TY - JOUR
T1 - Regression-based neural network for improving image reconstruction in diffuse optical tomography
AU - BALASUBRAMANIAM, GANESH M.
AU - Arnon, Shlomi
N1 - Funding Information: Acknowledgments. The authors thank the European Union’s Horizon 2020 research and innovation programme (Future and Emerging Technologies) for funding the CancerScan Project. The authors also thank the Kreitman School of Advanced Graduate Studies and the Ben-Gurion University of the Negev for providing fellowships to continue the research. Author Contributions: GMB and SA designed the study. GMB performed the numerical simulations. GMB and SA analyzed the data from numerical simulation. GMB constructed the feed-forward network and analyzed the result. Both the authors were involved in writing and reviewing the manuscript. GMB drew Figs. 1, 2, 3, and 4. The authors drew all the images or obtained the images through the study, and none of the images were taken from elsewhere. GMB drew Table 1. Publisher Copyright: © 2022 Optica Publishing Group.
PY - 2022/4/1
Y1 - 2022/4/1
N2 - Diffuse optical tomography (DOT) is a non-invasive imaging technique utilizing multi-scattered light at visible and infrared wavelengths to detect anomalies in tissues. However, the DOT image reconstruction is based on solving the inverse problem, which requires massive calculations and time. In this article, for the first time, to the best of our knowledge, a simple, regression-based cascaded feed-forward deep learning neural network is derived to solve the inverse problem of DOT in compressed breast geometry. The predicted data is subsequently utilized to visualize the breast tissues and their anomalies. The dataset in this study is created using a Monte-Carlo algorithm, which simulates the light propagation in the compressed breast placed inside a parallel plate source-detector geometry (forward process). The simulated DL-DOT system's performance is evaluated using the Pearson correlation coefficient (R) and the Mean squared error (MSE) metrics. Although a comparatively smaller dataset (50 nos.) is used, our simulation results show that the developed feed-forward network algorithm to solve the inverse problem delivers an increment of ∼30% over the analytical solution approach, in terms of R. Furthermore, the proposed network's MSE outperforms that of the analytical solution's MSE by a large margin revealing the robustness of the network and the adaptability of the system for potential applications in medical settings.
AB - Diffuse optical tomography (DOT) is a non-invasive imaging technique utilizing multi-scattered light at visible and infrared wavelengths to detect anomalies in tissues. However, the DOT image reconstruction is based on solving the inverse problem, which requires massive calculations and time. In this article, for the first time, to the best of our knowledge, a simple, regression-based cascaded feed-forward deep learning neural network is derived to solve the inverse problem of DOT in compressed breast geometry. The predicted data is subsequently utilized to visualize the breast tissues and their anomalies. The dataset in this study is created using a Monte-Carlo algorithm, which simulates the light propagation in the compressed breast placed inside a parallel plate source-detector geometry (forward process). The simulated DL-DOT system's performance is evaluated using the Pearson correlation coefficient (R) and the Mean squared error (MSE) metrics. Although a comparatively smaller dataset (50 nos.) is used, our simulation results show that the developed feed-forward network algorithm to solve the inverse problem delivers an increment of ∼30% over the analytical solution approach, in terms of R. Furthermore, the proposed network's MSE outperforms that of the analytical solution's MSE by a large margin revealing the robustness of the network and the adaptability of the system for potential applications in medical settings.
UR - http://www.scopus.com/inward/record.url?scp=85126976390&partnerID=8YFLogxK
U2 - https://doi.org/10.1364/BOE.449448
DO - https://doi.org/10.1364/BOE.449448
M3 - Article
C2 - 35519246
SN - 2156-7085
VL - 13
SP - 2006
EP - 2017
JO - Biomedical Optics Express
JF - Biomedical Optics Express
IS - 4
ER -