Development of Hybrid Optoacoustic and Ultrasound Imaging Systems

Abstract

Optoacoustic (OA) imaging provides a unique contrast to visualize and quantify light-absorbing tissue chromophores in living organisms. The light excitation of tissues together with consecutive sound wave acquisition combines optical contrast with ultrasound (US) resolution for functional and molecular imaging applications. Furthermore, multispectral optoacoustic tomography (MSOT) can provide spectral information in multiwavelength illumination mode by extracting the absorption properties of underlying tissues. Due to the unique contrast generated by OA images, this biomedical imaging modality is well-established in preclinical settings and clinical applications are introduced recently. However, wider adaptation and application of OA imaging in research and clinical settings as a standard tool require further improvements in contrast, spatial and temporal resolution, standardization of image processing methods, and open-source data sharing. The rich contrast in OA images originating from absorption and scattering in the tissues lacks the information about anatomy and elastic properties. Reflection US is a well-established method in biomedical imaging to monitor anatomical structures. Transmission US can provide information about speed of sound (SoS) changes, acoustic attenuation, and elastic properties of the tissues. Hybridization of these modalities in one imaging setup will benefit from the unique contrast provided by each of them. However, the development of such a hybrid imaging system imposes different constraints on transducer array designs, data transfer rates, and signal acquisition methods. These constraints can be mitigated to some extent by the development of optimized hardware solutions, image acquisition methods, or signal processing methods. The scope of this thesis is the development of new data acquisition and image processing methods to optimize hybrid OA and US imaging systems. Therefore, automated methods to segment boundaries of the anatomical structures in hybrid optoacoustic ultrasound (OPUS) images are proposed. The limited view or sparse acquisition artifacts in the spatial domain are further reduced by data-driven image processing methods for OA imaging. Signal domain limited view artifact removal is proposed for custom designed detector array optimized for hybrid OPUS imaging. US image acquisition sequences and data processing methods are developed to increase spatial and temporal resolution in tomographic acquisition settings. In addition, the applications of the developed transmission reflection optoacoustic ultrasound (TROPUS) system are explored for quantitative multimodal assessment of mammary tumors and non-alcoholic fatty liver disease (NAFLD). The proposed methods are expected to expedite the adaptation of hybrid OA and US systems in research and clinical settings.