mahbub1969/shamsul repository overview

⁠SHAMSUL: Systematic Holistic Analysis to investigate Medical Significance Utilizing Local interpretability methods in deep learning

The SHAMSUL Demo App explores interpretability for chest X-ray pathology predictions using methods—Grad-CAM, LIME, SHAP, and LRP. It provides heatmaps and evaluation metrics for better insights into the medical significance of predictions made by deep learning models.

⁠How to Run

⁠Online Access

Visit https://shamsul.serve.scilifelab.se/⁠ to use it directly.

⁠OR

⁠Run Locally

⁠Step 1: Install Docker

Install Docker Engine or Docker Desktop on your system by following the official Docker installation guide⁠.

⁠Step 2: Launch the App

1. Open a Terminal (or Windows Terminal).
2. Run this command to download and start the app:

   docker run --rm -p 7860:7860 mahbub1969/shamsul:v5
   

   Copy
3. Open your browser and go to http://localhost:7860/⁠ to use the app.

⁠Step 3: Stop the App

To stop the app, press Control+C in the terminal. Note that the session won’t be saved, so the app will reset to its default state the next time you run it.

⁠Step 4: Remove the Docker Image (Optional)

If you want to free up space, you can remove the Docker image. Use this command in your terminal:

docker image rm mahbub1969/shamsul:v5

For more details, check out the Docker image removal guide⁠.

⁠Key Features

• Multi-Method Interpretability: Incorporates four advanced interpretability methods—LIME, SHAP, Grad-CAM, and LRP—to provide diverse insights into deep learning model predictions.

• Focus on Medical Significancee: Designed specifically for chest radiography pathology prediction, ensuring results are meaningful for clinical applications.

• Comprehensive Visualizations: Generates heatmaps and segmentations to help identify the regions of interest linked to specific pathologies.

• Multi-Label, Multi-Class Analysis: Supports analyzing both single-label and multi-label instances, accommodating a variety of medical imaging needs.

• Quantitative and Qualitative Evaluation: Offers metrics like Intersection over Union (IoU) and detailed visual comparisons with expert annotations for robust performance assessment.

• Integration with CheXpert Dataset: One of the largest chest X-ray datasets to validate predictions and ensure high-quality results.

• User-Friendly Interface: Simplifies interaction by allowing users to upload images.

• Open-Source Access: Code and resources are available, promoting transparency and enabling further development by the research community.

⁠How to Use

1. Download the Dataset: Get the CheXpert dataset (high-resolution one) and the val_labels.csv file from the official CheXpert page: CheXpert Competition⁠.
2. Upload an Image: Select and upload a chest X-ray image listed in the val_labels.csv file.
3. Enter Image Name: Input the exact image name as it appears in the val_labels.csv file to ensure accurate analysis.
4. View Results: Access the predictions, heatmaps, and evaluation metrics generated by the model for the uploaded image.

⁠Examples

Simply click on the Examples section to try out a preloaded example. You can quickly see how the application works without needing to upload your own data!

⁠More Information

The acronym SHAMSUL comes from a Semitic word meaning “the Sun.” It symbolizes our approach to using heatmap-based interpretability methods to shed light on the medical significance behind predictions made by deep learning models. For detailed insights and methodology, please refer to the original research paper:
SHAMSUL: Systematic Holistic Analysis to investigate Medical Significance Utilizing Local interpretability methods in deep learning for chest radiography pathology prediction⁠

⁠Contact

For inquiries or further information, please contact:
Mahbub Ul Alam
Email: [email protected]⁠

⁠Citation

Please acknowledge the following work in papers or derivative software:

M. U. Alam, J. Hollmén, J. R. Baldvinsson, and R. Rahmani, “SHAMSUL: Systematic Holistic Analysis to investigate Medical Significance Utilizing Local interpretability methods in deep learning for chest radiography pathology prediction,” Nordic Machine Intelligence, vol. 3, pp. 27–47, 2023.
https://doi.org/10.5617/nmi.10471⁠

⁠Bibtex Format for Citation

@article{alam2023shamsul,
  title={SHAMSUL: Systematic Holistic Analysis to investigate Medical Significance Utilizing Local interpretability methods in deep learning for chest radiography pathology prediction},
  author={Ul Alam, Mahbub and Hollm{\'e}n, Jaakko and Baldvinsson, J{\'o}n R{\'u}nar and Rahmani, Rahim},
  journal={Nordic Machine Intelligence},
  volume={3},
  number={1},
  pages={27--47},
  year={2023},
  doi={10.5617/nmi.10471}
}