SegMNet v0.1.0

SegMNet is a Python-based web application designed for processing and segmenting kidney CT scans. Users can upload CT studies in .nii.gz format, which are processed by a segmentation model to generate 3D segmentations. While future versions aim to support interactive online visualization, the application currently enables downloading processed results for further use.

---

Table of Contents

1. Features
2. Technologies Used
3. Project Structure
4. Setup Instructions
5. Usage Guide
6. API Endpoints
7. Contributing
8. License

---

Features

• File Upload:
  Supports all popular images formats (.png, .jpg, .jpeg, .webp), numpy-arrays, torch-tensors and .nii.gz files for 3D CT scan uploads.
• Segmentation:
  Leverages a trained segmentation model for efficient predictions. Predict both kidney and tumor, if present.
• Visualization:
  Interactive visualization using Gradio functions.
• Download Results:
  Export segmentation outputs in .npy or .png.

---

Technologies Used

| Component | Technology/Library | Purpose |
|------------------------|--------------------|-----------------------------------|
| Backend Framework | Gradio | API management and web server | | Segmentation Model | PyTorch | 3D segmentation model |
| Visualization | Gradio | Interactive visualization | | Deployment | Poetry | Containerization and serving |

---

Setup Instructions

Prerequisites

• Python 3.11+
• PyTorch 2.4+
• Poetry
• CUDA-enabled GPU (optional, for faster inference)

Installation Steps

1. Clone the Repository

   git clone https://github.com/lskog7/SegMNet.git  
   cd SegMNet  
   

2. Create virtual environment and install dependencies

   poetry config virtualenvs.in-project true
   poetry install
   poetry shell
   

3. Run the Server

   gradio segmnet.py
   

---

Usage Guide

1. Upload CT Scans: Use simple Gradio interface to upload an image.
2. Process Scans: The segmentation model processes uploads to generate results.
3. Review and Download Results: Seen the model output and export processed files in desired formats.

---

Contributing

We welcome contributions! To contribute:

1. Fork the repository.
2. Create a feature branch.
3. Submit a pull request with detailed changes.

---

License

This project is licensed under the MIT License.

---

Note

Features and APIs marked as "Planned" are not yet implemented. Updates will be rolled out incrementally.