Asif Sohaib, Yan Yuqi, Feng Bojian, Wang Meiling, Zheng Yuxin, Jiang Tian, Fu Ruyi, Yao Jincao, Lv Lujiao, Song Mei, Sui Lin, Yin Zheng, Wang Vicky Yang, Xu Dong
Taizhou Key Laboratory of Minimally Invasive Interventional Therapy & Artificial Intelligence, Taizhou Campus of Zhejiang Cancer Hospital (Taizhou Cancer Hospital), Taizhou, Zhejiang 317502, China (S.A., Y.Y., B.F., L.S., D.X.); Center of Intelligent Diagnosis and Therapy (Taizhou), Hangzhou Institute of Medicine (HIM), Chinese Academy of Sciences, Taizhou, Zhejiang 317502, China (S.A., Y.Y., B.F., L.S., V.Y.W., D.X.); Wenling Institute of Big Data and Artificial Intelligence in Medicine, Taizhou, Zhejiang 317502, China (S.A., Y.Y., B.F., L.S., V.Y.W., D.X.); Department of Diagnostic Ultrasound Imaging & Interventional Therapy, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of Sciences, Hangzhou, Zhejiang 310022, China (S.A., Y.Y., B.F., T.J., J.Y., L.L., M.S., M.S., L.S., V.Y.W., D.X.).
Taizhou Key Laboratory of Minimally Invasive Interventional Therapy & Artificial Intelligence, Taizhou Campus of Zhejiang Cancer Hospital (Taizhou Cancer Hospital), Taizhou, Zhejiang 317502, China (S.A., Y.Y., B.F., L.S., D.X.); Center of Intelligent Diagnosis and Therapy (Taizhou), Hangzhou Institute of Medicine (HIM), Chinese Academy of Sciences, Taizhou, Zhejiang 317502, China (S.A., Y.Y., B.F., L.S., V.Y.W., D.X.); Wenling Institute of Big Data and Artificial Intelligence in Medicine, Taizhou, Zhejiang 317502, China (S.A., Y.Y., B.F., L.S., V.Y.W., D.X.); Department of Diagnostic Ultrasound Imaging & Interventional Therapy, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of Sciences, Hangzhou, Zhejiang 310022, China (S.A., Y.Y., B.F., T.J., J.Y., L.L., M.S., M.S., L.S., V.Y.W., D.X.); Zhejiang Provincial Research Center for Cancer Intelligent Diagnosis and Molecular Technology, Hangzhou 310022, China (Y.Y., T.J., J.Y., L.S., D.X.).
Acad Radiol. 2025 May 27. doi: 10.1016/j.acra.2025.05.007.
RATIONALE AND OBJECTIVES: Early detection of malignant lesions in ultrasound images is crucial for effective cancer diagnosis and treatment. While traditional methods rely on radiologists, deep learning models can improve accuracy, reduce errors, and enhance efficiency. This study explores the application of a deep learning model for classifying benign and malignant lesions, focusing on its performance and interpretability. MATERIALS AND METHODS: In this study, we proposed a feature fusion-based deep learning model for classifying benign and malignant lesions in ultrasound images. The model leverages advanced architectures such as MobileNetV2 and DenseNet121, enhanced with feature fusion and attention mechanisms to boost classification accuracy. The clinical dataset comprises 2171 images collected from 1758 patients between December 2020 and May 2024. Additionally, we utilized the publicly available BUSI dataset, consisting of 780 images from female patients aged 25 to 75, collected in 2018. To enhance interpretability, we applied Grad-CAM, Saliency Maps, and shapley additive explanations (SHAP) techniques to explain the model's decision-making. A comparative analysis with radiologists of varying expertise levels is also conducted. RESULTS: The proposed model exhibited the highest performance, achieving an AUC of 0.9320 on our private dataset and an area under the curve (AUC) of 0.9834 on the public dataset, significantly outperforming traditional deep convolutional neural network models. It also exceeded the diagnostic performance of radiologists, showcasing its potential as a reliable tool for medical image classification. The model's success can be attributed to its incorporation of advanced architectures, feature fusion, and attention mechanisms. The model's decision-making process was further clarified using interpretability techniques like Grad-CAM, Saliency Maps, and SHAP, offering insights into its ability to focus on relevant image features for accurate classification. CONCLUSION: The proposed deep learning model offers superior accuracy in classifying benign and malignant lesions in ultrasound images, outperforming traditional models and radiologists. Its strong performance, coupled with interpretability techniques, demonstrates its potential as a reliable and efficient tool for medical diagnostics. DATA AVAILABILITY: The datasets generated and analyzed during the current study are not publicly available due to the nature of this research and participants of this study, but may be available from the corresponding author on reasonable request.
原理与目标：超声图像中恶性病变的早期检测对于有效的癌症诊断和治疗至关重要。虽然传统方法依赖放射科医生，但深度学习模型可以提高准确性、减少误差并提高效率。本研究探讨了一种深度学习模型在良性和恶性病变分类中的应用，重点关注其性能和可解释性。 材料与方法：在本研究中，我们提出了一种基于特征融合的深度学习模型，用于对超声图像中的良性和恶性病变进行分类。该模型利用了诸如MobileNetV2和DenseNet121等先进架构，并通过特征融合和注意力机制进行增强，以提高分类准确性。临床数据集包含2020年12月至2024年5月期间从1758名患者收集的2171张图像。此外，我们使用了公开可用的BUSI数据集，该数据集由2018年收集的780张25至75岁女性患者的图像组成。为了提高可解释性，我们应用了Grad-CAM、显著性图和Shapley值加法解释（SHAP）技术来解释模型的决策过程。还对不同专业水平的放射科医生进行了对比分析。 结果：所提出的模型表现出最高的性能，在我们的私有数据集上的AUC为0.9320，在公共数据集上的曲线下面积（AUC）为0.9834，显著优于传统的深度卷积神经网络模型。它还超过了放射科医生的诊断性能，展示了其作为医学图像分类可靠工具的潜力。该模型的成功可归因于其采用了先进架构、特征融合和注意力机制。使用Grad-CAM、显著性图和SHAP等可解释性技术进一步阐明了模型的决策过程，深入了解了其专注于相关图像特征以进行准确分类的能力。 结论：所提出的深度学习模型在超声图像中良性和恶性病变的分类方面具有卓越的准确性，优于传统模型和放射科医生。其强大的性能以及可解释性技术证明了它作为医学诊断可靠且高效工具的潜力。 数据可用性：由于本研究的性质和本研究的参与者，在当前研究期间生成和分析的数据集不公开，但可根据合理请求从相应作者处获得。
Front Bioeng Biotechnol. 2025-6-25
Comput Biol Med. 2024-9
Zotero 插件