Stochastic numerical head phantoms to enable virtual imaging studies of transcranial photoacoustic computed tomography

TL;DR

This work tackles the lack of realistic evaluation frameworks for transcranial PACT by introducing a CT-based, stochastic framework to generate ensembles of 3D numerical head phantoms (NHPs) that integrate skull heterogeneity, vasculature, and tissue properties. The two-step process first builds anatomical NHPs from adjunct CT data and then assigns stochastic optical and acoustic-elastic properties to produce full optical and acoustic-elastic NHPs; a case study demonstrates how skull modeling errors degrade image quality and how incorporating scalp vasculature and realistic illumination is essential for valid assessments. The framework supports multiple skull-property models (CT-based, heterogeneous diploë, multi-plate, and homogeneous skull) and allows inter-subject variability, enabling rigorous in silico testing and uncertainty quantification for transcranial PACT reconstruction methods. By providing a public dataset of 50 NHPs and flexible workflows, the approach accelerates development and validation of image reconstruction techniques and system designs for transcranial PACT.

Abstract

Transcranial photoacoustic computed tomography (PACT) is an emerging neuroimaging modality, but skull-induced aberrations can result in severe image artifacts if not compensated for during image reconstruction. The development of advanced image reconstruction methods for transcranial PACT is hindered by the lack of well-characterized, clinically relevant evaluation frameworks. Virtual imaging studies offer a solution, but require realistic numerical phantoms. To address this need, this study introduces a framework for generating ensembles of realistic 3D numerical head phantoms for virtual imaging studies. The framework uses adjunct CT data to create anatomical phantoms, which are then enhanced with stochastically synthesized vasculature and assigned realistic optical and acoustic-elastic properties. The utility of the framework is demonstrated through a case study on the impact of skull modeling errors on transcranial PACT image quality. By allowing researchers to assess and refine reconstruction methods meaningfully, the presented framework is expected to accelerate the development of transcranial PACT.

Figures (9)

• Figure 1: Workflow for generating NHPs for transcranial PACT. The two-step process begins with the modeling of the skull and the establishment of the anatomical NHP, and is finalized by the definition and assignment of optical, elastic, and acoustic properties.
• Figure 2: Segmentation results for establishing the skull model. The left panels show CT intensity projection images along the frontal (top) and longitudinal axis (center), and a cross-section in the transverse plane (bottom) of the skull. The right panels present the corresponding segmented skull plates and layers.
• Figure 3: Four instances of scalp vasculature realization: (a) a vessel tree containing only the superficial temporal artery; (b) vasculature with both arteries and veins; (c) a dense artery vessel tree originating from the top of the head; and (d) vasculature with both dense arteries and veins concentrated in a targeted region. The superficial temporal arteries and veins were synthesized based on realistic anatomy.
• Figure 4: Examples of synthesized cortical vasculature. The left image shows dense cortical vasculature synthesized at the top of the head. The right image illustrates a model of the superior sagittal sinus, synthesized based on brain anatomy. This highlights the framework's capability to generate vasculature for diverse anatomical regions and imaging applications.
• Figure 5: Progressive, component-by-component visualization of the head phantom, illustrating its layered anatomical structure. The left panels show the internal cortical vasculature and brain, while the middle and right panels sequentially add the skull and scalp, respectively, to form the complete external anatomy.