Journal of Medical Imaging最新文献

{"title":"Synthesizing vocal tract magnetic resonance imaging sequences with phoneme-aware diffusion models.","authors":"Paula Andrea Perez-Toro, Tomas Arias-Vergara, Lukas Buess, Jana Hutter, Jonghye Woo, Andreas Maier","doi":"10.1117/1.JMI.13.S1.S11206","DOIUrl":"https://doi.org/10.1117/1.JMI.13.S1.S11206","url":null,"abstract":"<p><strong>Purpose: </strong>Real-time speech MRI offers critical insights into articulatory dynamics for diagnostics, therapy, and speech science, but direct speech-to-MRI mapping remains highly challenging. We present a diffusion-based framework that transforms spoken language into dynamic vocal tract MRI sequences with phonemic-aware attention. This improves phoneme-level fidelity and introduces the speech-to-image phonemic fidelity score (SIPFS), a new metric to assess articulatory precision.</p><p><strong>Approach: </strong>We trained audio-conditioned diffusion models with phonemic-aware attention and contrastive learning on the USC 75-Speaker Speech MRI Database. Input consisted of sagittal-view MRI sequences synchronized with speech. Three pre-trained audio encoders (Wav2Vec, WavLM, and Whisper) were used to provide embeddings. Performance was evaluated on unseen speakers and unseen speech segments using standard image/video metrics (FID, FVD, SSIM, and PSNR) and the proposed SIPFS.</p><p><strong>Results: </strong>Phonemic-aware attention consistently improved quality and phoneme-level precision. SemDiff produced high-quality static images, whereas SD3D excelled in dynamic video synthesis. SIPFS revealed up to 15% higher phonemic accuracy compared with unconditioned models. Excluding articulatory-region ROI reduced fidelity of vocal tract motion, highlighting its importance. The models generalized across new speakers and missing segments, demonstrating robustness.</p><p><strong>Conclusions: </strong>Our framework demonstrates the feasibility of synthesizing accurate, phoneme-sensitive vocal tract MRI from speech using diffusion models. By improving image fidelity and introducing SIPFS, this approach supports clinical applications in diagnostics, speech therapy, and articulatory research. Future work will focus on expanding datasets, optimizing computational efficiency, and refining evaluation metrics to further enhance clinical integration.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"13 Suppl 1","pages":"S11206"},"PeriodicalIF":1.7,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13186585/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147983838","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":"Conditional generative diffusion model for 3D trabecular bone synthesis with tunable microstructure.","authors":"Xin Wang, Gengxin Shi, Peiqin Teng, Aswath Sivakumar, Tianyi Ye, Adam D Sylvester, J Webster Stayman, Wojciech B Zbijewski","doi":"10.1117/1.JMI.13.S1.S11204","DOIUrl":"10.1117/1.JMI.13.S1.S11204","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Purpose: &lt;/strong&gt;We aim to develop a conditional generative diffusion model capable of producing three-dimensional (3D) trabecular bone samples that can be tuned to achieve specific structural characteristics prescribed in terms of three geometric metrics of trabecular microarchitecture: bone volume fraction (BV/TV), trabecular thickness (Tb.Th), and spacing (Tb.Sp).&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Approach: &lt;/strong&gt;The generative model is based on 3D latent diffusion. The latent representation of trabecular patches is obtained by a dedicated variational autoencoder (VAE). To control the microstructure characteristics of the synthetic samples, the model is conditioned on BV/TV, Tb.Th, and Tb.Sp. In addition, a shifting slab inference method is employed to generate extended volumes with locally tunable microstructure in a computationally efficient manner. The training data involved 3551 &lt;math&gt;&lt;mrow&gt;&lt;mn&gt;128&lt;/mn&gt; &lt;mo&gt;×&lt;/mo&gt; &lt;mn&gt;128&lt;/mn&gt; &lt;mo&gt;×&lt;/mo&gt; &lt;mn&gt;128&lt;/mn&gt;&lt;/mrow&gt; &lt;/math&gt; volumes of interest (VOIs) extracted from micro-CT volumes ( &lt;math&gt;&lt;mrow&gt;&lt;mn&gt;50&lt;/mn&gt; &lt;mtext&gt;  &lt;/mtext&gt; &lt;mi&gt;μ&lt;/mi&gt; &lt;mi&gt;m&lt;/mi&gt;&lt;/mrow&gt; &lt;/math&gt; voxel size) of 20 femoral bone specimens, paired with trabecular metrics computed within each VOI; the split for training and validation data was 9:1. For testing, 2000 synthetic bone samples were generated using single slab inference over a wide range of condition (target) microstructure metrics. Results were evaluated in terms of (i) consistency across multiple realizations of reverse diffusion for a fixed condition, measured by the coefficient of variation (CV) of trabecular measurements; (ii) agreement between BV/TV, Tb.Th, and Tb.Sp values provided as a condition and those measured in the corresponding synthetic samples, assessed using Pearson correlation coefficient (PCC); and (iii) overlap between the distributions of trabecular parameters of real and synthetic bone patches; this coverage analysis included both the conditioning parameters of BV/TV, Tb.Th, and Tb.Sp, as well as unconditioned metrics of degree of anisotropy, ellipsoid factor, and connectivity. Further, extended volumes ( &lt;math&gt;&lt;mrow&gt;&lt;mn&gt;128&lt;/mn&gt; &lt;mo&gt;×&lt;/mo&gt; &lt;mn&gt;128&lt;/mn&gt; &lt;mo&gt;×&lt;/mo&gt; &lt;mn&gt;256&lt;/mn&gt; &lt;mrow&gt;&lt;mtext&gt;  &lt;/mtext&gt;&lt;/mrow&gt; &lt;mrow&gt;&lt;mtext&gt;voxels&lt;/mtext&gt;&lt;/mrow&gt; &lt;/mrow&gt; &lt;/math&gt; ) were generated using shifting-slab inference with spatially invariant and spatially varying conditioning and evaluated in terms of local agreement between the prescribed and achieved trabecular parameters.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;Visually, the synthesized cancellous bone patches appear highly similar to the training micro-CT data. The conditioned parameters of the generated volumes agree well with their target values (PCC of 0.99, 0.97, and 0.95 for BV/TV, Tb.Th, and Tb.Sp, respectively). There is a trend toward generating trabeculae that are slightly thicker than prescribed, but this bias is typically on the order of one voxel ( &lt;math&gt;&lt;mrow&gt;&lt;mn&gt;50&lt;/mn&gt; &lt;mtext&gt;  &lt;/mtext&gt; &lt;mi&gt;μ&lt;/mi&gt; &lt;mi&gt;m&lt;/mi&gt;&lt;/mr","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"13 Suppl 1","pages":"S11204"},"PeriodicalIF":2.3,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12907505/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146214560","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":"Challenges to the management of oncologic theranostics clinical trials: recommendations for the conduct of theranostics trials at investigational sites.","authors":"Nicholas P Gruszauskas, Joseph Steiner, Krista Dillingham","doi":"10.1117/1.JMI.13.1.013502","DOIUrl":"10.1117/1.JMI.13.1.013502","url":null,"abstract":"<p><strong>Purpose: </strong>Advancements in radionuclide imaging and therapy techniques have created a groundswell of enthusiasm in the recently designated field of theranostics. This has increased the need for facilities that are able to participate in clinical trials for investigational theranostic agents. Theranostics clinical trials present several unique challenges that will tax the resources and staff of most medical centers. Our purpose is to describe the unique logistical and administrative challenges associated with theranostics clinical trials, propose strategies for addressing them, and make recommendations regarding trial conduct to the community at large.</p><p><strong>Approach: </strong>The authors' experiences reviewing, implementing, and managing theranostics trials at their institution were used to identify common activities and challenges.</p><p><strong>Results: </strong>Several key categories of requirements and challenges were identified. Multidisciplinary teams consisting of nuclear medicine, oncology, nursing, clinical research, and administrative staff are necessary to adequately perform all trial-related activities. Strategies are proposed to address these challenges and activities at the institutional and industry levels.</p><p><strong>Conclusion: </strong>The unique challenges inherent to theranostics clinical trials require a focused investment of time, effort, and resources from all stakeholders. Institutions that wish to participate in these trials must develop the infrastructure necessary to fully support the breadth of activities they require. Implementation of the strategies and recommendations presented here will ensure the successful conduct of these trials and will improve efficiency across the community.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"13 1","pages":"013502"},"PeriodicalIF":1.7,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12928531/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147285697","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":"Improving Clarity and Impact in JMI.","authors":"Bennett A Landman","doi":"10.1117/1.JMI.13.1.010101","DOIUrl":"https://doi.org/10.1117/1.JMI.13.1.010101","url":null,"abstract":"<p><p>JMI Editor-in-Chief Bennett Landman offers guidance to help authors achieve higher impact, clearer assessment of contributions, and more useful and direct reviews from the peer review community.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"13 1","pages":"010101"},"PeriodicalIF":1.7,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12946671/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147327812","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":"LiteMIL: a computationally efficient cross-attention multiple instance learning for cancer subtyping on whole-slide images.","authors":"Haitham Kussaibi","doi":"10.1117/1.JMI.13.1.017501","DOIUrl":"https://doi.org/10.1117/1.JMI.13.1.017501","url":null,"abstract":"<p><strong>Purpose: </strong>Accurate cancer subtyping is essential for precision medicine but challenged by the computational demands of gigapixel whole-slide images (WSIs). Although transformer-based multiple instance learning (MIL) methods achieve strong performance, their quadratic complexity limits clinical deployment. We introduce LiteMIL, a computationally efficient cross-attention MIL, optimized for WSIs classification.</p><p><strong>Approach: </strong>LiteMIL employs a single learnable query with multi-head cross-attention for bag-level aggregation from extracted features. We evaluated LiteMIL against five baselines (mean/max pooling, ABMIL, MAD-MIL, and TransMIL) on four TCGA datasets (breast: <math><mrow><mi>n</mi> <mo>=</mo> <mn>875</mn></mrow> </math> , kidney: <math><mrow><mi>n</mi> <mo>=</mo> <mn>906</mn></mrow> </math> , lung: <math><mrow><mi>n</mi> <mo>=</mo> <mn>958</mn></mrow> </math> , and TUPAC16: <math><mrow><mi>n</mi> <mo>=</mo> <mn>821</mn></mrow> </math> ) using nested cross-validation with patient-level splitting. Systematic ablation studies evaluated multi-query variants, attention heads, dropout rates, and architectural components.</p><p><strong>Results: </strong>LiteMIL achieved competitive accuracy (average 83.5%), matching TransMIL, while offering substantial efficiency gains: <math><mrow><mn>4.8</mn> <mo>×</mo></mrow> </math> fewer parameters (560K versus 2.67M), <math><mrow><mn>2.9</mn> <mo>×</mo></mrow> </math> faster inference (1.6s versus 4.6s per fold), and <math><mrow><mn>6.7</mn> <mo>×</mo></mrow> </math> lower Graphics Processing Unit (GPU) memory usage (1.15 GB versus 7.77 GB). LiteMIL excelled on lung (86.3% versus 85.0%), TUPAC16 (72% versus 71.4%), and matched kidney performance (89.9% versus 89.7%). Ablation studies revealed task-dependent multi-query performance benefits: <math><mrow><mi>Q</mi> <mo>=</mo> <mn>4</mn></mrow> </math> versus <math><mrow><mi>Q</mi> <mo>=</mo> <mn>1</mn></mrow> </math> improved morphologically heterogeneous tasks (breast/lung <math><mrow><mo>+</mo> <mn>1.3</mn> <mo>%</mo></mrow> </math> each, <math><mrow><mi>p</mi> <mo><</mo> <mn>0.05</mn></mrow> </math> ) but degraded on grading tasks (TUPAC16: <math><mrow><mo>-</mo> <mn>1.6</mn> <mo>%</mo></mrow> </math> ), validating single-query optimality for focused attention scenarios.</p><p><strong>Conclusions: </strong>LiteMIL provides a resource-efficient solution for WSI classification. The cross-attention architecture matches complex transformer performance while enabling deployment on consumer GPUs. Task-dependent design insights, single query for sparse discriminating features, multi-query for heterogeneous patterns, guide practical implementation. The architecture's efficiency, combined with compact features, makes LiteMIL suitable for clinical integration in settings with limited computational infrastructure.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"13 1","pages":"017501"},"PeriodicalIF":1.7,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12772520/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145918838","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":"Asynchronous federated learning for web-based OCT image analysis.","authors":"Hasan Md Tusfiqur Alam, Tim Maurer, Abdulrahman Mohamed Selim, Matthias Eiletz, Michael Barz, Daniel Sonntag","doi":"10.1117/1.JMI.13.1.014501","DOIUrl":"https://doi.org/10.1117/1.JMI.13.1.014501","url":null,"abstract":"<p><strong>Purpose: </strong>Centralized machine learning often struggles with limited data access and expert involvement. We investigate decentralized approaches that preserve data privacy while enabling collaborative model training for medical imaging tasks.</p><p><strong>Approach: </strong>We explore asynchronous federated learning (FL) using the FL with buffered asynchronous aggregation (FedBuff) algorithm for classifying optical coherence tomography (OCT) retina images. Unlike synchronous algorithms such as FedAvg, which require all clients to participate simultaneously, FedBuff supports independent client updates. We compare its performance to both centralized models and FedAvg. In addition, we develop a browser-based proof-of-concept system using modern web technologies to assess the feasibility and limitations of interactive, collaborative learning in real-world settings.</p><p><strong>Results: </strong>FedBuff performs well in binary OCT classification tasks but shows reduced accuracy in more complex, multiclass scenarios. FedAvg achieves results comparable to centralized training, consistent with previous findings. Although FedBuff underperforms compared with FedAvg and centralized models, it still delivers acceptable accuracy in less complex settings. The browser-based prototype demonstrates the potential for accessible, user-driven FL systems but also highlights technical limitations in current web standards, especially regarding local computation and communication efficiency.</p><p><strong>Conclusion: </strong>Asynchronous FL via FedBuff offers a promising, privacy-preserving approach for medical image classification, particularly when synchronous participation is impractical. However, its scalability to complex classification tasks remains limited. Web-based implementations have the potential to broaden access to collaborative AI tools, but limitations of the current technologies need to be further investigated.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"13 1","pages":"014501"},"PeriodicalIF":1.7,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12767620/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145913135","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":"Characterizing the effects of noncontrast head CT reconstruction kernel and slice thickness parameters on the performance of an automated AI algorithm in the evaluation of ischemic stroke.","authors":"Spencer H Welland, Grace Hyun J Kim, Anil Yadav, Kambiz Nael, John M Hoffman, Matthew S Brown, Michael F McNitt-Gray, William Hsu","doi":"10.1117/1.JMI.13.1.014503","DOIUrl":"10.1117/1.JMI.13.1.014503","url":null,"abstract":"<p><strong>Purpose: </strong>There are multiple commercially available, Food and Drug Administration (FDA)-cleared, artificial intelligence (AI)-based tools automating stroke evaluation in noncontrast computed tomography (NCCT). This study assessed the impact of variations in reconstruction kernel and slice thickness on two outputs of such a system: hypodense volume and Alberta Stroke Program Early CT Score (ASPECTS).</p><p><strong>Approach: </strong>The NCCT series image data of 67 patients imaged with a CT stroke protocol were reconstructed with four kernels (H10s-smooth, H40s-medium, H60s-sharp, and H70h-very sharp) and three slice thicknesses (1.5, 3.0, and 5.0 mm) to create 1 reference condition (H40s/5.0 mm) and 11 nonreference conditions. The 12 reconstructions per patient were processed with a commercially available FDA-cleared software package that yields total hypodense volume (mL) and ASPECTS. A mixed-effect model was used to test the difference in hypodense volume, and an ordered logistic model was used to test the difference in e-ASPECTS.</p><p><strong>Results: </strong>Hypodense volume differences from the reference condition ranged from <math><mrow><mo>-</mo> <mn>14.6</mn></mrow> </math> to 1.1 mL and were significant for all nonreference kernels (H10s <math><mrow><mi>p</mi> <mo>=</mo> <mn>0.025</mn></mrow> </math> , H60s <math><mrow><mi>p</mi> <mo><</mo> <mn>0.001</mn></mrow> </math> , and H70h <math><mrow><mi>p</mi> <mo><</mo> <mn>0.001</mn></mrow> </math> ) and for thinner slices (1.5 mm <math><mrow><mi>p</mi> <mo><</mo> <mn>0.001</mn></mrow> </math> and 3.0 mm <math><mrow><mi>p</mi> <mo>=</mo> <mn>0.002</mn></mrow> </math> ). e-ASPECTS was invariant to the nonreference kernels and slice thicknesses, with a mean difference ranging from <math><mrow><mo>-</mo> <mn>0.1</mn></mrow> </math> to 0.5. No significant differences were found for any kernel or slice thickness (all <math><mrow><mi>p</mi> <mo>></mo> <mn>0.05</mn></mrow> </math> ).</p><p><strong>Conclusions: </strong>Automated hypodense volume measured with a commercially available, FDA-cleared software package is substantially impacted by reconstruction kernel and slice thickness. Conversely, automated ASPECTS is invariant to these reconstruction parameters.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"13 1","pages":"014503"},"PeriodicalIF":2.3,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12782953/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145953582","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":"Utility of the virtual imaging trials methodology for objective characterization of AI systems and training data.","authors":"Fakrul Islam Tushar, Lavsen Dahal, Saman Sotoudeh-Paima, Ehsan Abadi, William P Segars, Joseph Y Lo, Ehsan Samei","doi":"10.1117/1.JMI.13.1.014506","DOIUrl":"10.1117/1.JMI.13.1.014506","url":null,"abstract":"<p><strong>Purpose: </strong>The credibility of artificial intelligence (AI) models for medical imaging continues to be a challenge, affected by the diversity of models, the data used to train the models, and the applicability of their combination to produce reproducible results for new data. We aimed to explore whether emerging virtual imaging trial (VIT) methodologies can provide an objective resource to approach this challenge.</p><p><strong>Approach: </strong>We conducted this study for the case example of COVID-19 diagnosis using clinical and virtual computed tomography (CT) and chest radiography (CXR) processed with convolutional neural networks. Multiple AI models were developed and tested using 3D ResNet-like and 2D EfficientNetv2 architectures across diverse datasets.</p><p><strong>Results: </strong>Model performance was evaluated using the area under the curve (AUC) and the DeLong method for AUC confidence intervals. The models trained on the most diverse datasets showed the highest external testing performance, with AUC values ranging from 0.73 to 0.76 for CT and 0.70 to 0.73 for CXR. Internal testing yielded higher AUC values (0.77 to 0.85 for CT and 0.77 to 1.0 for CXR), highlighting a substantial drop in performance during external validation, which underscores the importance of diverse and comprehensive training and testing data. Most notably, the VIT approach provided an objective assessment of the utility of diverse models and datasets while offering insight into the influence of dataset characteristics, patient factors, and imaging physics on AI efficacy.</p><p><strong>Conclusions: </strong>The VIT approach enhances model transparency and reliability, offering nuanced insights into the factors driving AI performance and bridging the gap between experimental and clinical settings.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"13 1","pages":"014506"},"PeriodicalIF":2.3,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12906637/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146208146","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":"From preoperative computed tomography to postmastoidectomy mesh construction: mastoidectomy shape prediction for cochlear implant surgery.","authors":"Yike Zhang, Eduardo Davalos, Dingjie Su, Ange Lou, Jack Noble","doi":"10.1117/1.JMI.13.1.014004","DOIUrl":"10.1117/1.JMI.13.1.014004","url":null,"abstract":"<p><strong>Purpose: </strong>Cochlear implant (CI) surgery treats severe hearing loss by inserting an electrode array into the cochlea to stimulate the auditory nerve. An important step in this procedure is mastoidectomy, which removes part of the mastoid region of the temporal bone to provide surgical access. Accurate mastoidectomy shape prediction from preoperative imaging improves presurgical planning, reduces risks, and enhances surgical outcomes. Despite its importance, there are limited deep-learning-based studies regarding this topic due to the challenges of acquiring ground-truth labels. We address this gap by investigating self-supervised and weakly-supervised learning models to predict the mastoidectomy region without human annotations.</p><p><strong>Approach: </strong>We propose a hybrid self-supervised and weakly-supervised learning framework to predict the mastoidectomy region directly from preoperative CT scans, where the mastoid remains intact. Our self-supervised learning approach reconstructs the postmastoidectomy 3D surface from preoperative imaging, aiming to align with the corresponding intraoperative microscope views for future surgical navigation-related applications. Postoperative CT scans are used in the self-supervised learning model to assist training procedures despite additional challenges such as metal artifacts and low signal-to-noise ratios introduced by them. To further improve the accuracy and robustness, we introduce a Mamba-based weakly-supervised model that refines mastoidectomy shape prediction by using 3D T-distribution loss function, inspired by the student- <math><mrow><mi>t</mi></mrow> </math> distribution. Weak supervision is achieved by leveraging segmentation results from the prior self-supervised framework, eliminating the manual data labeling process.</p><p><strong>Results: </strong>Our hybrid method achieves a mean Dice score of 0.72 when predicting the complex and boundary-less mastoidectomy shape, surpassing state-of-the-art approaches and demonstrating strong performance. The method provides groundwork for constructing 3D postmastoidectomy surfaces directly from the corresponding preoperative CT scans.</p><p><strong>Conclusion: </strong>To our knowledge, this is the first work that integrates self-supervised and weakly-supervised learning for mastoidectomy shape prediction, offering a robust and efficient solution for CI surgical planning while leveraging 3D T-distribution loss in weakly-supervised medical imaging.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"13 1","pages":"014004"},"PeriodicalIF":2.3,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12838396/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146094516","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":"Deep learning-based mediastinal lymph node assessment on PET/CT images without pixel-level annotations.","authors":"Sofija Engelson, Yannic Elser, Malte Maria Sieren, Jan Ehrhardt, Julia Andresen, Stefanie Schierholz, Tobias Keck, Daniel Drömann, Jörg Barkhausen, Heinz Handels","doi":"10.1117/1.JMI.13.1.014507","DOIUrl":"https://doi.org/10.1117/1.JMI.13.1.014507","url":null,"abstract":"<p><strong>Purpose: </strong><math><mrow><mi>N</mi></mrow> </math> -staging, a critical component in cancer diagnostics, quantifies metastatic involvement of lymph nodes and plays an important role in guiding treatment decisions. Manual assessment of lymph nodes on PET/CT scans is time-consuming due to minimal contrast to surrounding tissue and strong heterogeneity of the lymph node's morphology. To streamline the <math><mrow><mi>N</mi></mrow> </math> -staging process, we propose a deep learning-based algorithm that localizes lymph node stations through atlas-to-patient registration, classifies mediastinal lymph node stations as malignant or benign, and subsequently performs automated <math><mrow><mi>N</mi></mrow> </math> -staging. Notably, our model is trained without any pixel-level annotations, i.e., using image-level classification labels only.</p><p><strong>Approach: </strong>To address the challenge of training without annotations at the pixel level, we use prior knowledge of the lymph node station locations through atlas-to-patient registration and deduce pseudo-labels for lymph node station groups from the <math><mrow><mi>N</mi></mrow> </math> -stage to enable weakly supervised network training.</p><p><strong>Results: </strong>The proposed algorithm achieves an accuracy of <math><mrow><mn>0.88</mn> <mo>±</mo> <mn>0.02</mn></mrow> </math> , a sensitivity of <math><mrow><mn>0.72</mn> <mo>±</mo> <mn>0.08</mn></mrow> </math> , and a specificity of <math><mrow><mn>0.90</mn> <mo>±</mo> <mn>0.03</mn></mrow> </math> for lymph node station classification, which is significantly better than the performance of the standard threshold-based approach used for lymph node assessment in radiological images and an algorithm for PET lesion segmentation that was trained with segmentation masks. For automatic <math><mrow><mi>N</mi></mrow> </math> -staging, the accuracy of <math><mrow><mn>0.63</mn> <mo>±</mo> <mn>0.04</mn></mrow> </math> is on par with an algorithm that was trained with segmentation masks.</p><p><strong>Conclusions: </strong>The division of the problem setting into subtasks as well as the integration of prior knowledge enables better or comparable performance of models trained with and without segmentation masks.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"13 1","pages":"014507"},"PeriodicalIF":1.7,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12915757/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146229382","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}