UPAMNet: A unified network with deep knowledge priors for photoacoustic microscopy

IF 7.1 1区医学 Q1 ENGINEERING, BIOMEDICAL

Photoacoustics Pub Date : 2024-04-25 DOI:10.1016/j.pacs.2024.100608

Yuxuan Liu , Jiasheng Zhou , Yating Luo , Jinkai Li , Sung-Liang Chen , Yao Guo , Guang-Zhong Yang

{"title":"UPAMNet: A unified network with deep knowledge priors for photoacoustic microscopy","authors":"Yuxuan Liu , Jiasheng Zhou , Yating Luo , Jinkai Li , Sung-Liang Chen , Yao Guo , Guang-Zhong Yang","doi":"10.1016/j.pacs.2024.100608","DOIUrl":null,"url":null,"abstract":"<div>Photoacoustic microscopy (PAM) has gained increasing popularity in biomedical imaging, providing new opportunities for tissue monitoring and characterization. With the development of deep learning techniques, convolutional neural networks have been used for PAM image resolution enhancement and denoising. However, there exist several inherent challenges for this approach. This work presents a Unified PhotoAcoustic Microscopy image reconstruction Network (UPAMNet) for both PAM image super-resolution and denoising. The proposed method takes advantage of deep image priors by incorporating three effective attention-based modules and a mixed training constraint at both pixel and perception levels. The generalization ability of the model is evaluated in details and experimental results on different PAM datasets demonstrate the superior performance of the method. Experimental results show improvements of 0.59 dB and 1.37 dB, respectively, for 1/4 and 1/16 sparse image reconstruction, and 3.9 dB for image denoising in peak signal-to-noise ratio.</div>","PeriodicalId":56025,"journal":{"name":"Photoacoustics","volume":"38 ","pages":"Article 100608"},"PeriodicalIF":7.1000,"publicationDate":"2024-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2213597924000259/pdfft?md5=6e9fbd1a1d430dc8823039b6c73b9b70&pid=1-s2.0-S2213597924000259-main.pdf","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Photoacoustics","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2213597924000259","RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, BIOMEDICAL","Score":null,"Total":0}

引用次数: 0

Abstract

Photoacoustic microscopy (PAM) has gained increasing popularity in biomedical imaging, providing new opportunities for tissue monitoring and characterization. With the development of deep learning techniques, convolutional neural networks have been used for PAM image resolution enhancement and denoising. However, there exist several inherent challenges for this approach. This work presents a Unified PhotoAcoustic Microscopy image reconstruction Network (UPAMNet) for both PAM image super-resolution and denoising. The proposed method takes advantage of deep image priors by incorporating three effective attention-based modules and a mixed training constraint at both pixel and perception levels. The generalization ability of the model is evaluated in details and experimental results on different PAM datasets demonstrate the superior performance of the method. Experimental results show improvements of 0.59 dB and 1.37 dB, respectively, for 1/4 and 1/16 sparse image reconstruction, and 3.9 dB for image denoising in peak signal-to-noise ratio.

查看原文本刊更多论文

UPAMNet：用于光声显微镜的具有深度知识先验的统一网络

光声显微镜（PAM）在生物医学成像领域越来越受欢迎，为组织监测和表征提供了新的机会。随着深度学习技术的发展，卷积神经网络已被用于 PAM 图像分辨率增强和去噪。然而，这种方法存在一些固有的挑战。本研究提出了一种用于 PAM 图像超分辨率和去噪的统一光声显微图像重建网络（UPAMNet）。所提出的方法利用了深度图像先验，在像素和感知水平上纳入了三个有效的基于注意力的模块和混合训练约束。对模型的泛化能力进行了详细评估，不同 PAM 数据集的实验结果证明了该方法的卓越性能。实验结果表明，1/4 和 1/16 稀疏图像重建的性能分别提高了 0.59 dB 和 1.37 dB，图像去噪的峰值信噪比提高了 3.9 dB。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Photoacoustics Physics and Astronomy-Atomic and Molecular Physics, and Optics

CiteScore

11.40

自引率

16.50%

发文量

审稿时长

53 days

期刊介绍： The open access Photoacoustics journal (PACS) aims to publish original research and review contributions in the field of photoacoustics-optoacoustics-thermoacoustics. This field utilizes acoustical and ultrasonic phenomena excited by electromagnetic radiation for the detection, visualization, and characterization of various materials and biological tissues, including living organisms. Recent advancements in laser technologies, ultrasound detection approaches, inverse theory, and fast reconstruction algorithms have greatly supported the rapid progress in this field. The unique contrast provided by molecular absorption in photoacoustic-optoacoustic-thermoacoustic methods has allowed for addressing unmet biological and medical needs such as pre-clinical research, clinical imaging of vasculature, tissue and disease physiology, drug efficacy, surgery guidance, and therapy monitoring. Applications of this field encompass a wide range of medical imaging and sensing applications, including cancer, vascular diseases, brain neurophysiology, ophthalmology, and diabetes. Moreover, photoacoustics-optoacoustics-thermoacoustics is a multidisciplinary field, with contributions from chemistry and nanotechnology, where novel materials such as biodegradable nanoparticles, organic dyes, targeted agents, theranostic probes, and genetically expressed markers are being actively developed. These advanced materials have significantly improved the signal-to-noise ratio and tissue contrast in photoacoustic methods.