Journal of Medical Imaging最新文献

{"title":"Frequency-based texture analysis of non-Gaussian properties of digital breast tomosynthesis images and comparison across two vendors.","authors":"Kai Yang, Craig K Abbey, Bruno Barufaldi, Xinhua Li, Theodore A Marschall, Bob Liu","doi":"10.1117/1.JMI.12.S2.S22004","DOIUrl":"10.1117/1.JMI.12.S2.S22004","url":null,"abstract":"<p><strong>Purpose: </strong>We aim to analyze higher-order textural components of digital breast tomosynthesis (DBT) images to quantify differences in the appearance of breast parenchyma produced by different vendors.</p><p><strong>Approach: </strong>We included consecutive women who had normal screening DBT exams in January 2018 from a GE system and in adjacent years from Hologic systems. Laplacian fractional entropy (LFE), as a measure of non-Gaussian statistical properties of breast tissue texture, was calculated from for-presentation Craniocaudal (CC) view DBT slices and synthetic mammograms (SMs) through frequency-based filtering with Gabor filters, which were considered mathematical models for human visual response to image textures. The LFE values were compared within and across subjects and vendors along with secondary parameters (laterality, year-to-year, modality, and breast density) via two-way analysis of variance (ANOVA) tests using frequency as one of the two independent variables, and a <math><mrow><mi>P</mi></mrow> </math> -value <math><mrow><mo><</mo> <mn>0.05</mn></mrow> </math> was considered statistically significant.</p><p><strong>Results: </strong>A total of 8529 CC view DBT slices and SM images from 73 screening exams in 25 women were analyzed. Significant differences in LFE were observed for different frequencies ( <math><mrow><mi>P</mi> <mo><</mo> <mn>0.001</mn></mrow> </math> ) and across vendors (GE versus Hologic DBT: <math><mrow><mi>P</mi> <mo><</mo> <mn>0.001</mn></mrow> </math> , GE versus Hologic SM: <math><mrow><mi>P</mi> <mo><</mo> <mn>0.001</mn></mrow> </math> ).</p><p><strong>Conclusion: </strong>Significant differences in perception of breast parenchyma textures among two DBT vendors were demonstrated via higher-order non-Gaussian statistical properties. This finding extends previously observed differences in anatomical noise power spectra in DBT images and provides quantitative evidence to support caution in across-vendor comparative reading and will be beneficial to facilitate future development of vendor-neutral artificial intelligence algorithms for breast cancer screening.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"12 Suppl 2","pages":"S22004"},"PeriodicalIF":1.9,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11925074/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143694062","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":"Investigating the effect of adding comparisons with prior mammograms to standalone digital breast tomosynthesis screening.","authors":"Pontus Timberg, Gustav Hellgren, Magnus Dustler, Anders Tingberg","doi":"10.1117/1.JMI.12.S2.S22003","DOIUrl":"10.1117/1.JMI.12.S2.S22003","url":null,"abstract":"<p><strong>Purpose: </strong>The purpose is to retrospectively investigate how the addition of prior and concurrent mammograms affects wide-angle digital breast tomosynthesis (DBT) screening false-positive recall rates, malignancy scoring, and recall agreement.</p><p><strong>Approach: </strong>A total of 200 cases were selected from the Malmö Breast Tomosynthesis Screening Trial. They consist of 150 recalled cases [30 true positives (TPs), 120 false positives (FPs), and 50 healthy, non-recalled true-negative (TN) cases]. The positive cases were categorized based on being recalled by either DBT, digital mammography (DM), or both. Each case had DBT, synthetic mammography (SM), and DM (prior screening round) images. Five radiologists participated in a reading study where detection, risk of malignancy, and recall were assessed. They read each case twice, once using only DBT and once using DBT together with SM and DM priors.</p><p><strong>Results: </strong>The results showed a significant reduction in recall rates for all FP categories, as well as for the TN cases, when adding SM and prior DM to DBT. This resulted also in a significant increase in recall agreement for these categories, with more of the negative cases being recalled by few or no readers. These categories were overall rated as appearing more malignant in the DBT reading arm. For the TP categories, there was a significant decrease in recalls for DM-recalled cancers ( <math><mrow><mi>p</mi> <mo>=</mo> <mn>0.047</mn></mrow> </math> ), but no significant difference for DBT-recalled cancers ( <math><mrow><mi>p</mi> <mo>=</mo> <mn>0.063</mn></mrow> </math> ), or DBT/DM-recalled cancers ( <math><mrow><mi>p</mi> <mo>=</mo> <mn>0.208</mn></mrow> </math> ).</p><p><strong>Conclusions: </strong>Similar to the documented effect of priors in DM screening, we suggest that added two-dimensional priors improve the specificity of DBT screening but may reduce the sensitivity.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"12 Suppl 2","pages":"S22003"},"PeriodicalIF":1.9,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11931293/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143711591","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 the impact of image restoration in simulating higher dose mammography: effects on the detectability of microcalcifications across different sizes using model observer analysis.","authors":"Renann F Brandão, Lucas E Soares, Lucas R Borges, Predrag R Bakic, Anders Tingberg, Marcelo A C Vieira","doi":"10.1117/1.JMI.12.S2.S22013","DOIUrl":"10.1117/1.JMI.12.S2.S22013","url":null,"abstract":"<p><strong>Purpose: </strong>Breast cancer is one of the leading causes of cancer-related deaths among women, and digital mammography plays a key role in screening and early detection. The radiation dose on mammographic exams directly influences image quality and radiologists' performance. We evaluate the impact of an image restoration pipeline-designed to simulate higher dose acquisitions-on the detectability of microcalcifications of various sizes in mammograms acquired at different radiation doses.</p><p><strong>Approach: </strong>The restoration pipeline denoises the image using a Poisson-Gaussian noise model, combining it with the noisy image to achieve a signal-to-noise ratio comparable with an acquisition at twice the original dose. We created a database of images using a physical breast phantom at doses ranging from 50% to 200% of the standard dose. Clustered microcalcifications were computationally inserted into the phantom images. The channelized Hotelling observer was employed in a four-alternative forced-choice to evaluate the detectability of microcalcifications across different sizes and exposure levels.</p><p><strong>Results: </strong>The restoration of low-dose images acquired at <math><mrow><mo>∼</mo> <mn>75</mn> <mo>%</mo></mrow> </math> of the standard dose resulted in detectability levels comparable with those of images acquired at the standard dose. Moreover, images restored at the standard dose demonstrated detectability similar to those acquired at 160% of the nominal radiation dose, with no statistically significant differences.</p><p><strong>Conclusions: </strong>We demonstrate the potential of an image restoration pipeline to simulate higher quality mammography images. The results indicate that reducing noise through denoising and restoration impacts the detectability of microcalcifications. This method improves image quality without hardware modifications or additional radiation exposure.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"12 Suppl 2","pages":"S22013"},"PeriodicalIF":1.9,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12175087/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144334186","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":"Hybrid simulation of breast CT for assessing microcalcification detectability.","authors":"Su Hyun Lyu, Andrey Makeev, Dan Li, Andreu Badal, Andrew M Hernandez, John M Boone, Stephen J Glick","doi":"10.1117/1.JMI.12.S2.S22015","DOIUrl":"https://doi.org/10.1117/1.JMI.12.S2.S22015","url":null,"abstract":"<p><strong>Purpose: </strong>Virtual imaging trials (VITs) are of interest for regulatory evaluation because they enable faster and more cost-effective evaluation of new imaging technologies than patient clinical trials. Our purpose is to develop a hybrid VIT methodology for breast computed tomography (CT) applications and use it to investigate microcalcification detectability.</p><p><strong>Approach: </strong>Ray tracing was used to generate projection images of clusters of five microcalcifications which varied in diameter, chemical composition, and density. These simulated projection images were added to patient projection images acquired with the fourth-generation breast CT scanner from UC Davis (Doheny) and reconstructed using the Feldkamp filtered backprojection algorithm with varying apodization kernels. Volumes of interest and maximum intensity projections were extracted from the reconstructed volumes. Human observers (HOs) and deep learning model observers (DLMOs) were used to detect calcification clusters, and receiver operating characteristic curve analysis was used to analyze detection performance.</p><p><strong>Results: </strong>DLMO detected 0.18-mm type I calcifications with AUC = 0.80 and 0.21 mm calcifications with <math><mrow><mi>AUC</mi> <mo>=</mo> <mn>0.99</mn></mrow> </math> . HO performance was inferior to deep learning model observer performance, but both HO and DLMO detected 0.21-mm type I calcifications with <math><mrow><mi>AUC</mi> <mo>></mo> <mn>0.90</mn></mrow> </math> and 0.24-mm type I calcifications with near-perfect performance. Microcalcification clusters embedded in adipose tissue were more conspicuous than clusters embedded in fibroglandular tissue. There was superior detection performance for clusters located anteriorly within the breast compared with clusters located posteriorly within the breast.</p><p><strong>Conclusions: </strong>A hybrid approach for virtual imaging trials shows promise for the assessment of imaging systems across a broad range of parameters.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"12 Suppl 2","pages":"S22015"},"PeriodicalIF":1.9,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12225739/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144576688","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":"Workload reduction of digital breast tomosynthesis screening using artificial intelligence and synthetic mammography: a simulation study.","authors":"Victor Dahlblom, Magnus Dustler, Sophia Zackrisson, Anders Tingberg","doi":"10.1117/1.JMI.12.S2.S22005","DOIUrl":"10.1117/1.JMI.12.S2.S22005","url":null,"abstract":"<p><strong>Purpose: </strong>To achieve the high sensitivity of digital breast tomosynthesis (DBT), a time-consuming reading is necessary. However, synthetic mammography (SM) images, equivalent to digital mammography (DM), can be generated from DBT images. SM is faster to read and might be sufficient in many cases. We investigate using artificial intelligence (AI) to stratify examinations into reading of either SM or DBT to minimize workload and maximize accuracy.</p><p><strong>Approach: </strong>This is a retrospective study based on double-read paired DM and one-view DBT from the Malmö Breast Tomosynthesis Screening Trial. DBT examinations were analyzed with the cancer detection AI system ScreenPoint Transpara 1.7. For low-risk examinations, SM reading was simulated by assuming equality with DM reading. For high-risk examinations, the DBT reading results were used. Different combinations of single and double reading were studied.</p><p><strong>Results: </strong>By double-reading the DBT of 30% (4452/14,772) of the cases with the highest risk, and single-reading SM for the rest, 122 cancers would be detected with the same reading workload as DM double reading. That is 28% (27/95) more cancers would be detected than with DM double reading, and in total, 96% (122/127) of the cancers detectable with full DBT double reading would be found.</p><p><strong>Conclusions: </strong>In a DBT-based screening program, AI could be used to select high-risk cases where the reading of DBT is valuable, whereas SM is sufficient for low-risk cases. Substantially, more cancers could be detected compared with DM only, with only a limited increase in reading workload. Prospective studies are necessary.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"12 Suppl 2","pages":"S22005"},"PeriodicalIF":1.9,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12042222/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144003543","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":"Glo-In-One-v2: holistic identification of glomerular cells, tissues, and lesions in human and mouse histopathology.","authors":"Lining Yu, Mengmeng Yin, Ruining Deng, Quan Liu, Tianyuan Yao, Can Cui, Junlin Guo, Yu Wang, Yaohong Wang, Shilin Zhao, Haichun Yang, Yuankai Huo","doi":"10.1117/1.JMI.12.6.061406","DOIUrl":"10.1117/1.JMI.12.6.061406","url":null,"abstract":"<p><strong>Purpose: </strong>Segmenting intraglomerular tissue and glomerular lesions traditionally depends on detailed morphological evaluations by expert nephropathologists, a labor-intensive process susceptible to interobserver variability. Our group previously developed the Glo-In-One toolkit for integrated glomerulus detection and segmentation. We leverage the Glo-In-One toolkit to version 2 (Glo-In-One-v2), which adds fine-grained segmentation capabilities. We curated 14 distinct labels spanning tissue regions, cells, and lesions across 23,529 annotated glomeruli from human and mouse histopathology data. To our knowledge, this dataset is among the largest of its kind to date.</p><p><strong>Approach: </strong>We present a single dynamic-head deep learning architecture for segmenting 14 classes within partially labeled images from human and mouse kidney pathology. The model was trained on data derived from 368 annotated kidney whole-slide images with five key intraglomerular tissue types and nine glomerular lesion types.</p><p><strong>Results: </strong>The glomerulus segmentation model achieved a decent performance compared with baselines and achieved a 76.5% average Dice similarity coefficient. In addition, transfer learning from rodent to human for the glomerular lesion segmentation model has enhanced the average segmentation accuracy across different types of lesions by more than 3%, as measured by Dice scores.</p><p><strong>Conclusions: </strong>We introduce a convolutional neural network for multiclass segmentation of intraglomerular tissue and lesions. The Glo-In-One-v2 model and pretrained weight are publicly available at https://github.com/hrlblab/Glo-In-One_v2.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"12 6","pages":"061406"},"PeriodicalIF":1.7,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12303538/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144733981","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":"Impact of synthetic data on training a deep learning model for lesion detection and classification in contrast-enhanced mammography.","authors":"Astrid Van Camp, Henry C Woodruff, Lesley Cockmartin, Marc Lobbes, Michael Majer, Corinne Balleyguier, Nicholas W Marshall, Hilde Bosmans, Philippe Lambin","doi":"10.1117/1.JMI.12.S2.S22006","DOIUrl":"10.1117/1.JMI.12.S2.S22006","url":null,"abstract":"<p><strong>Purpose: </strong>Predictive models for contrast-enhanced mammography often perform better at detecting and classifying enhancing masses than (non-enhancing) microcalcification clusters. We aim to investigate whether incorporating synthetic data with simulated microcalcification clusters during training can enhance model performance.</p><p><strong>Approach: </strong>Microcalcification clusters were simulated in low-energy images of lesion-free breasts from 782 patients, considering local texture features. Enhancement was simulated in the corresponding recombined images. A deep learning (DL) model for lesion detection and classification was trained with varying ratios of synthetic and real (850 patients) data. In addition, a handcrafted radiomics classifier was trained using delineations and class labels from real data, and predictions from both models were ensembled. Validation was performed on internal (212 patients) and external (279 patients) real datasets.</p><p><strong>Results: </strong>The DL model trained exclusively with synthetic data detected over 60% of malignant lesions. Adding synthetic data to smaller real training sets improved detection sensitivity for malignant lesions but decreased precision. Performance plateaued at a detection sensitivity of 0.80. The ensembled DL and radiomics models performed worse than the standalone DL model, decreasing the area under this receiver operating characteristic curve from 0.75 to 0.60 on the external validation set, likely due to falsely detected suspicious regions of interest.</p><p><strong>Conclusions: </strong>Synthetic data can enhance DL model performance, provided model setup and data distribution are optimized. The possibility to detect malignant lesions without real data present in the training set confirms the utility of synthetic data. It can serve as a helpful tool, especially when real data are scarce, and it is most effective when complementing real data.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"12 Suppl 2","pages":"S22006"},"PeriodicalIF":1.9,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12036226/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144001778","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":"Comparing percent breast density assessments of an AI-based method with expert reader estimates: inter-observer variability.","authors":"Stepan Romanov, Sacha Howell, Elaine Harkness, Dafydd Gareth Evans, Sue Astley, Martin Fergie","doi":"10.1117/1.JMI.12.S2.S22011","DOIUrl":"10.1117/1.JMI.12.S2.S22011","url":null,"abstract":"<p><strong>Purpose: </strong>Breast density estimation is an important part of breast cancer risk assessment, as mammographic density is associated with risk. However, density assessed by multiple experts can be subject to high inter-observer variability, so automated methods are increasingly used. We investigate the inter-reader variability and risk prediction for expert assessors and a deep learning approach.</p><p><strong>Approach: </strong>Screening data from a cohort of 1328 women, case-control matched, was used to compare between two expert readers and between a single reader and a deep learning model, Manchester artificial intelligence - visual analog scale (MAI-VAS). Bland-Altman analysis was used to assess the variability and matched concordance index to assess risk.</p><p><strong>Results: </strong>Although the mean differences for the two experiments were alike, the limits of agreement between MAI-VAS and a single reader are substantially lower at +SD (standard deviation) 21 (95% CI: 19.65, 21.69) -SD 22 (95% CI: <math><mrow><mo>-</mo> <mn>22.71</mn></mrow> </math> , <math><mrow><mo>-</mo> <mn>20.68</mn></mrow> </math> ) than between two expert readers +SD 31 (95% CI: 32.08, 29.23) -SD 29 (95% CI: <math><mrow><mo>-</mo> <mn>29.94</mn></mrow> </math> , <math><mrow><mo>-</mo> <mn>27.09</mn></mrow> </math> ). In addition, breast cancer risk discrimination for the deep learning method and density readings from a single expert was similar, with a matched concordance of 0.628 (95% CI: 0.598, 0.658) and 0.624 (95% CI: 0.595, 0.654), respectively. The automatic method had a similar inter-view agreement to experts and maintained consistency across density quartiles.</p><p><strong>Conclusions: </strong>The artificial intelligence breast density assessment tool MAI-VAS has a better inter-observer agreement with a randomly selected expert reader than that between two expert readers. Deep learning-based density methods provide consistent density scores without compromising on breast cancer risk discrimination.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"12 Suppl 2","pages":"S22011"},"PeriodicalIF":1.9,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12159425/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144303313","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":"Robust evaluation of tissue-specific radiomic features for classifying breast tissue density grades.","authors":"Vincent Dong, Walter Mankowski, Telmo M Silva Filho, Anne Marie McCarthy, Despina Kontos, Andrew D A Maidment, Bruno Barufaldi","doi":"10.1117/1.JMI.12.S2.S22010","DOIUrl":"10.1117/1.JMI.12.S2.S22010","url":null,"abstract":"<p><strong>Purpose: </strong>Breast cancer risk depends on an accurate assessment of breast density due to lesion masking. Although governed by standardized guidelines, radiologist assessment of breast density is still highly variable. Automated breast density assessment tools leverage deep learning but are limited by model robustness and interpretability.</p><p><strong>Approach: </strong>We assessed the robustness of a feature selection methodology (RFE-SHAP) for classifying breast density grades using tissue-specific radiomic features extracted from raw central projections of digital breast tomosynthesis screenings ( <math> <mrow> <msub><mrow><mi>n</mi></mrow> <mrow><mi>I</mi></mrow> </msub> <mo>=</mo> <mn>651</mn></mrow> </math> , <math> <mrow> <msub><mrow><mi>n</mi></mrow> <mrow><mi>II</mi></mrow> </msub> <mo>=</mo> <mn>100</mn></mrow> </math> ). RFE-SHAP leverages traditional and explainable AI methods to identify highly predictive and influential features. A simple logistic regression (LR) classifier was used to assess classification performance, and unsupervised clustering was employed to investigate the intrinsic separability of density grade classes.</p><p><strong>Results: </strong>LR classifiers yielded cross-validated areas under the receiver operating characteristic (AUCs) per density grade of [ <math><mrow><mi>A</mi></mrow> </math> : <math><mrow><mn>0.909</mn> <mo>±</mo> <mn>0.032</mn></mrow> </math> , <math><mrow><mi>B</mi></mrow> </math> : <math><mrow><mn>0.858</mn> <mo>±</mo> <mn>0.027</mn></mrow> </math> , <math><mrow><mi>C</mi></mrow> </math> : <math><mrow><mn>0.927</mn> <mo>±</mo> <mn>0.013</mn></mrow> </math> , <math><mrow><mi>D</mi></mrow> </math> : <math><mrow><mn>0.890</mn> <mo>±</mo> <mn>0.089</mn></mrow> </math> ] and an AUC of <math><mrow><mn>0.936</mn> <mo>±</mo> <mn>0.016</mn></mrow> </math> for classifying patients as nondense or dense. In external validation, we observed per density grade AUCs of [ <math><mrow><mi>A</mi></mrow> </math> : 0.880, <math><mrow><mi>B</mi></mrow> </math> : 0.779, <math><mrow><mi>C</mi></mrow> </math> : 0.878, <math><mrow><mi>D</mi></mrow> </math> : 0.673] and nondense/dense AUC of 0.823. Unsupervised clustering highlighted the ability of these features to characterize different density grades.</p><p><strong>Conclusions: </strong>Our RFE-SHAP feature selection methodology for classifying breast tissue density generalized well to validation datasets after accounting for natural class imbalance, and the identified radiomic features properly captured the progression of density grades. Our results potentiate future research into correlating selected radiomic features with clinical descriptors of breast tissue density.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"12 Suppl 2","pages":"S22010"},"PeriodicalIF":1.9,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12120562/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144200544","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":"TFKT V2: task-focused knowledge transfer from natural images for computed tomography perceptual image quality assessment.","authors":"Kazi Ramisa Rifa, Md Atik Ahamed, Jie Zhang, Abdullah Imran","doi":"10.1117/1.JMI.12.5.051805","DOIUrl":"10.1117/1.JMI.12.5.051805","url":null,"abstract":"<p><strong>Purpose: </strong>The accurate assessment of computed tomography (CT) image quality is crucial for ensuring diagnostic reliability while minimizing radiation dose. Radiologists' evaluations are time-consuming and labor-intensive. Existing automated approaches often require large CT datasets with predefined image quality assessment (IQA) scores, which often do not align well with clinical evaluations. We aim to develop a reference-free, automated method for CT IQA that closely reflects radiologists' evaluations, reducing the dependency on large annotated datasets.</p><p><strong>Approach: </strong>We propose Task-Focused Knowledge Transfer (TFKT), a deep learning-based IQA method leveraging knowledge transfer from task-similar natural image datasets. TFKT incorporates a hybrid convolutional neural network-transformer model, enabling accurate quality predictions by learning from natural image distortions with human-annotated mean opinion scores. The model is pre-trained on natural image datasets and fine-tuned on low-dose computed tomography perceptual image quality assessment data to ensure task-specific adaptability.</p><p><strong>Results: </strong>Extensive evaluations demonstrate that the proposed TFKT method effectively predicts IQA scores aligned with radiologists' assessments on in-domain datasets and generalizes well to out-of-domain clinical pediatric CT exams. The model achieves robust performance without requiring high-dose reference images. Our model is capable of assessing the quality of <math><mrow><mo>∼</mo> <mn>30</mn></mrow> </math> CT image slices in a second.</p><p><strong>Conclusions: </strong>The proposed TFKT approach provides a scalable, accurate, and reference-free solution for CT IQA. The model bridges the gap between traditional and deep learning-based IQA, offering clinically relevant and computationally efficient assessments applicable to real-world clinical settings.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"12 5","pages":"051805"},"PeriodicalIF":1.9,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12116730/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144182165","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}