Biomedical optics express最新文献_第3页

Prior photon measurement density functions constructed in standard channel space for functional near-infrared spectroscopy. 功能近红外光谱在标准通道空间中构造的先验光子测量密度函数。

IF 3.2 2区医学

Biomedical optics express Pub Date : 2026-09-04 eCollection Date: 2026-10-01 DOI: 10.1364/BOE.604796

Yang Zhao, Li-Jiang Wei, Xue-Ting Sun, Fa-Rui Liu, Lei Zhang, Chao-Zhe Zhu

{"title":"Prior photon measurement density functions constructed in standard channel space for functional near-infrared spectroscopy.","authors":"Yang Zhao, Li-Jiang Wei, Xue-Ting Sun, Fa-Rui Liu, Lei Zhang, Chao-Zhe Zhu","doi":"10.1364/BOE.604796","DOIUrl":"10.1364/BOE.604796","url":null,"abstract":"Photon measurement density function (PMDF) determines the measurement locations and their associated sensitivity to the hemodynamic changes for an fNIRS channel. It is therefore essential for pre-experimental optode arrangement and post-experimental signal analysis and interpretation in fNIRS studies. However, accurate PMDF estimation requires subject-specific MRI data, which are unavailable in common fNIRS studies. Previous studies often rely on PMDFs derived from a single head template (template-PMDF), ignoring inter-subject variability in anatomy and optode placement, and are limited to sparse predefined scalp locations. Here, we propose a methodological framework to construct prior PMDFs (prior-PMDF) by synthesizing PMDFs derived from multiple subjects in a structural MRI database. This approach incorporates inter-subject variability and enables estimation of PMDFs in the standard brain space for arbitrary scalp locations. We demonstrate the utility of the prior-PMDF in estimating regional sensitivities, including voxel-wise sensitivity, sensitivity to intra- and extracerebral tissues, and sensitivity to atlas-defined brain regions. The estimation accuracy of the prior-PMDF is evaluated by comparing it with subject-specific PMDFs derived from MRI data that are not included in the construction database. Furthermore, comparison with template-PMDF showed that the proposed method significantly improves accuracy, thereby supporting more reliable PMDF estimation in MRI-free fNIRS studies.","PeriodicalId":8969,"journal":{"name":"Biomedical optics express","volume":"17 10","pages":"5034-5047"},"PeriodicalIF":3.2,"publicationDate":"2026-09-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13671689/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149215024","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

FBS-DiffNet: a deep unfolding network with diffusion prior for joint reconstruction of optical absorption and sound speed in photoacoustic tomography. FBS-DiffNet：一个具有扩散先验的深度展开网络，用于光声断层成像中光吸收和声速的联合重建。

IF 3.2 2区医学

Biomedical optics express Pub Date : 2026-09-04 eCollection Date: 2026-10-01 DOI: 10.1364/BOE.608902

Sun Zheng, Ding Gang'ao

{"title":"FBS-DiffNet: a deep unfolding network with diffusion prior for joint reconstruction of optical absorption and sound speed in photoacoustic tomography.","authors":"Sun Zheng, Ding Gang'ao","doi":"10.1364/BOE.608902","DOIUrl":"10.1364/BOE.608902","url":null,"abstract":"Quantitative photoacoustic tomography faces challenges arising from ill-posedness caused by acoustic heterogeneity and multi-parameter coupling. We propose the forward-backward splitting diffusion network (FBS-DiffNet), to simultaneously reconstruct optical absorption coefficient and sound speed distributions. This framework unfolds the classical forward-backward splitting algorithm into a cascaded architecture. Each level integrates a data consistency module driven by a residual gradient step network and a generative joint diffusion module. This configuration ensures physical fidelity to measured signals while exploiting spatial correlations between parameters at anatomical boundaries to provide synergistic regularization. Experimental results on simulation, phantom, and in vivo datasets demonstrate that FBS-DiffNet surpasses uniform sound speed methods and prominent sound speed estimation techniques such as K-means-GMM or NF-APACT concerning both visual quality and quantitative metrics. Ablation studies verify the contribution of each key component to these results. By combining physical interpretability with deep learning representational capacity, FBS-DiffNet offers a promising approach for high-precision quantitative photoacoustic tomography.","PeriodicalId":8969,"journal":{"name":"Biomedical optics express","volume":"17 10","pages":"4992-5014"},"PeriodicalIF":3.2,"publicationDate":"2026-09-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13671699/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149215009","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Erythema and hyperpigmentation delineation for cutaneous chronic graft-versus-host disease using hyperspectral reconstruction. 皮肤慢性移植物抗宿主病的红斑和色素沉着的高光谱重建。

IF 3.2 2区医学

Biomedical optics express Pub Date : 2026-09-02 eCollection Date: 2026-10-01 DOI: 10.1364/BOE.610164

Mayaank Pillai, Ruijing E Han, Inga Saknite, Eric Tkaczyk, Audrey K Bowden

引用次数: 0

DASHING-LSCI: deep learning-based anatomical multiclass segmentation for human intraoperative neurosurgical guidance - laser speckle contrast imaging. dash - lsci：基于深度学习的人体术中神经外科引导解剖多类分割-激光散斑对比成像。

IF 3.2 2区医学

Biomedical optics express Pub Date : 2026-09-01 DOI: 10.1364/BOE.605293

Ebenezer Raj Selvaraj Mercyshalinie, Paul Calle, Jacqueline O Stovall, Brayden J Girard, Diogo Pinheiro Cordeiro, Andrew M Bauer, Johannes Goldberg, Christopher S Graffeo, Andreas Raabe, Michael T Lawton, Andrew K Dunn, Chongle Pan, David R Miller

{"title":"DASHING-LSCI: deep learning-based anatomical multiclass segmentation for human intraoperative neurosurgical guidance - laser speckle contrast imaging.","authors":"Ebenezer Raj Selvaraj Mercyshalinie, Paul Calle, Jacqueline O Stovall, Brayden J Girard, Diogo Pinheiro Cordeiro, Andrew M Bauer, Johannes Goldberg, Christopher S Graffeo, Andreas Raabe, Michael T Lawton, Andrew K Dunn, Chongle Pan, David R Miller","doi":"10.1364/BOE.605293","DOIUrl":"10.1364/BOE.605293","url":null,"abstract":"Laser speckle contrast imaging (LSCI) provides real-time, label-free visualization of cerebral blood flow in vessels and perfusion in cortical tissue during neurosurgery. At present, surgeons interpret LSCI-derived flow and perfusion changes qualitatively between microsurgical maneuvers, which limits continuous quantitative analysis. Automated delineation of blood vessels and cortical tissue is a prerequisite for such analysis. Here we evaluate a deep learning approach for multiclass segmentation of intraoperative LSCI images into blood vessels, cortical tissue, and background (e.g., surgical instruments, skull). We curated 90 intraoperative LSCI images acquired during 18 unique human neurosurgery procedures across three prospective observational clinical studies (30 images per site) and trained nnU-Net (Version 2) to segment images. To assess cross-site generalization, one site was reserved exclusively for testing while the images from the other two were pooled and split into training and validation sets. Across three held-out sites, the model achieved mean Dice scores of 0.88 for background and 0.73 for cortical tissue, but a lower and more variable 0.41 for blood vessels, reflecting the difficulty of segmenting thin, sparse vessel structures. These results demonstrate the feasibility of deep-learning segmentation for extracting anatomical structure from neurosurgical LSCI images, a capability needed for real-time, continuous blood flow and perfusion analysis to support surgical guidance.","PeriodicalId":8969,"journal":{"name":"Biomedical optics express","volume":"17 9","pages":"4933-4946"},"PeriodicalIF":3.2,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13592254/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149015567","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Personalized digital pipeline for basal cell carcinoma delineation using hyperspectral imaging. 使用高光谱成像对基底细胞癌进行个性化数字管道描绘。

IF 3.2 2区医学

Biomedical optics express Pub Date : 2026-09-01 DOI: 10.1364/BOE.592641

Paulina Ibek, Alexander Degener, Magne Stridh, Marcus Tegnér, Patrik Edén, Victor Olariu, Malin Malmsjö, Aboma Merdasa

{"title":"Personalized digital pipeline for basal cell carcinoma delineation using hyperspectral imaging.","authors":"Paulina Ibek, Alexander Degener, Magne Stridh, Marcus Tegnér, Patrik Edén, Victor Olariu, Malin Malmsjö, Aboma Merdasa","doi":"10.1364/BOE.592641","DOIUrl":"10.1364/BOE.592641","url":null,"abstract":"Basal cell carcinoma (BCC) is the most common form of skin cancer with the current diagnostic procedures relying on surgical excisions and histopathological analyses. Incomplete removal of the tumor necessitates additional surgery, which is both costly and time-consuming. To address this, non-invasive diagnostic techniques capable of determining the dimensions of skin cancer preoperatively are urgently needed. This work presents the development of an automated diagnostic pipeline for delineating BCC, using hyperspectral imaging (HSI), combined with machine learning (ML). HSI is a non-invasive, high-resolution imaging technique capable of differentiating tissue structures based on their molecular composition. ML is a branch of AI where algorithms learn patterns from data to make predictions without being explicitly programmed. The proposed pipeline is designed to adapt for individual patients, as universal skin cancer characteristic signals are challenging to define. The pipeline predicted tumor sizes with high accuracy, as validated by histopathological measurements in the 21 samples evaluated. The median absolute errors were low (0.25-0.30 mm), meaning that the predicted tumor dimensions were within the clinically accepted safety margins for radical excisions (≤0.4 mm). This work is the first to use unsupervised ML for tumor delineation in BCC using HSI, which enables personalized preoperative assessment and facilitates surgical precision.","PeriodicalId":8969,"journal":{"name":"Biomedical optics express","volume":"17 9","pages":"4916-4932"},"PeriodicalIF":3.2,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13592271/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149015331","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Multi-frequency asynchronous optical coherence elastography by direct demodulation. 直接解调的多频异步光学相干弹性成像。

IF 3.2 2区医学

Biomedical optics express Pub Date : 2026-09-01 DOI: 10.1364/BOE.607897

Ginger Schmidt, Ryan McAuley, Brett E Bouma, Néstor Uribe-Patarroyo

{"title":"Multi-frequency asynchronous optical coherence elastography by direct demodulation.","authors":"Ginger Schmidt, Ryan McAuley, Brett E Bouma, Néstor Uribe-Patarroyo","doi":"10.1364/BOE.607897","DOIUrl":"10.1364/BOE.607897","url":null,"abstract":"Optical coherence elastography has demonstrated promising results in highly specialized, research-grade OCT systems and in systems with custom, slow scan patterns. However, these approaches are fundamentally incompatible with commercial, FDA-approved OCT systems already deployed in clinics worldwide. Asynchronous optical coherence elastography addresses this limitation by using raster scanning-induced amplitude modulation to spatially encode harmonic shear wave displacement fields. This principle enables recovery of the coherent shear wave field within seconds using conventional frame-rate systems, despite severe temporal undersampling across B-scans. In this work, we extend this approach to show that asynchronous imaging can recover multiple simultaneous excitation frequencies, effectively multiplexing the determination of shear wave dispersion curves. We further introduce a new recovery framework, termed direct demodulation, based on the two-dimensional spatial Fourier transform of raster-scanned displacement data. We demonstrate experimentally that this framework enables efficient and intuitive isolation of up to 12 excitation frequencies within a single acquisition. We also identify practical limits on frequency multiplexing and provide guidelines and code for selecting imaging and excitation parameters that minimize spectral interference. The framework is validated through agreement between sequential and multiplexed measurements of shear wave number. Finally, we demonstrate the importance of multi-frequency elastography for robust mechanical characterization of complex tissues, showing sensitivity to boundary effects, viscoelastic dispersion, and improved mechanical contrast through incoherent frequency averaging in heterogeneous samples. This work enables multi-frequency, shear wave-based OCE in conventional frame-rate, raster-scanning OCT systems, further paving a path toward widespread adoption of elastography in existing imaging platforms.","PeriodicalId":8969,"journal":{"name":"Biomedical optics express","volume":"17 9","pages":"4947-4967"},"PeriodicalIF":3.2,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13592244/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149015358","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Accelerated immunostaining of thick biological tissues using bidirectional electric fields with reversible deformation. 用可逆变形的双向电场加速厚生物组织的免疫染色。

IF 3.2 2区医学

Biomedical optics express Pub Date : 2026-08-31 eCollection Date: 2026-09-01 DOI: 10.1364/BOE.613433

Ziwen Zhou, Xiao Xiao, Lijuan Du, Jiazhu Zhu, Zhenhong Du, Yao Zheng, Ke Si, Wei Gong

{"title":"Accelerated immunostaining of thick biological tissues using bidirectional electric fields with reversible deformation.","authors":"Ziwen Zhou, Xiao Xiao, Lijuan Du, Jiazhu Zhu, Zhenhong Du, Yao Zheng, Ke Si, Wei Gong","doi":"10.1364/BOE.613433","DOIUrl":"10.1364/BOE.613433","url":null,"abstract":"Immunostaining of thick biological tissues is often limited by slow antibody transport, resulting in long processing times and depth-dependent labeling. We describe an automated immunostaining approach based on an alternating bidirectional electric field, termed ARDEI, that improves labeling performance in millimeter-scale tissues. Under the applied electric-field conditions, tissue samples undergo reversible cyclic deformation, which may contribute to antibody redistribution by transiently altering tissue geometry. Using ARDEI, 1000 μm-thick mouse brain sections were labeled within 2.5 h, with improved signal penetration and spatial uniformity compared with passive diffusion and unidirectional electric-field staining under time-matched conditions. The system operates at 40 V under temperature-controlled conditions, minimizing thermal and structural perturbation. ARDEI supports three-dimensional imaging of neuronal and glial structures and can be applied to SHANEL-pretreated human brain tissue, where improved labeling uniformity was observed in 1000 μm-thick sections within 4 h. Under the tested conditions, repeated deformation cycles did not produce detectable structural distortion or signal degradation. These results show that ARDEI improves thick-tissue immunostaining performance and are consistent with a contribution from coupled electric-field-assisted transport and reversible tissue deformation. ARDEI provides a practical approach that can be integrated with existing optical imaging workflows for volumetric tissue analysis.","PeriodicalId":8969,"journal":{"name":"Biomedical optics express","volume":"17 9","pages":"4901-4915"},"PeriodicalIF":3.2,"publicationDate":"2026-08-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13592247/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149015560","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

OCT-guided angle-resolved low-coherence interferometry for microstructural analysis of the human retina. oct引导的角度分辨低相干干涉法用于人眼视网膜显微结构分析。

IF 3.2 2区医学

Biomedical optics express Pub Date : 2026-08-28 eCollection Date: 2026-09-01 DOI: 10.1364/BOE.605759

David A Miller, Hillel B Price, Wan Wang, Adam Wax

{"title":"OCT-guided angle-resolved low-coherence interferometry for microstructural analysis of the human retina.","authors":"David A Miller, Hillel B Price, Wan Wang, Adam Wax","doi":"10.1364/BOE.605759","DOIUrl":"10.1364/BOE.605759","url":null,"abstract":"Angle-resolved low-coherence interferometry (a/LCI) is an optical technique for extracting depth-resolved angular scattering profiles from biological tissues. Analysis of these angular scattering profiles can be used to compute the two-point spatial correlation function, which enables quantification of tissue structural heterogeneity at sub-cellular length scales. Building on previous work that combined a/LCI and optical coherence tomography (OCT) to detect biomarkers of Alzheimer's disease (AD) in ex vivo murine models, this study seeks to translate the platform for noninvasive scattering analysis of the human retina. The resulting system combines OCT image guidance, enabled by our high-performance, low-cost OCT platform, with parafoveal angular scattering measurements in each retinal quadrant, using 2D a/LCI. Measurements from 11 healthy participants were used to characterize retinal microstructure at length scales ranging from 4.8 μm to 9.5 μm. Our results revealed consistent variations in the structural correlations across retinal layers and quadrants, establishing baseline healthy measurements for future studies. This dual-modality approach offers a promising framework for improving diagnostic sensitivity in a variety of retinal and neurological disorders.","PeriodicalId":8969,"journal":{"name":"Biomedical optics express","volume":"17 9","pages":"4870-4883"},"PeriodicalIF":3.2,"publicationDate":"2026-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13592265/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149015405","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Phase-sensitive photothermal optical coherence tomography: decoupling optical and thermal tissue properties. 相敏光热光学相干层析成像：光学和热组织特性去耦。

IF 3.2 2区医学

Biomedical optics express Pub Date : 2026-08-28 eCollection Date: 2026-09-01 DOI: 10.1364/BOE.608642

Mohammadhossein Salimi, Amirhossein Khazaei, Nima Tabatabaei

{"title":"Phase-sensitive photothermal optical coherence tomography: decoupling optical and thermal tissue properties.","authors":"Mohammadhossein Salimi, Amirhossein Khazaei, Nima Tabatabaei","doi":"10.1364/BOE.608642","DOIUrl":"10.1364/BOE.608642","url":null,"abstract":"Photothermal optical coherence tomography (PT-OCT) has emerged as a promising extension of OCT for molecular contrast imaging, yet conventional amplitude-based PT-OCT is responsive, as opposed to being specific, to molecular contrast due to the multifactorial dependence of phase signals. Here, we present a phase-sensitive extension of PT-OCT that exploits PT phase to generate molecular contrast based on intrinsic thermal transport properties. We develop a theoretical model that links the PT-OCT phase response to the sample's thermal diffusivity and effusivity, and show that, in contrast to amplitude-based PT-OCT, the phase response is largely invariant to excitation intensity and optical attenuation. Systematic phantom experiments validate these predictions, demonstrating that while PT-OCT amplitude scales with absorption strength and excitation power, the phase signal remains independent of excitation conditions and imaging depth, instead encoding intrinsic thermal diffusivity. Using lipid- and water-rich phantoms, we further show that phase contrast provides additional separation between constituents with overlapping absorption spectra and reveals smooth interfacial transitions arising from effusivity mismatches. These results demonstrate that phase contrast can distinguish materials that are indistinguishable in conventional amplitude-based PT-OCT imaging. Together, these findings establish phase-sensitive PT-OCT as a practical extension of OCT for label-free imaging of tissue thermal properties, enabling high-contrast mapping governed by intrinsic thermophysical parameters rather than optical absorption alone. The approach has potential applications in cardiovascular diagnostics, oncology, and materials characterization, although translation to in vivo imaging will require improved phase stability and noise suppression.","PeriodicalId":8969,"journal":{"name":"Biomedical optics express","volume":"17 9","pages":"4884-4900"},"PeriodicalIF":3.2,"publicationDate":"2026-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13592260/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149015427","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Retinal fluid segmentation in real-world OCT imaging of neovascular AMD using hybrid deep learning and graph-based optimization. 使用混合深度学习和基于图的优化在新生血管性AMD的真实OCT成像中进行视网膜液分割。

IF 3.2 2区医学

Biomedical optics express Pub Date : 2026-08-26 eCollection Date: 2026-09-01 DOI: 10.1364/BOE.599076

Zhi Chen, Bernardo Bach, Honghai Zhang, Andreas Wahle, Ian C Han, H C Boldt, Stephen R Russell, Milan Sonka, Elliott H Sohn

{"title":"Retinal fluid segmentation in real-world OCT imaging of neovascular AMD using hybrid deep learning and graph-based optimization.","authors":"Zhi Chen, Bernardo Bach, Honghai Zhang, Andreas Wahle, Ian C Han, H C Boldt, Stephen R Russell, Milan Sonka, Elliott H Sohn","doi":"10.1364/BOE.599076","DOIUrl":"10.1364/BOE.599076","url":null,"abstract":"Accurate segmentation and quantification of retinal fluid in optical coherence tomography (OCT) images are important for assessing disease activity and treatment response in neovascular age-related macular degeneration (nvAMD). Yet, manual delineation of intraretinal fluid (IRF), subretinal fluid (SRF), and pigment epithelial detachment (PED) is labor-intensive and subject to inter-grader variability. We developed a hybrid framework for robust retinal fluid segmentation in clinical-grade longitudinal OCT, combining anatomically constrained retinal-region extraction using Deep LOGISMOS with nnU-Net-based 2-D and 3-D fluid segmentation. To improve adaptation to heterogeneous clinical data while reducing annotation burden, an iterative expert-guided labeling strategy was used, starting with 70 RETOUCH OCT scans and 100 scans from subjects with macular neovascularization at the University of Iowa and expanding to a final aggregated training set of 291 scans. The method was validated on a held-out test set of 50 OCT volumes from 50 nvAMD subjects, with a subset of 20 volumes independently annotated by a second expert for inter-observer analysis. It was further applied to a longitudinal dataset of 14,500 Heidelberg Spectralis macular OCT scans from 221 independent subjects. The ensemble model achieved Dice similarity coefficients (%) of 89.5±11.7 for IRF, 88.0±10.7 for SRF, and 82.0±13.0 for PED against the primary expert annotations, with performance approaching expert-level agreement. In longitudinal inference, our developed framework supports robust large-scale quantification of fluid burden and spatial extent across heterogeneous scan protocols and extended follow-up. The artificial intelligence-based approach has the potential to enable scalable and clinically meaningful longitudinal analysis of retinal fluid in nvAMD.","PeriodicalId":8969,"journal":{"name":"Biomedical optics express","volume":"17 9","pages":"4834-4853"},"PeriodicalIF":3.2,"publicationDate":"2026-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13592241/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149015523","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0