mar-eval 0.1.0

Metal Artifact Reduction Evaluation Toolkit (Annex GG)

mar-eval

A Python toolkit for evaluating Metal Artifact Reduction (MAR) performance in CT imaging.

---

Overview

mar-eval provides a reproducible framework for evaluating MAR performance using objective, task-based metrics.
It supports both digital and quantitative physical test methods consistent with the procedures described in IEC 60601-2-44, Annex GG (informative).

The toolkit implements:

• Channelized Hotelling Observer (CHO) analysis for lesion-detection tasks
• AUC computation as the figure of merit for detectability
• Bias and statistical comparison modules for ΔAUC analysis
• Modular support for simulated (digital) and scanned (physical) image datasets
• A transparent and open-source foundation for regulatory and manufacturer use

---

Installation

Option 1 – Install directly from GitHub

pip install git+https://github.com/cdc15000/mar-eval.git

Option 2 – Clone and install locally

git clone https://github.com/cdc15000/mar-eval.git
cd mar-eval
pip install .

---

Example Usage

from mareval.cho import compute_cho
from mareval.stats import compute_auc

# Example: run CHO analysis on lesion-present and lesion-absent image sets
auc = compute_auc(lesion_present, lesion_absent)
print(f"AUC = {auc:.3f}")

For a complete demonstration, see the example script:
examples/synthetic_demo.py

---

Features

Category	Description
CHO Analysis	Implements a channelized Hotelling observer for model-based detectability tasks
AUC Metrics	Computes area under the ROC curve for lesion-detection performance
Bias Assessment	Quantifies ΔAUC between MAR-enabled and non-MAR reconstructions
Statistical Comparison	Supports one-tailed paired t-tests or nonparametric equivalents
Extensibility	Designed for integration with validated simulators and test devices

---

Contributing

Contributions are welcome.
If you identify issues, propose improvements, or want to extend the toolkit, please open an Issue or submit a Pull Request.

---

Citation

If you use this toolkit in academic or regulatory work, please cite:

Cocchiaraley, C.D., mar-eval: A Python Toolkit for Objective Evaluation of Metal Artifact Reduction in CT Imaging (2025).
Available at: https://github.com/cdc15000/mar-eval

---

License

This project is licensed under the MIT License.
See LICENSE for details.

---

Acknowledgment

Development of this toolkit is informed by ongoing work within IEC TC 62 / SC 62B WG 30 and related DICOM initiatives on Metal Artifact Reduction (MAR) in CT imaging.