installation

Ensure you have the required dependencies installed:

#Create Environment
conda create -n MINDSETS python=3.8
#Activate Environment
conda activate MINDSETS
#Clone Repo
git clone https://github.com/salmasoma/MINDSETS/
cd MINDSETS
#Install requirements
pip install -r requirements.txt

Download model weights: Here

🚀 Usage

To run the MRI classification pipeline, use the following command:

Basic Usage:

python scripts/inference.py --input <path_to_mri.nii.gz> --output <output_directory> --model <model_path>

With verbose output:

python scripts/inference.py --input <path_to_mri.nii.gz> --output <output_directory> --model <model_path> --verbose

📂 Output Files

After running the pipeline, the following output files will be generated:

• 🏷 Segmentation Mask: <span><filename>_synthseg.nii.gz</span>
• 📜 Extracted Features: <span><filename>_radiomics.csv</span>
• 📑 Classification Results: <span><filename>_classification.txt and <filename>_classification.json</span>

📥 Demo & Paper

Live Demo: HuggingFace Space

Paper: Link

📝 Citation

If you use this project in your research, please cite:

@article{hassan2025mindsets,
  title={MINDSETS: Multi-omics Integration with Neuroimaging for Dementia Subtyping and Effective Temporal Study},
  author={Hassan, Salma and Akaila, Dawlat and Arjemandi, Maryam and Papineni, Vijay and Yaqub, Mohammad},
  journal={Scientific Reports},
  volume={15},
  number={1},
  pages={1--12},
  year={2025},
  publisher={Nature Publishing Group}
}