C_ADMM

Demo code for the paper:

Constrained and Regularized Quantitative Ultrasound Parameter Estimation using ADMM
Accepted at SPIE Medical Imaging 2026

This repository provides a MATLAB implementation of an ADMM-based framework for constrained and regularized estimation of quantitative ultrasound (QUS) parameters from power spectrum data.

---

Repository Structure

C_ADMM/ │ ├── Demo.m # Main demo script ├── Demo.asv # Autosave file ├── AdmmPhantom.m # ADMM solver for QUS parameter estimation ├── extract_matrices.m # Constructs system matrices for optimization ├── lasso4.m # LASSO solver (ℓ1 regularization) ├── lasso5.m # Alternative LASSO implementation ├── wd.m # Weighting / windowing utilities ├── PowerSpectrumLocationCM.mat# Example power spectrum & spatial locations ├── LICENSE └── README.md

---

Requirements

• MATLAB (R2019b or newer recommended)
• No external toolboxes are required beyond standard MATLAB functionality

---

Getting Started

1. Download Required Data

Download the required power spectrum and spatial location .mat files from the following link

Place the downloaded .mat files in the same directory as Demo.m.

---

2. Run the Demo

In MATLAB, navigate to the repository directory and run:

Demo

The script will:

• Load the power spectrum and spatial location data
• Perform constrained and regularized QUS parameter estimation using ADMM
• Estimate parameters on a line-by-line basis
• Save the estimated parameters for further analysis and visualization

---

Output

The demo computes QUS parameters (e.g., attenuation and backscatter-related quantities) for each RF line.
The resulting parameters are stored as MATLAB variables and can be directly used to reproduce the figures shown in the paper.

---

Method Overview

The proposed framework formulates QUS parameter estimation as a constrained inverse problem solved using the Alternating Direction Method of Multipliers (ADMM).

Key features include:

• Explicit physical constraints
• ℓ1-regularization for robustness
• Improved stability under noise and ill-conditioned scenarios

Full methodological details and validation results are provided in the associated SPIE publication.

---

Citation

If you use this code in your research, please cite:

Ali K. Z. Tehrani et al.
Constrained and Regularized Quantitative Ultrasound Parameter Estimation using ADMM
Proceedings of SPIE Medical Imaging, 2026.

A BibTeX entry will be added upon final publication.

---

Contact

For questions, suggestions, or issues, please open a GitHub issue or contact the authors.