|
Due to a demonstrated capability to assess tumor angiogenesis and hypoxia in mammalian systems, there is great interest in applying optoacoustic tomography (OAT) to the study and screening of breast cancer. In order to translate OAT to clinical applications, in silico studies are crucial for studying imaging system parameters that might be impossible to assess via direct experimentation. Previous numerical phantoms have proven to be too unrealistic for rigorous testing of modern image reconstruction methods and clinically relevant signal detection tasks. Recently, the U.S. Food and Drug Administration has released software to generate realistic three-dimensional numerical realizations of the human female breast as part of the Virtual Imaging Clinical Trials for Regulatory Evaluation (VICTRE) project. By careful selection of physical attributes and material coefficients, the VICTRE breast phantom can be customized for particular imaging tasks, but no such customization has been given for OAT. We propose a general framework of in silico studies for OAT breast imaging using the VICTRE breast phantom. We will create an ensemble of OAT breast phantoms, using appropriate optical and acoustic parameters, that have typical sizes and tissue densities. Various lesions will be created and embedded based on clinical scenarios. We will define and perform several signal detection tasks by which the system performance may be compared. Generation of such an ensemble requires substantial computation but once produced, it can be utilized in other numerical simulation studies of the configuration of OAT imaging systems customized for diverse tasks. We will make this ensemble of phantoms publicly available online. The proposed framework will permit standardization of the assessment of 3D OAT data-acquisition parameters and image reconstruction methods.
|
