Setup

pip install -r requirements_.txt

Ablation study

python ablation_study_sirt_.py

Overview

This project focuses on evaluating the PDM-DART algorithm for image reconstruction in cases with continuous and complex grey-level variations. We designed four distinct 2D phantoms—Layered, Resolution, CT, and Filled—to simulate medical imaging scenarios including soft tissue gradients, sharp edges, and overlapping anatomical structures.

Each phantom is designed to test specific aspects of the reconstruction process:

• Layered Phantom: Tests segmentation accuracy for sharp edges and smooth grey-level transitions.
• Resolution Phantom: Adds small structures to test the algorithm’s ability to handle high-frequency detail.
• CT Phantom: Mimics real CT cross-sections of the human body with diverse anatomical shapes and intensity values.
• Filled Phantom: Contains overlapping and free-form shapes to challenge the method’s pixel classification under complexity.

---

Reconstruction of CT phantom with gradient based optimization

python pinns.py