Dual-encoder deep learning networks to enhance diffuse optical imaging for highly scattered nonhomogeneous media

Nazish Murad, Ya Fen Hsu, Min Chun Pan

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

This paper aims to demonstrate a novel deep-learning network that addresses the prediction of breast tumors for diffuse optical imaging. Two learning schemes, signal encoder and image encoder, in the proposed network are designed for reconstructing optical-property images. The former processing method takes boundary data directly to deep networks, and predicts the optical-coefficient distribution, while the latter feeds images obtained by inverse image reconstruction with artifacts and sometimes hard-to-localized tumors. All 10,000 samples of synthesized homogeneous and heterogeneous phantoms were randomly selected for training, validation, and testing of performance. Twelve phantom samples were employed to justify its effectiveness in real applications.

Original languageEnglish
Title of host publicationComputational Optical Imaging and Artificial Intelligence in Biomedical Sciences
EditorsLiang Gao, Guoan Zheng, Seung Ah Lee
PublisherSPIE
ISBN (Electronic)9781510669734
DOIs
StatePublished - 2024
EventComputational Optical Imaging and Artificial Intelligence in Biomedical Sciences 2024 - San Francisco, United States
Duration: 27 Jan 202429 Jan 2024

Publication series

NameProgress in Biomedical Optics and Imaging - Proceedings of SPIE
Volume12857
ISSN (Print)1605-7422

Conference

ConferenceComputational Optical Imaging and Artificial Intelligence in Biomedical Sciences 2024
Country/TerritoryUnited States
CitySan Francisco
Period27/01/2429/01/24

Keywords

  • Diffuse optical imaging
  • dual-encoder deep learning network
  • measured phantom justification
  • signal/image encoder

Fingerprint

Dive into the research topics of 'Dual-encoder deep learning networks to enhance diffuse optical imaging for highly scattered nonhomogeneous media'. Together they form a unique fingerprint.

Cite this