<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Publishing DTD v1.3 20210610//EN" "JATS-journalpublishing1-3.dtd">
<article article-type="research-article" dtd-version="1.3" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xml:lang="ru"><front><journal-meta><journal-id journal-id-type="publisher-id">innosfera</journal-id><journal-title-group><journal-title xml:lang="ru">Наука и инновации</journal-title><trans-title-group xml:lang="en"><trans-title>Science and Innovations</trans-title></trans-title-group></journal-title-group><issn pub-type="ppub">1818-9857</issn><issn pub-type="epub">2412-9372</issn><publisher><publisher-name>Издательский дом «Белорусская наука»</publisher-name></publisher></journal-meta><article-meta><article-id pub-id-type="doi">10.29235/1818-9857-2026-04-78-83</article-id><article-id custom-type="elpub" pub-id-type="custom">innosfera-1027</article-id><article-categories><subj-group subj-group-type="heading"><subject>Research Article</subject></subj-group><subj-group subj-group-type="section-heading" xml:lang="ru"><subject>ДИССЕРТАЦИОННЫЕ ИССЛЕДОВАНИЯ</subject></subj-group><subj-group subj-group-type="section-heading" xml:lang="en"><subject>DISSERTATION RESEARCH</subject></subj-group></article-categories><title-group><article-title>Автоматическая сегментация гемангиомы печени на МРТ с использованием модифицированной архитектуры U-Net</article-title><trans-title-group xml:lang="en"><trans-title>Automated segmentation of liver hemangiomas on MRI using a modified U-Net architecture</trans-title></trans-title-group></title-group><contrib-group><contrib contrib-type="author" corresp="yes"><name-alternatives><name name-style="eastern" xml:lang="ru"><surname>Кадан</surname><given-names>А.</given-names></name><name name-style="western" xml:lang="en"><surname>Kadan</surname><given-names>A.</given-names></name></name-alternatives><bio xml:lang="ru"><p>Александр Кадан, заведующий кафедрой системного программирования и компьютерной безопасности, кандидат технических наук, доцент</p></bio><xref ref-type="aff" rid="aff-1"/></contrib><contrib contrib-type="author" corresp="yes"><name-alternatives><name name-style="eastern" xml:lang="ru"><surname>Петров</surname><given-names>С.</given-names></name><name name-style="western" xml:lang="en"><surname>Petrov</surname><given-names>S.</given-names></name></name-alternatives><bio xml:lang="ru"><p>Сергей Петров, доцент кафедры системного программирования и компьютерной безопасности, кандидат медицинских наук, магистр технических наук, доцент</p></bio><xref ref-type="aff" rid="aff-1"/></contrib><contrib contrib-type="author" corresp="yes"><name-alternatives><name name-style="eastern" xml:lang="ru"><surname>Прокопович</surname><given-names>А.</given-names></name><name name-style="western" xml:lang="en"><surname>Prokopovich</surname><given-names>A.</given-names></name></name-alternatives><bio xml:lang="ru"><p>Андрей Прокопович, врач лучевой диагностики кабинета магнитно-резонансной томографии; старший преподаватель кафедры лучевой диагностики и терапии</p></bio><xref ref-type="aff" rid="aff-2"/></contrib><contrib contrib-type="author" corresp="yes"><name-alternatives><name name-style="eastern" xml:lang="ru"><surname>Зеньков</surname><given-names>О.</given-names></name><name name-style="western" xml:lang="en"><surname>Zenkov</surname><given-names>O.</given-names></name></name-alternatives><bio xml:lang="ru"><p>Олег Зеньков, заместитель главного врача по организационно-методической работе</p><p> </p></bio><xref ref-type="aff" rid="aff-3"/></contrib><contrib contrib-type="author" corresp="yes"><name-alternatives><name name-style="eastern" xml:lang="ru"><surname>Стурейко</surname><given-names>А.</given-names></name><name name-style="western" xml:lang="en"><surname>Stureiko</surname><given-names>A.</given-names></name></name-alternatives><bio xml:lang="ru"><p>Артем Стурейко, инженер электросвязи</p></bio><xref ref-type="aff" rid="aff-4"/></contrib></contrib-group><aff xml:lang="ru" id="aff-1"><institution>ГрГУ им. Янки Купалы</institution><country>Belarus</country></aff><aff xml:lang="ru" id="aff-2"><institution>Гродненская университетская клиника; Гродненский государственный медицинский университет</institution><country>Belarus</country></aff><aff xml:lang="ru" id="aff-3"><institution>Гродненская университетская клиника</institution><country>Belarus</country></aff><aff xml:lang="ru" id="aff-4"><institution>РУП Белтелеком, Гродненский филиал</institution><country>Belarus</country></aff><pub-date pub-type="collection"><year>2026</year></pub-date><pub-date pub-type="epub"><day>21</day><month>05</month><year>2026</year></pub-date><volume>0</volume><issue>4</issue><fpage>78</fpage><lpage>83</lpage><permissions><copyright-statement>Copyright &amp;#x00A9; Издательский дом «Белорусская наука», 2026</copyright-statement><copyright-year>2026</copyright-year><copyright-holder xml:lang="ru">Издательский дом «Белорусская наука»</copyright-holder><copyright-holder xml:lang="en">Издательский дом «Белорусская наука»</copyright-holder><license xlink:href="https://innosfera.belnauka.by/jour/about/submissions#copyrightNotice" xlink:type="simple"><license-p>https://innosfera.belnauka.by/jour/about/submissions#copyrightNotice</license-p></license></permissions><self-uri xlink:href="https://innosfera.belnauka.by/jour/article/view/1027">https://innosfera.belnauka.by/jour/article/view/1027</self-uri><abstract><p>В работе представлен разработанный авторами метод автоматической сегментации гемангиомы печени на магнитно- резонансных снимках, основанный на модификациях сверточной архитектуры U-Net. Проведено экспериментальное сравнение базовой U-Net, U-Net++ и Attention U-Net. Подтверждено, что модель Attention U-Net с пространственно-канальным механизмом внимания (SCSE) и комбинированной функцией потерь (Dice + Focal Loss) достигает наилучших результатов, демонстрируя средний коэффициент Дайса (DSC) 84,65% на приватном наборе данных. Этот показатель сопоставим с результатами современных аналогов, особенно при работе с данными, характеризующимися значительным дисбалансом классов. Для апробации метода в условиях, приближенных к клиническим, разработан прототип программного обеспечения с графическим интерфейсом и возможностью интеграции в PACS-системы через протокол DICOMweb.</p></abstract><trans-abstract xml:lang="en"><p>This paper presents a method for the automatic segmentation of liver hemangioma in magnetic resonance imaging (MRI) developed by the authors, based on modifications of the U-Net convolutional architecture. An experimental comparison was conducted between the baseline U-Net, U-Net++, and Attention U-Net. It was confirmed that the Attention U-Net model, utilizing a Spatial and Channel Squeeze-and-Excitation (scSE) attention mechanism and a combined loss function (Dice + Focal Loss), achieves the best performance, demonstrating a mean Dice Similarity Coefficient (DSC) of 84.65% on a private dataset. This result is comparable to state-of-the-art benchmarks, particularly when dealing with data characterized by significant class imbalance. To validate the method in conditions approximating clinical practice, a software prototype was developed featuring a graphical user interface and the capability for integration into PACS systems via the DICOMweb protocol.</p></trans-abstract><kwd-group xml:lang="ru"><kwd>сегментация медицинских изображений</kwd><kwd>гемангиома печени</kwd><kwd>магнитно-резонансная томография</kwd><kwd>сверточные нейронные сети</kwd><kwd>глубокое обучение</kwd><kwd>МРТ</kwd><kwd>U-Net</kwd><kwd>Attention U-Net</kwd><kwd>DICOM</kwd><kwd>PACS</kwd></kwd-group><kwd-group xml:lang="en"><kwd>medical image segmentation</kwd><kwd>liver hemangioma</kwd><kwd>magnetic resonance imaging</kwd><kwd>convolutional neural networks</kwd><kwd>deep learning</kwd><kwd>MRI</kwd><kwd>U-Net</kwd><kwd>Attention U-Net</kwd><kwd>DICOM</kwd><kwd>PACS</kwd></kwd-group></article-meta></front><back><ref-list><title>References</title><ref id="cit1"><label>1</label><citation-alternatives><mixed-citation xml:lang="ru">Kacała A. et al, Evaluation of Predictive Factors for Transarterial Bleomycin–Lipiodol Embolization Success in Treating Giant Hepatic Hemangiomas // Cancers. 2025. №17(1). P. 42. Https://doi.org/10.3390/cancers17010042.</mixed-citation><mixed-citation xml:lang="en">Kacała A. et al, Evaluation of Predictive Factors for Transarterial Bleomycin–Lipiodol Embolization Success in Treating Giant Hepatic Hemangiomas // Cancers. 2025. №17(1). P. 42. Https://doi.org/10.3390/cancers17010042.</mixed-citation></citation-alternatives></ref><ref id="cit2"><label>2</label><citation-alternatives><mixed-citation xml:lang="ru">A survey on deep learning in medical image analysis / G. Litjens [et al.] // Medical Image Analysis. 2017. V. 42. P. 60–88.</mixed-citation><mixed-citation xml:lang="en">A survey on deep learning in medical image analysis / G. Litjens [et al.] // Medical Image Analysis. 2017. V. 42. P. 60–88.</mixed-citation></citation-alternatives></ref><ref id="cit3"><label>3</label><citation-alternatives><mixed-citation xml:lang="ru">Albiin N. MRI of Focal Liver Lesions // Curr Med Imaging Rev. 2012. May, 8(2). P. 107–116. Doi: 10.2174/157340512800672216. PMID: 23049491; PMCID: PMC3462338.</mixed-citation><mixed-citation xml:lang="en">Albiin N. MRI of Focal Liver Lesions // Curr Med Imaging Rev. 2012. May, 8(2). P. 107–116. Doi: 10.2174/157340512800672216. PMID: 23049491; PMCID: PMC3462338.</mixed-citation></citation-alternatives></ref><ref id="cit4"><label>4</label><citation-alternatives><mixed-citation xml:lang="ru">Ronneberger O., Fischer P., Brox T. U-Net: Convolutional Networks for Biomedical Image Segmentation // Medical Image Computing and Computer-Assisted Intervention (MICCAI). 2015. P. 234–241.</mixed-citation><mixed-citation xml:lang="en">Ronneberger O., Fischer P., Brox T. U-Net: Convolutional Networks for Biomedical Image Segmentation // Medical Image Computing and Computer-Assisted Intervention (MICCAI). 2015. P. 234–241.</mixed-citation></citation-alternatives></ref><ref id="cit5"><label>5</label><citation-alternatives><mixed-citation xml:lang="ru">Attention U-Net: Learning Where to Look for the Pancreas / O. Oktay [et al.] // Https://doi.org/10.48550/arXiv.1804.03999.</mixed-citation><mixed-citation xml:lang="en">Attention U-Net: Learning Where to Look for the Pancreas / O. Oktay [et al.] // Https://doi.org/10.48550/arXiv.1804.03999.</mixed-citation></citation-alternatives></ref><ref id="cit6"><label>6</label><citation-alternatives><mixed-citation xml:lang="ru">Deep Residual Learning for Image Recognition / K. He [et al.] // CVPR. 2016. P. 770–778. Doi: 10.1109/ CVPR.2016.90.</mixed-citation><mixed-citation xml:lang="en">Deep Residual Learning for Image Recognition / K. He [et al.] // CVPR. 2016. P. 770–778. Doi: 10.1109/ CVPR.2016.90.</mixed-citation></citation-alternatives></ref><ref id="cit7"><label>7</label><citation-alternatives><mixed-citation xml:lang="ru">Coarse-to-Fine Liver Tumor Segmentation using 3D U-Net / H. Wu [et al.] // Medical Image Analysis. 2023. V. 85. P. 102734.</mixed-citation><mixed-citation xml:lang="en">Coarse-to-Fine Liver Tumor Segmentation using 3D U-Net / H. Wu [et al.] // Medical Image Analysis. 2023. V. 85. P. 102734.</mixed-citation></citation-alternatives></ref><ref id="cit8"><label>8</label><citation-alternatives><mixed-citation xml:lang="ru">Multi-modal learning with an anisotropic 3D U-Net for the segmentation of liver tumors in DCE-MRI / A. Hänsch [et al.] // Computerized Medical Imaging and Graphics. 2022. V. 95. P. 102015.</mixed-citation><mixed-citation xml:lang="en">Multi-modal learning with an anisotropic 3D U-Net for the segmentation of liver tumors in DCE-MRI / A. Hänsch [et al.] // Computerized Medical Imaging and Graphics. 2022. V. 95. P. 102015.</mixed-citation></citation-alternatives></ref><ref id="cit9"><label>9</label><citation-alternatives><mixed-citation xml:lang="ru">Improving automatic liver tumor segmentation in late-phase MRI using multi-model training and 3D convolutional neural networks / Hänsch [et al.] // Scientific Reports. 2022. V.12. P. 12262. Https://doi.org/10.1038/s41598-022-16388-9.</mixed-citation><mixed-citation xml:lang="en">Improving automatic liver tumor segmentation in late-phase MRI using multi-model training and 3D convolutional neural networks / Hänsch [et al.] // Scientific Reports. 2022. V.12. P. 12262. Https://doi.org/10.1038/s41598-022-16388-9.</mixed-citation></citation-alternatives></ref><ref id="cit10"><label>10</label><citation-alternatives><mixed-citation xml:lang="ru">Explainable and Robust Deep Learning for Liver Segmentation Through U-Net Network / M. C. Brunese // Diagnostics (Basel). 2025. Mar 31. V. 15(7). P. 878. Doi: 10.3390/diagnostics15070878. PMID: 40218228; PMCID: PMC11989174.</mixed-citation><mixed-citation xml:lang="en">Explainable and Robust Deep Learning for Liver Segmentation Through U-Net Network / M. C. Brunese // Diagnostics (Basel). 2025. Mar 31. V. 15(7). P. 878. Doi: 10.3390/diagnostics15070878. PMID: 40218228; PMCID: PMC11989174.</mixed-citation></citation-alternatives></ref></ref-list><fn-group><fn fn-type="conflict"><p>The authors declare that there are no conflicts of interest present.</p></fn></fn-group></back></article>
