Imaging inside highly scattering media using hybrid deep learning and analytical algorithm

• Published in: Journal of biophotonics Vol. 16; no. 10; pp. e202300127 - n/a
• Main Authors: Wiesel, Ben, Arnon, Shlomi
• Format: Journal Article
• Language: English
• Published: Weinheim WILEY‐VCH Verlag GmbH & Co. KGaA 01.10.2023; Wiley Subscription Services, Inc
• Subjects: Algorithms; computational imaging; Deep learning; deep‐learning‐DOT; diffuse tomography; Machine learning; Mathematical analysis; Remote sensing; Scattering; time‐of‐flight; diffuse tomography; computational imaging; deep-learning-DOT; time-of-flight
• ISSN: 1864-063X; 1864-0648
• DOI: 10.1002/jbio.202300127

Imaging through highly scattering media is a challenging problem with numerous applications in biomedical and remote‐sensing fields. Existing methods that use analytical or deep learning tools are limited by simplified forward models or a requirement for prior physical knowledge, resulting in blurry images or a need for large training databases. To address these limitations, we propose a hybrid scheme called Hybrid‐DOT that combines analytically derived image estimates with a deep learning network. Our analysis demonstrates that Hybrid‐DOT outperforms a state‐of‐the‐art ToF‐DOT algorithm by improving the PSNR ratio by 4.6 dB and reducing the resolution by a factor of 2.5. Furthermore, when compared to a deep learning stand‐alone model, Hybrid‐DOT achieves a 0.8 dB increase in PSNR, 1.5 times the resolution, and a significant reduction in the required dataset size (factor of 1.6–3). The proposed model remains effective at higher depths, providing similar improvements for up to 160 mean‐free paths. This paper introduces a hybrid scheme called Hybrid‐DOT, which combines analytically derived image estimates with a deep learning network. The architecture of Hybrid‐DOT is X‐shaped, consisting of two parallel arms. Each arm is responsible for extracting information either from the raw data or from an analytically derived estimate. Through our analysis, we have shown that Hybrid‐DOT surpasses a state‐of‐the‐art ToF‐DOT algorithm by enhancing the PSNR ratio and improving resolution. Furthermore, the proposed model demonstrates consistent effectiveness at greater depths, delivering similar improvements for depths of up to 160 mean‐free paths.