An improved physics model for multi-material identification in photon counting CT

TL;DR

This paper introduces a physics-based model for accurately calculating effective atomic number and electron density in photon-counting CT, enabling improved multi-material identification with high precision and potential clinical applications.

Contribution

The paper presents a novel physics-based model for material identification in PCCT, surpassing semi-empirical methods in accuracy and enabling simultaneous multi-material discrimination.

Findings

Relative standard deviations for effective atomic number and electron density are less than 1%.

The model accurately separates five different materials in a combined map.

Validated across various materials and energy bin configurations.

Abstract

Photon-counting computed tomography (PCCT) with energy discrimination capabilities hold great potentials to improve the limitations of the conventional CT, including better signal-to-noise ratio (SNR), improved contrast-to-noise ratio (CNR), lower radiation dose, and most importantly, simultaneous multiple material identification. One potential way of material identification is via calculation of effective atomic number and effective electron density from PCCT image data. However, the current methods for calculating effective atomic number and effective electron density from PCCT image data are mostly based on semi-empirical models and accordingly are not sufficiently accurate. Here, we present a physics-based model to calculate the effective atomic number and effective electron density of various matters, including single element substances, molecular compounds, and multi-material mixtures as well. The model was validated over several materials under various combinations of energy bins. A PCCT system was simulated to generate the PCCT image data, and the proposed model was applied to the PCCT image data. Our model yielded a relative standard deviations for effective atomic numbers and effective electron densities at less than 1%. Our results further showed that five different materials can be simultaneously identified and well separated in a effective atomic number - effective electron density map. The model could serve as a basis for simultaneous material identification from PCCT.