Ultrasound in Medicine and Biology最新文献

{"title":"Assessment of Coronary Microcirculation with High Frame-Rate Contrast-Enhanced Echocardiography","authors":"Geraldi Wahyulaksana ,&nbsp;Luxi Wei ,&nbsp;Jason Voorneveld ,&nbsp;Maaike te Lintel Hekkert ,&nbsp;Daniel J. Bowen ,&nbsp;Mihai Strachinaru ,&nbsp;Dirk J. Duncker ,&nbsp;Antonius F.W. van der Steen ,&nbsp;Hendrik J. Vos","doi":"10.1016/j.ultrasmedbio.2024.12.002","DOIUrl":"10.1016/j.ultrasmedbio.2024.12.002","url":null,"abstract":"<div><h3>Objective</h3><div>Assessing myocardial perfusion in acute myocardial infarction is important for guiding clinicians in choosing appropriate treatment strategies. Echocardiography can be used due to its direct feedback and bedside nature, but it currently faces image quality issues and an inability to differentiate coronary macro- from micro-circulation. We previously developed an imaging scheme using high frame-rate contrast-enhanced ultrasound (HFR CEUS) with higher order singular value decomposition (HOSVD) that provides dynamic perfusion and vascular flow visualization. In this study, we aim to show the ability of this technique to image perfusion deficits and investigate the potential occurrence of false-positive contrast detection.</div></div><div><h3>Methods</h3><div>We used a porcine model comprising occlusion and release of the left anterior descending coronary artery. During slow contrast agent infusion, the afore-mentioned imaging scheme was used to capture and process the data offline using HOSVD.</div></div><div><h3>Results</h3><div>Fast and slow coronary flow was successfully differentiated, presumably representing the different compartments of the micro-circulation. Low perfusion was seen in the area that was affected, as expected by vascular occlusion. Furthermore, we also imaged coronary flow dynamics before, during and after release of the occlusion, the latter showing hyperemia as expected. A contrast agent destruction test showed that the processed images contained actual contrast signal in the cardiac phases with minimal motion. With larger tissue motion, tissue signal leaked into the contrast-enhanced images.</div></div><div><h3>Conclusion</h3><div>Our results demonstrate the feasibility of HFR CEUS with HOSVD as a viable option for assessing myocardial perfusion. Flow dynamics were resolved, which potentially helped to directly evaluate coronary flow deficits.</div></div>","PeriodicalId":49399,"journal":{"name":"Ultrasound in Medicine and Biology","volume":"51 3","pages":"Pages 585-591"},"PeriodicalIF":2.4,"publicationDate":"2025-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142933179","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Machine Learning Models for Predicting Clinically and Ultrasound-Negative Axillary Lymph Node Metastasis in Early-Stage Breast Cancer","authors":"Yu Du,&nbsp;Xuehong Diao","doi":"10.1016/j.ultrasmedbio.2024.12.017","DOIUrl":"10.1016/j.ultrasmedbio.2024.12.017","url":null,"abstract":"","PeriodicalId":49399,"journal":{"name":"Ultrasound in Medicine and Biology","volume":"51 3","pages":"Page 606"},"PeriodicalIF":2.4,"publicationDate":"2025-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142933183","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Real-Time Volumetric Free-Hand Ultrasound Imaging for Large-Sized Organs: A Study of Imaging the Whole Spine","authors":"Caozhe Li ,&nbsp;Enxiang Shen ,&nbsp;Haoyang Wang ,&nbsp;Yuxin Wang ,&nbsp;Jie Yuan ,&nbsp;Li Gong ,&nbsp;Di Zhao ,&nbsp;Weijing Zhang ,&nbsp;Zhibin Jin","doi":"10.1016/j.ultrasmedbio.2024.12.015","DOIUrl":"10.1016/j.ultrasmedbio.2024.12.015","url":null,"abstract":"<div><h3>Objectives</h3><div>Three-dimensional (3D) ultrasound imaging can overcome the limitations of conventional two-dimensional (2D) ultrasound imaging in structural observation and measurement. However, conducting volumetric ultrasound imaging for large-sized organs still faces difficulties including long acquisition time, inevitable patient movement, and 3D feature recognition. In this study, we proposed a real-time volumetric free-hand ultrasound imaging system optimized for the above issues and applied it to the clinical diagnosis of scoliosis.</div></div><div><h3>Methods</h3><div>This study employed an incremental imaging method coupled with algorithmic acceleration to enable real-time processing and visualization of the large amounts of data generated when scanning large-sized organs. Furthermore, to deal with the difficulty of image feature recognition, we proposed two tissue segmentation algorithms to reconstruct and visualize the spinal anatomy in 3D space by approximating the depth at which the bone structures are located and segmenting the ultrasound images at different depths.</div></div><div><h3>Results</h3><div>We validated the adaptability of our system by deploying it to multiple models of ultrasound equipment and conducting experiments using different types of ultrasound probes. We also conducted experiments on six scoliosis patients and 10 normal volunteers to evaluate the performance of our proposed method. Ultrasound imaging of a volunteer spine from shoulder to crotch (more than 500 mm) was performed in 2 minutes, and the 3D imaging results displayed in real-time were compared with the corresponding X-ray images with a correlation coefficient of 0.96 in spinal curvature.</div></div><div><h3>Conclusion</h3><div>Our proposed volumetric ultrasound imaging system might hold the potential to be clinically applied to other large-sized organs.</div></div>","PeriodicalId":49399,"journal":{"name":"Ultrasound in Medicine and Biology","volume":"51 3","pages":"Pages 598-605"},"PeriodicalIF":2.4,"publicationDate":"2025-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142933270","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Elasticity Characteristics of Thenar Muscles in Carpal Tunnel Syndrome","authors":"Naoya Iida ,&nbsp;Hicham El Hor ,&nbsp;Dirk R. Larson ,&nbsp;Marianne T. Luetmer ,&nbsp;Ruple S. Laughlin ,&nbsp;Peter C. Amadio","doi":"10.1016/j.ultrasmedbio.2024.12.003","DOIUrl":"10.1016/j.ultrasmedbio.2024.12.003","url":null,"abstract":"<div><h3>Objective</h3><div>Measurement of thenar muscle elasticity by ultrasound shear wave elastography (SWE) may be useful for the diagnosis and evaluation of carpal tunnel syndrome (CTS), but there is a paucity of information on SWE of the thenar muscles in patients with CTS. The purpose of this study was to investigate the elasticity of the thenar muscles in patients with CTS.</div></div><div><h3>Methods</h3><div>Twenty-two adult patients with a referral diagnosis of CTS (27 hands) and 20 healthy volunteers as a control (20 dominant hands) participated in this study. The elastic modulus of the thenar muscles was measured with SWE in two conditions, rest and pinch. The elastic modulus and percent change between the two conditions for each muscle were compared between groups.</div></div><div><h3>Results</h3><div>The elastic modulus of the abductor pollicis brevis (APB) in the patient group was lower than that in the control group at rest (20.6 ± 5.6 kPa vs. 25.0 ± 8.5 kPa; <em>p</em> = 0.034). However, the elastic modulus (87.9 ± 49.0 kPa vs. 62.5 ± 28.5 kPa; <em>p</em> = 0.044) and percent change (373.3 ± 336.4% vs. 169.2 ± 116.0%; <em>p</em> = 0.006) of the APB in the patient group were higher than those in the control group with pinch. For the opponens pollicis and adductor pollicis, there was no significant difference in the elastic modulus between groups.</div></div><div><h3>Conclusion</h3><div>Patients with CTS showed differences in SWE of the APB. Future studies need to investigate the accuracy of SWE in diagnosing CTS and assessing prognosis.</div></div>","PeriodicalId":49399,"journal":{"name":"Ultrasound in Medicine and Biology","volume":"51 3","pages":"Pages 592-597"},"PeriodicalIF":2.4,"publicationDate":"2025-01-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142928544","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Ultrasound of the Median Nerve: A Pictorial Review. Normal Ultrasound Findings and Variations","authors":"Marco Becciolini ,&nbsp;Giorgio Tamborrini ,&nbsp;Christopher Pivec ,&nbsp;Georg Riegler","doi":"10.1016/j.ultrasmedbio.2024.12.001","DOIUrl":"10.1016/j.ultrasmedbio.2024.12.001","url":null,"abstract":"<div><div>This is the first of a two-part article in which we focus on the Ultrasound (US) appearance of the normal median nerve (MN) and its main branches. The detailed anatomy and US anatomy of the MN course are presented with high-resolution images obtained with the latest-generation US machines and transducers. Variations are discussed to avoid misinterpretation of normal findings.</div></div>","PeriodicalId":49399,"journal":{"name":"Ultrasound in Medicine and Biology","volume":"51 3","pages":"Pages 439-451"},"PeriodicalIF":2.4,"publicationDate":"2025-01-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142927621","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Added Value of Dynamic Contrast-Enhanced Ultrasound Analysis for Differential Diagnosis of Small (≤20 mm) Solid Pancreatic Lesions","authors":"Xiu-Yun Lu ,&nbsp;Jia-Ying Cao ,&nbsp;Sheng Chen ,&nbsp;Ying Wang ,&nbsp;Li Wei ,&nbsp;Wei Gong ,&nbsp;Wen-Hui Lou ,&nbsp;Yi Dong","doi":"10.1016/j.ultrasmedbio.2024.11.015","DOIUrl":"10.1016/j.ultrasmedbio.2024.11.015","url":null,"abstract":"<div><h3>Objective</h3><div>To evaluate the added value of dynamic contrast-enhanced ultrasound (DCE-US) analysis in pre-operative differential diagnosis of small (≤20 mm) solid pancreatic lesions (SPLs).</div></div><div><h3>Methods</h3><div>In this retrospective study, patients with biopsy or surgerical resection and histopathologically confirmed small (≤20 mm) SPLs were included. One wk before biopsy/surgery, pre-operative B-mode ultrasound and contrast-enhanced ultrasound were performed. An ultrasonic system (ACUSON Sequoia, Siemens Medical Solutions, PA, USA) equipped with a 5C1 MHz convex array transducer was utilized. A dose of 1.5 ml SonoVue (Bracco, Italy) was injected as the contrast agent. Time-intensity curves were generated using VueBox software (Bracco) and various DCE-US quantitative parameters were subsequently calculated after curve fitting. Univariate and multivariate logistic regression analysis were utilized.</div></div><div><h3>Results</h3><div>From August 2020 to November 2023, a total of 76 patients (31 males and 45 females; mean age: 61.9 ± 10.5 y) with 76 small (≤20 mm) SPLs were included. Mean size of the lesions was 16.4 ± 0.4 mm (range: 7–20 mm). Final diagnosis included 37 benign and 39 malignant small SPLs. On B-mode ultrasound, the majority of malignant (37/39, 94.9%) and benign SPLs (30/37, 81.1%) were hypo-echoic lesions with ill-defined borders and irregular shapes (<em>p</em> &gt; 0.05). During the arterial phase of contrast-enhanced ultrasound, most SPLs (59/76, 77.6%) exhibited iso-enhancement when compared with surrounding pancreatic parenchyma. Subsequently, 82.1% (32/39) of malignant SPLs and 35.1% (13/37) of benign SPLs demonstrated wash-out in the venous phase and showed hypo-enhancement in venous and late phases (<em>p</em> &gt; 0.05). Compared with benign SPLs, the time-intensity curves of small malignant SPLs revealed earlier and lower enhancement in the arterial phase, and a faster decline during the venous phase with a decreased area under the curve. Among the quantitative parameters, a lower peak enhancement ratio and higher fall time ratio were more common in small malignant SPLs (<em>p</em> &lt; 0.05). For DCE-US analysis, the combined areas under the curve of significant quantitative parameters was 0.919, with 87.2% sensitivity and 86.5% specificity when differentiating between small malignant and benign SPLs. This result was better than contrast-enhanced computed tomography, which has a sensitivity of 74.4% and a specificity of 75.7%.</div></div><div><h3>Conclusion</h3><div>DCE-US analysis provides added value for the pre-operative differential diagnosis of small malignant SPLs.</div></div>","PeriodicalId":49399,"journal":{"name":"Ultrasound in Medicine and Biology","volume":"51 3","pages":"Pages 535-542"},"PeriodicalIF":2.4,"publicationDate":"2025-01-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142928542","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Effect of Interventional Ultrasound in the Treatment of Postthyroidectomy Vocal Cord Dysfunction: A Preliminary Exploratory Study","authors":"Zhihao Jiang ,&nbsp;Yifan Zhu ,&nbsp;Bin Li ,&nbsp;Zhen Shan ,&nbsp;Shasha Huang ,&nbsp;Jie Deng ,&nbsp;Zhu Wang","doi":"10.1016/j.ultrasmedbio.2024.11.021","DOIUrl":"10.1016/j.ultrasmedbio.2024.11.021","url":null,"abstract":"<div><h3>Purpose</h3><div>To explore the efficacy of interventional ultrasound treatment for hoarseness caused by nondisconnected recurrent laryngeal nerve injury after thyroidectomy.</div></div><div><h3>Materials and methods</h3><div>In this retrospective study, we analysed the clinical data of 21 patients who underwent interventional ultrasound therapy (ultrasound-guided injection of a Diprospan and saline mixture) for postthyroidectomy hoarseness at our hospital between August 1, 2023, and January 31, 2024 (the Diprospan group) and randomly selected 21 patients who did not receive any treatment for postthyroidectomy hoarseness during the same period as the control group. The average vocal cord activity improvement time for the Diprospan group was calculated and compared with that of untreated patients from previous studies. Vocal cord activity improvement rates were assessed at 1, 2, and 3 mo after thyroidectomy, and changes in the voice handicap index (VHI) and the subjective listening perception scale (GRBAS) scores before and after treatment were evaluated. Additionally, the efficacy was compared between patients treated within 2 wk (timely intervention group, TI group) and those treated 2 wk after thyroidectomy (delayed intervention group, DI group) to explore the impact of treatment timing on outcomes.</div></div><div><h3>Results</h3><div>In the Diprospan group, the average vocal cord activity improvement time was 44.6 ± 4.7 d for those treated within 2 wk after thyroidectomy and 17.7 ± 1.5 d for those treated 2 wk after thyroidectomy, both of which were significantly shorter than the 6 mo to 1 y reported in the literature. The improvement rates at 1, 2, and 3 mo after thyroidectomy were significantly greater in the Diprospan group than in the control group (0.0% vs. 28.6%, 23.8% vs. 81.0%, 33.3% vs. 90.5%, <em>p</em> &lt; 0.05). The VHI score (72.6 ± 21.2 vs. 12.6 ± 24.7, <em>p</em> &lt; 0.001) and GRBAS score (12.0 vs. 0.0, <em>p</em> &lt; 0.001) significantly improved after treatment. Compared with the DI group, the TI group also showed faster voice recovery (19.8 ± 5.8 d vs. 14.0 ± 2.6 d, <em>p</em> &lt; 0.05). No treatment-related adverse reactions were observed.</div></div><div><h3>Conclusion</h3><div>Interventional ultrasound is a safe and effective treatment for hoarseness after thyroidectomy, as it effectively improves vocal cord mobility and accelerates voice recovery.</div></div>","PeriodicalId":49399,"journal":{"name":"Ultrasound in Medicine and Biology","volume":"51 3","pages":"Pages 578-584"},"PeriodicalIF":2.4,"publicationDate":"2024-12-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142899906","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Ultrasonographic Measurement of Common Carotid Artery Wall Pulse Dynamics and Longitudinal Motion – Method Validation and a Novel Parameter Ratio","authors":"Artturi Petäjä ,&nbsp;Yuxiang Zhu ,&nbsp;Azin Khodaverdi ,&nbsp;Tobias Erlöv ,&nbsp;Åsa Rydén Ahlgren ,&nbsp;Magnus Cinthio","doi":"10.1016/j.ultrasmedbio.2024.11.019","DOIUrl":"10.1016/j.ultrasmedbio.2024.11.019","url":null,"abstract":"<div><h3>Objectives</h3><div>The enormous burden that cardiovascular diseases put on individuals and societies warrants reliable biomarkers of disease risk to optimize disease prevention. We studied longitudinal movement (LMov) in arterial walls using ultrasound of the common carotid artery (CCA). We believe that LMov could be a sensitive biomarker of cardiovascular health and in this study, we evaluate the intra-observer repeatability and inter-observer precision of our method. We also present a novel parameter ratio.</div></div><div><h3>Method</h3><div>Four independent researchers, called “observers”, analyzed CINE-loops of two separate ultrasound recordings of the CCA in 20 healthy individuals using an in-house developed program (ARTIC) based on MATLAB. Vessel wall displacement was measured in two dimensions and results between repeated measurements for one research subject and between observers were compared. We also present and evaluate a novel parameter ratio, the notch ratio (NR) which by combining velocities of phases B and X reduces variability and improves comparability of LMov analysis in late systole.</div></div><div><h3>Results</h3><div>Most LMov parameters showed good repeatability with coefficient of variation (CV) values 10%–26%. CV values for radial and IMT measures were excellent between 1% and 7%. Intra-class correlation coefficients (ICC) for inter-observer precision of single measures showed excellent values ICC &gt; 0.9. NR improved LMov analysis in late systole and showed homogenous values of 1.36–1.45 between observers and good CV values of 9%–20%.</div></div><div><h3>Conclusion</h3><div>We demonstrate good intra-observer repeatability and excellent inter-observer precision of LMov analysis using ARTIC. NR displays similar consistency. This promotes further research on larger study populations using the same method focusing on the possible pathophysiological link between LMov and cardiovascular disease.</div></div>","PeriodicalId":49399,"journal":{"name":"Ultrasound in Medicine and Biology","volume":"51 3","pages":"Pages 559-567"},"PeriodicalIF":2.4,"publicationDate":"2024-12-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142899939","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Interactively Fusing Global and Local Features for Benign and Malignant Classification of Breast Ultrasound Images","authors":"Wenhan Wang ,&nbsp;Jiale Zhou ,&nbsp;Jin Zhao ,&nbsp;Xun Lin ,&nbsp;Yan Zhang ,&nbsp;Shan Lu ,&nbsp;Wanchen Zhao ,&nbsp;Shuai Wang ,&nbsp;Wenzhong Tang ,&nbsp;Xiaolei Qu","doi":"10.1016/j.ultrasmedbio.2024.11.014","DOIUrl":"10.1016/j.ultrasmedbio.2024.11.014","url":null,"abstract":"<div><h3>Objective</h3><div>Breast ultrasound (BUS) is used to classify benign and malignant breast tumors, and its automatic classification can reduce subjectivity. However, current convolutional neural networks (CNNs) face challenges in capturing global features, while vision transformer (ViT) networks have limitations in effectively extracting local features. Therefore, this study aimed to develop a deep learning method that enables the interaction and updating of intermediate features between CNN and ViT to achieve high-accuracy BUS image classification.</div></div><div><h3>Methods</h3><div>This study introduced the CNN and transformer multi-stage fusion network (CTMF-Net) consisting of two branches: a CNN branch and a transformer branch. The CNN branch employs visual geometry group as its backbone, while the transformer branch utilizes ViT as its base network. Both branches were divided into four stages. At the end of each stage, a proposed feature interaction module facilitated feature interaction and fusion between the two branches. Additionally, the convolutional block attention module was employed to enhance relevant features after each stage of the CNN branch. Extensive experiments were conducted using various state-of-the-art deep-learning classification methods on three public breast ultrasound datasets (SYSU, UDIAT and BUSI).</div></div><div><h3>Results</h3><div>For the internal validation on SYSU and UDIAT, our proposed method CTMF-Net achieved the highest accuracy of 90.14 ± 0.58% on SYSU and 92.04 ± 4.90% on UDIAT, which showed superior classification performance over other state-of-art networks (<em>p</em> &lt; 0.05). Additionally, for external validation on BUSI, CTMF-Net showed outstanding performance, achieving the highest area under the curve score of 0.8704 when trained on SYSU, marking a 0.0126 improvement over the second-best visual geometry group attention ViT method. Similarly, when applied to UDIAT, CTMF-Net achieved an area under the curve score of 0.8505, surpassing the second-best global context ViT method by 0.0130.</div></div><div><h3>Conclusion</h3><div>Our proposed method, CTMF-Net, outperforms all existing methods and can effectively assist doctors in achieving more accurate classification performance of breast tumors.</div></div>","PeriodicalId":49399,"journal":{"name":"Ultrasound in Medicine and Biology","volume":"51 3","pages":"Pages 525-534"},"PeriodicalIF":2.4,"publicationDate":"2024-12-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142873093","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Attention-based Fusion Network for Breast Cancer Segmentation and Classification Using Multi-modal Ultrasound Images","authors":"Yoonjae Cho ,&nbsp;Sampa Misra ,&nbsp;Ravi Managuli ,&nbsp;Richard G. Barr ,&nbsp;Jeongmin Lee ,&nbsp;Chulhong Kim","doi":"10.1016/j.ultrasmedbio.2024.11.020","DOIUrl":"10.1016/j.ultrasmedbio.2024.11.020","url":null,"abstract":"<div><h3>Objective</h3><div>Breast cancer is one of the most commonly occurring cancers in women. Thus, early detection and treatment of cancer lead to a better outcome for the patient. Ultrasound (US) imaging plays a crucial role in the early detection of breast cancer, providing a cost-effective, convenient, and safe diagnostic approach. To date, much research has been conducted to facilitate reliable and effective early diagnosis of breast cancer through US image analysis. Recently, with the introduction of machine learning technologies such as deep learning (DL), automated lesion segmentation and classification, the identification of malignant masses in US breasts has progressed, and computer-aided diagnosis (CAD) technology is being applied in clinics effectively. Herein, we propose a novel deep learning-based “segmentation + classification” model based on B- and SE-mode images.</div></div><div><h3>Methods</h3><div>For the segmentation task, we propose a Multi-Modal Fusion U-Net (MMF-U-Net), which segments lesions by mixing B- and SE-mode information through fusion blocks. After segmenting, the lesion area from the B- and SE-mode images is cropped using a predicted segmentation mask. The encoder part of the pre-trained MMF-U-Net model is then used on the cropped B- and SE-mode breast US images to classify benign and malignant lesions.</div></div><div><h3>Results</h3><div>The experimental results using the proposed method showed good segmentation and classification scores. The dice score, intersection over union (IoU), precision, and recall are 78.23%, 68.60%, 82.21%, and 80.58%, respectively, using the proposed MMF-U-Net on real-world clinical data. The classification accuracy is 98.46%.</div></div><div><h3>Conclusion</h3><div>Our results show that the proposed method will effectively segment the breast lesion area and can reliably classify the benign from malignant lesions.</div></div>","PeriodicalId":49399,"journal":{"name":"Ultrasound in Medicine and Biology","volume":"51 3","pages":"Pages 568-577"},"PeriodicalIF":2.4,"publicationDate":"2024-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142856428","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}