Proceedings of SPIE--the International Society for Optical Engineering最新文献

{"title":"Task-specific deep learning-based denoising for UHR cardiac PCD-CT adaptive to imaging conditions and patient characteristics: Impact on image quality and clinical diagnosis and quantitative assessment.","authors":"Madeleine Wilson, Shaojie Chang, Emily K Koons, Cynthia H McCollough, Shuai Leng","doi":"10.1117/12.3047283","DOIUrl":"https://doi.org/10.1117/12.3047283","url":null,"abstract":"<p><p>Ultra-high-resolution (UHR) photon-counting detector (PCD) CT offers superior spatial resolution compared to conventional CT, benefiting various clinical areas. However, the UHR resolution also significantly increases image noise, which can limit its clinical adoption in areas such as cardiac CT. In clinical practice, this image noise varies substantially across imaging conditions, such as different diagnostic tasks, patient characteristics (e.g., size), scan protocols, and image reconstruction settings. To address these challenges and provide the full potential of PCD-CT for optimal clinical performance, a convolutional neural network (CNN) denoising algorithm was developed, optimized, and tailored to each specific set of conditions. The algorithm's effectiveness in reducing noise and its impact on coronary artery stenosis quantification across different patient size categories (small: water equivalent diameter <300 mm, medium: 300-320 mm, and large: >320 mm) were objectively assessed. Reconstruction kernels at different sharpness, from Bv60 to Bv76, were investigated to determine optimal settings for each patient size regarding image quality and quantitative assessment of coronary stenosis (in terms of percent diameter stenosis). Our findings indicate that for patients with a water equivalent diameter less than 320 mm, CNN-denoised Bv72 images provide optimal image quality, less blooming artifact, and reduced percent diameter stenosis compared to routine images, while for patients with water equivalent diameter over 320 mm, CNN-denoised Bv60 images are preferable. Quantitatively, the CNN reduces noise-by 85% compared to the input images and 53% compared to commercial iterative reconstructions at strength 4 (QIR4)-while maintaining high spatial resolution and a natural noise texture. Moreover, it enhances stenosis quantification by reducing the percent diameter stenosis measurement by 52% relative to the input and 24% relative to QIR4. These improvements demonstrate the capability of CNN denoising in UHR PCD-CT to enhance image quality and quantitative assessment of coronary artery disease in a manner that is adaptive to patient characteristics and imaging conditions.</p>","PeriodicalId":74505,"journal":{"name":"Proceedings of SPIE--the International Society for Optical Engineering","volume":"13405 ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12076256/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144082624","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Vasculature-informed spatial smoothing of white matter functional magnetic resonance imaging.","authors":"Adam M Saunders, Michael E Kim, Kurt G Schilling, John C Gore, Bennett A Landman, Yurui Gao","doi":"10.1117/12.3047140","DOIUrl":"https://doi.org/10.1117/12.3047140","url":null,"abstract":"<p><p>Blood oxygenation level-dependent (BOLD) signals in white matter in the brain are anisotropically oriented, so that typical isotropic Gaussian spatial smoothing (GSS) of functional magnetic resonance images (fMRI) blurs across anatomical distributions. Abramian et al. developed a graph signal processing approach to smooth fMRI data along white matter fibers using diffusion MRI (diffusion-informed spatial smoothing, DSS). BOLD signals are modulated by the volume and oxygenation of blood carried by the vasculature, so we extend this method to provide vasculature-informed spatial smoothing (VSS). We collected susceptibility-weighted images and applied a Frangi filter to identify the peak vasculature direction in each voxel, alongside co-registered diffusion MRI and resting-state fMRI, weighting the VSS graph by the agreement of the vasculature directions aligned onto the graph's edges. We acquired resting-state fMRI at 7T using a repetition time of 1.5 seconds and 400 time points. Applying the DSS and VSS filters significantly increased the local functional connectivity measured using regional homogeneity (ReHo) compared to GSS (<i>p</i> < 0.01 using a paired <i>t</i>-test), but not when comparing DSS and VSS (<i>p</i> = 0.06). Independent component analysis resulted in less noisy components that agree better with labels from a white matter atlas with a significantly higher Dice score from the VSS filter compared to GSS (<i>p</i> < 0.05 using the Mann-Whitney U-test), and the VSS filter and DSS filter performed comparably (<i>p</i> = 0.06). In this pilot analysis, we find that fMRI data smoothed using VSS are comparable to results generated using DSS. The filtering code is available online (https://github.com/MASILab/vss_fmri).</p>","PeriodicalId":74505,"journal":{"name":"Proceedings of SPIE--the International Society for Optical Engineering","volume":"13406 ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12074660/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144082654","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Attention Rings for Shape Analysis and Application to MRI Quality Control.","authors":"Florian Davaux, Lucas Valladon, Lucie Dole, Jean Christophe Fillion, Beatriz Paniagua, Martin Styner, Juan Carlos Prieto","doi":"10.1117/12.3047233","DOIUrl":"10.1117/12.3047233","url":null,"abstract":"<p><p>The Adolescent Brain Cognitive Development (ABCD) Study collects extensive neuroimaging data, including over 20,000 MRI sessions, to understand brain development in children. Ensuring high-quality MRI data is essential for accurate analysis, but manual Quality Control (QC) is impractical for large datasets due to time and resource constraints. We propose an automated QC method using an ensemble model that leverages metrics from FSQC and a novel deep learning model for brain shape analysis that uses cortical thickness, curvature, sulcal depth, and surface area as input features. The ensemble model achieved an accuracy of 76%, while our method achieved an accuracy of 72.62%, with balanced precision, recall, and F1 scores for both classes. This automated method promises to improve QC processes and accelerate the analysis of ABCD data.</p>","PeriodicalId":74505,"journal":{"name":"Proceedings of SPIE--the International Society for Optical Engineering","volume":"13410 ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12096335/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144129840","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Design of Multiplexed, Live Cell Imaging Experiments Using Excitation Scan-Based Hyperspectral Imaging Microscopy.","authors":"Naga Annamdevula, Rebecca Tang-Holmes, Robert LeDoux, Taylor Jackson, Peyton Baker, Andrea L Britain, Thomas C Rich, Silas J Leavesley","doi":"10.1117/12.3042349","DOIUrl":"https://doi.org/10.1117/12.3042349","url":null,"abstract":"<p><p>In the last 20 years there have been remarkable advances in our ability to track movement and activities of proteins within cells. This is largely due to improved chemical probes and fluorescent proteins, and technical advances in microscopy. A remaining challenge is real-time multiplexed imaging. Excitation scan-based hyperspectral imaging (HSI) approaches are well suited for multiplexed imaging. However, excitation scan-based HSI has not been widely adopted, in part due to a lack of protocols for selection of combinations of fluorescent labels and proteins, and determining the range of excitation wavelengths and dichroic filters. Here we address this issue by outlining considerations for the selection of multiple labels for excitation scan-based HSI. HEK-293 cells were transfected with fluorescent protein constructs and/or loaded with dyes or labels for measurement of excitation spectra. Cells were imaged using a custom-built excitation scan-based HSI microscope that utilizes tunable thin film filters to filter fluorescence excitation from 360 nm to 550 nm in 5 nm increments in conjunction with a long pass dichroic filter and long pass emission filter. We observed that we can effectively quantify the relative abundance and spatial distributions of NucBlue, AlexaFluor 488, AlexaFluor 514, and AlexaFluor 555, Cal520, Cal590, as well as the fluorescent proteins GFP, Cerulean, Turquoise, Venus, tdTomato, and mCherry, individually and in combinations. We are currently assessing the spectra of these fluorophores using excitation scan-based HSI microscope systems.</p>","PeriodicalId":74505,"journal":{"name":"Proceedings of SPIE--the International Society for Optical Engineering","volume":"13323 ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12023988/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144026057","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A naturally brighter approach to colorectal cancer detection.","authors":"R Tang-Holmes, J Bond, N Annamdevula, M Verde, D Chakroborty, M Schuler, T C Rich, N Gong, C Sarkar, S J Leavesley","doi":"10.1117/12.3042063","DOIUrl":"https://doi.org/10.1117/12.3042063","url":null,"abstract":"<p><p>Colorectal cancer (CRC) is the second leading cause of cancer-related deaths worldwide. The gold standard for diagnosis is tissue biopsy during colonoscopy and subsequent histopathology. Limitations of current techniques include the turnaround time required for histopathology and the limited ability to detect flat lesions due to inadequate contrast provided by traditional white light endoscopy (WLE). The focus of this work was to assess detection accuracy for differentiating CRC and noncancerous tissues using excitation-scanning hyperspectral imaging (Ex-HSI) of autofluorescence compared to current diagnostic methods. Fluorescence Ex-HSI permits detection of all emitted light above a cut-off wavelength. Ex-HSI has been shown to reduce acquisition time, improve signal-to-noise ratio, and increase spectral information compared to emission-scanning HSI. This study utilized a mouse CRC model in which Azoxymethane/Dextran sodium sulfate (AOM/DSS) treatments induced colitis with subsequent nodule formation. Ex-HSI images were validated using transmitted light images, confocal \"z-stack\" images, and histology sectioning with H&E staining. Ex-HSI images were corrected to a flat spectral response, and excitation spectra were extracted from selected regions within each field of view (FOV). Inflammation and rectal bleeding were observed in the initial 31-day timepoint consistent with the AOM/DSS treatment. Colorectal nodules were visible using 4x and 20x magnification objectives and confocal \"z-stack\" imaging. Extracted spectra displayed two to several peak excitation wavelengths, likely indicating the presence of multiple autofluorescent molecules. Further investigation will utilize principal component analysis (PCA) and convolutional neural networks (CNN) to assess detection performance.</p>","PeriodicalId":74505,"journal":{"name":"Proceedings of SPIE--the International Society for Optical Engineering","volume":"13323 ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11996042/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144030779","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Errors Induced by Partial Pathlength Variations Due to Mammographic Compression in Dynamic Diffuse Optical Breast Imaging.","authors":"Kaiser Niknam, Mannu Bardhan Paul, Anthony Donaldson, Mini Das","doi":"10.1117/12.3048012","DOIUrl":"10.1117/12.3048012","url":null,"abstract":"<p><p>Optical imaging methods have the potential to overcome many of the drawbacks posed by current breast imaging modalities. Previous studies have found that mammographic compression induces different hemodynamic effects in cancerous and healthy breast tissue. This effect could be exploited in continuous-wave near-infrared spectroscopic imaging (CW-NIRS) for fast and accurate breast cancer screening. The primary issue with this approach is that breast tissue (and the cancerous mass) is displaced during the compression process, potentially introducing a considerable amount of errors and noise into the NIRS measurements with the current simple models used in estimating blood volume (and/or oxy/deoxy Hb) concentrations. In this work, we examine how these errors change with signal depth with breast compression and investigate methods to correct these based on simulations and experiments.</p>","PeriodicalId":74505,"journal":{"name":"Proceedings of SPIE--the International Society for Optical Engineering","volume":"13314 ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11956789/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143766031","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Probing bacterial membranes with polarization-resolved second harmonic scattering.","authors":"Eleanor F Page, Marea J Blake, Tessa R Calhoun","doi":"10.1117/12.3028197","DOIUrl":"10.1117/12.3028197","url":null,"abstract":"<p><p>For antibiotics that target Gram-positive bacterial cell structures, optimizing their interaction with the cytoplasmic membrane is of paramount importance. Recent time-resolved second harmonic scattering (trSHS) experiments with living bacterial cells have shown that some amphiphilic small molecules display signals consistent with organization within the membrane environment. Such organization could arise, for example, from aggregation, solvent interactions, and/or environmental rigidity. To expand our study of this system, we turn to polarization-resolved SHS (pSHS). PSHS has previously been used with model membranes to extract information about the angular distribution of integrated small molecules. Here we apply pSHS, for the first time, to cells, specifically living <i>Staphylococcus aureus</i>. In doing so, we aim to address contributions ascribed to the organization of amphiphilic molecules in bacterial membranes.</p>","PeriodicalId":74505,"journal":{"name":"Proceedings of SPIE--the International Society for Optical Engineering","volume":"13139 ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12070223/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144044343","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Exploring advanced 2D acquisitions in breast tomosynthesis: T-shaped and Pentagon geometries.","authors":"Priyash Singh, Chloe J Choi, Bruno Barufaldi, Andrew D A Maidment, Raymond J Acciavatti","doi":"10.1117/12.3027054","DOIUrl":"10.1117/12.3027054","url":null,"abstract":"<p><p>In this study, we investigate the performance of advanced 2D acquisition geometries - Pentagon and T-shaped - in digital breast tomosynthesis (DBT) and compare them against the conventional 1D geometry. Unlike the conventional approach, our proposed 2D geometries also incorporate anterior projections away from the chest wall. Implemented on the Next-Generation Tomosynthesis (NGT) prototype developed by X-ray Physics Lab (XPL), UPenn, we utilized various phantoms to compare three geometries: a Defrise slab phantom with alternating plastic slabs to study low-frequency modulation; a Checkerboard breast phantom (a 2D adaptation of the Defrise phantom design) to study the ability to reconstruct the fine features of the checkerboard squares; and the 360° Star-pattern phantom to assess aliasing and compute the Fourier-spectral distortion (FSD) metric that assesses spectral leakage and the contrast transfer function. We find that both Pentagon and T-shaped scans provide greater modulation amplitude of the Defrise phantom slabs and better resolve the squares of the Checkerboard phantom against the conventional scan. Notably, the Pentagon geometry exhibited a significant reduction in aliasing of spatial frequencies oriented in the right-left (RL) medio-lateral direction, which was corroborated by a near complete elimination of spectral leakage in the FSD plot. Conversely T-shaped scan redistributes the aliasing between both posteroanterior (PA) and RL directions thus maintaining non-inferiority against the conventional scan which is predominantly affected by PA aliasing. The results of this study underscore the potential of incorporating advanced 2D geometries in DBT systems, offering marked improvements in imaging performance over the conventional 1D approach.</p>","PeriodicalId":74505,"journal":{"name":"Proceedings of SPIE--the International Society for Optical Engineering","volume":"13174 ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11236198/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141592639","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A general approach to improve adversarial robustness of DNNs for medical image segmentation and detection.","authors":"Linhai Ma, Jiasong Chen, Linchen Qian, Liang Liang","doi":"10.1117/12.3006534","DOIUrl":"10.1117/12.3006534","url":null,"abstract":"<p><p>It is known that deep neural networks (DNNs) are vulnerable to adversarial noises. Improving adversarial robustness of DNNs is essential. This is not only because unperceivable adversarial noise is a threat to the performance of DNNs models, but also adversarially robust DNNs have a strong resistance to the white noises that may present everywhere in the actual world. To improve adversarial robustness of DNNs, a variety of adversarial training methods have been proposed. Most of the previous methods are designed under one single application scenario: image classification. However, image segmentation, landmark detection, and object detection are more commonly observed than classifying the entire images in the medical imaging field. Although classification tasks and other tasks (e.g., regression) share some similarities, they also differ in certain ways, e.g., some adversarial training methods use misclassification criteria, which is well-defined in classification but not in regression. These restrictions/limitations hinder application of adversarial training for many medical imaging analysis tasks. In our work, the contributions are as follows: (1) We investigated the existing adversarial training methods and discovered the challenges that make those methods unsuitable for adaptation in segmentation and detection tasks. (2) We modified and adapted some existing adversarial training methods for medical image segmentation and detection tasks. (3) We proposed a general adversarial training method for medical image segmentation and detection. (4) We implemented our method in diverse medical imaging tasks using publicly available datasets, including MRI segmentation, Cephalometric landmark detection, and blood cell detection. The experiments substantiated the effectiveness of our method.</p>","PeriodicalId":74505,"journal":{"name":"Proceedings of SPIE--the International Society for Optical Engineering","volume":"12926 ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11491114/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142482678","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"CDPNet: a radiomic feature learning method with epigenetic application to estimating MGMT promoter methylation status in glioblastoma.","authors":"Jun Guo, Fanyang Yu, MacLean P Nasrallah, Christos Davatzikos","doi":"10.1117/12.3009716","DOIUrl":"https://doi.org/10.1117/12.3009716","url":null,"abstract":"<p><p>Radiomics has been widely recognized for its effectiveness in decoding tumor phenotypes through the extraction of quantitative imaging features. However, the robustness of radiomic methods to estimate clinically relevant biomarkers non-invasively remains largely untested. In this study, we propose Cascaded Data Processing Network (CDPNet), a radiomic feature learning method to predict tumor molecular status from medical images. We apply CDPNet to an epigenetic case, specifically targeting the estimation of <i>O6-methylguanine-DNA-methyltransferase</i> (<i>MGMT</i>) promoter methylation from Magnetic Resonance Imaging (MRI) scans of glioblastoma patients. CDPNet has three components: 1) Principal Component Analysis (PCA), 2) Fisher Linear Discriminant (FLD), and 3) a combination of hashing and blockwise histograms. The outlined architectural framework capitalizes on PCA to reconstruct input image patches, followed by FLD to extract discriminative filter banks, and finally using binary hashing and blockwise histogram module for indexing, pooling, and feature generation. To validate the effectiveness of CDPNet, we conducted an exhaustive evaluation on a comprehensive retrospective cohort comprising 484 <i>IDH</i>-wildtype glioblastoma patients with pre-operative multi-parametric MRI scans (T1, T1-Gd, T2, and T2-FLAIR). The prediction of <i>MGMT</i> promoter methylation status was cast as a binary classification problem. The developed model underwent rigorous training via 10-fold cross-validation on a discovery cohort of 446 patients. Subsequently, the model's performance was evaluated on a distinct and previously unseen replication cohort of 38 patients. Our method achieved an accuracy of 70.11% and an area under the curve of 0.71 (95% CI: 0.65 - 0.74).</p>","PeriodicalId":74505,"journal":{"name":"Proceedings of SPIE--the International Society for Optical Engineering","volume":"12930 ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11034757/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140859305","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}