ProSecCo - Prostate Segmentation and Conversion Tool

A comprehensive toolkit for processing, converting, and analyzing prostate-related medical imaging data, particularly focusing on CT and CBCT images.

Overview

This repository contains a collection of tools for handling medical imaging data, specifically focused on prostate-related imaging and segmentation. The toolkit provides functionality for:

• Converting DICOM files to NIfTI format
• Processing RT Structure files
• Creating training datasets
• Generating visualizations
• Evaluating segmentation results
• Handling both CT and CBCT images

Key Components

Data Processing

• dicom2nii.py: Converts DICOM files to NIfTI format
• create_training_data*.py: Prepares training datasets from medical images
• unifyData.py: Standardizes data formats across different sources
• match_image_space.py: Aligns images to a common space

Visualization

• generateFigures.py: Creates visual representations of the data
• foldervisualization.py: Generates folder structure visualizations
• images2video.py: Converts image sequences to video format

Evaluation

• evaluation.py: Contains tools for evaluating segmentation results
• inspectData.py: Utility for examining dataset contents

Usage

This toolkit was developed and used for preprocessing data in the MICCAI paper "Cross-Modality Supervised Prostate Segmentation on CBCT for Adaptive Radiotherapy". The code was specifically designed to:

Requirements

The project uses Python and requires several key dependencies:

• pydicom
• rt_utils
• numpy
• nibabel
• nrrd
• matplotlib

Note

This toolkit is specifically designed for medical imaging data and should be used in accordance with relevant medical data handling protocols and privacy regulations.