Scholarlink

<p>Ovarian cysts are common conditions in women, potentially leading to infertility, malignancy, or torsion if not identified and treated early. Artificial intelligence (AI) has proven effective in medical imaging; however, most research focuses on distinguishing between benign and malignant cysts. it is equally important to classify specific cyst types to aid in ovarian cancer prevention. This study aims to develop an efficient model to classify seven benign cyst types—Polycystic Ovary Syndrome (PCOS), Serous, Carcinoma, Dermoid, Hemorrhagic, Complex, and Simple cysts using ultrasound images. A novel deep transfer learning model, EfficientOvaNet, was proposed and trained on ultrasound images of ovarian cysts. Diverse data augmentation techniques were applied to address class imbalance issues. The model’s performance was benchmarked against established deep learning architectures, including VGG16, VGG19, ResNet50, DenseNet201, and InceptionV3, using metrics such as accuracy, precision, recall, and F1-score. EfficientOvaNet achieved superior performance across all evaluated metrics, outperforming state-of-the-art models in accurately identifying all seven types of ovarian cysts. The model demonstrated high reliability and robustness, addressing class imbalance effectively and ensuring accurate predictions. EfficientOvaNet offers a promising solution for accurate classification of ovarian cysts, contributing to early diagnosis, improved treatment strategies, and ovarian cancer prevention. This study highlights the model’s potential for real-world applications in medical imaging and its capability to overcome significant challenges such as data imbalance.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

EfficientOvaNet: efficient deep learning model for multiclass classification of benign ovarian cyst using ultrasound images

  • 期刊论文

  • 发表日期：

    2025年9月21日

查看全文
  • Aarti Parekh,
  • Madhavi Desai

摘要

Ovarian cysts are common conditions in women, potentially leading to infertility, malignancy, or torsion if not identified and treated early. Artificial intelligence (AI) has proven effective in medical imaging; however, most research focuses on distinguishing between benign and malignant cysts. it is equally important to classify specific cyst types to aid in ovarian cancer prevention. This study aims to develop an efficient model to classify seven benign cyst types—Polycystic Ovary Syndrome (PCOS), Serous, Carcinoma, Dermoid, Hemorrhagic, Complex, and Simple cysts using ultrasound images. A novel deep transfer learning model, EfficientOvaNet, was proposed and trained on ultrasound images of ovarian cysts. Diverse data augmentation techniques were applied to address class imbalance issues. The model’s performance was benchmarked against established deep learning architectures, including VGG16, VGG19, ResNet50, DenseNet201, and InceptionV3, using metrics such as accuracy, precision, recall, and F1-score. EfficientOvaNet achieved superior performance across all evaluated metrics, outperforming state-of-the-art models in accurately identifying all seven types of ovarian cysts. The model demonstrated high reliability and robustness, addressing class imbalance effectively and ensuring accurate predictions. EfficientOvaNet offers a promising solution for accurate classification of ovarian cysts, contributing to early diagnosis, improved treatment strategies, and ovarian cancer prevention. This study highlights the model’s potential for real-world applications in medical imaging and its capability to overcome significant challenges such as data imbalance.