Scholarlink

<p>PCOS is a hormonal disorder that primarily results in female infertility in women who are of childbearing age. Increased body hair, irregular periods, increased acne, and obesity are common signs of PCOS. To control the symptoms and lower the hazards to one’s health, PCOS must be detected early. The diagnosis is made using the Rotterdam criteria, which include polycystic ovaries on ultrasound imaging, ovulation failure, and a high level of androgen hormones. PCOS identification is a challenging PCOS diagnostic criterion. To manually identify PCOSs, doctors and radiologists currently employ ovarian ultrasonography to count the number of follicles and assess their volume in the ovaries. In addition to the patient’s medical complaints, these healthcare professionals must do additional tests and examinations for biochemical and clinical signs in order to identify PCOS. Two adapted deep learning architectures like a feature-sequence convolutional network and a gated feature-learning network are employed as base learners to automatically extract discriminative patterns from tabular clinical variables. After that, this work created a deep learning model with the TextNet (TxNet) and GateNet (GNet) with a 97% accuracy rate in identifying the PCOS from the provided. Then, this work suggested a hybrid model that uses Clinical data to ascertain if a patient has polycystic ovary syndrome (PCOS). By utilizing the TxNet and GNet architecture to extract the clinical features and combining them with PCOS information, the best model that has been constructed attained an accuracy of 96%.</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

An automated PCOS prediction model using hybrid text and gate network model using learning approaches

  • 期刊论文

  • 发表日期：

    2026年3月17日

查看全文
  • A. Smithakranthi,
  • D. Haritha

摘要

PCOS is a hormonal disorder that primarily results in female infertility in women who are of childbearing age. Increased body hair, irregular periods, increased acne, and obesity are common signs of PCOS. To control the symptoms and lower the hazards to one’s health, PCOS must be detected early. The diagnosis is made using the Rotterdam criteria, which include polycystic ovaries on ultrasound imaging, ovulation failure, and a high level of androgen hormones. PCOS identification is a challenging PCOS diagnostic criterion. To manually identify PCOSs, doctors and radiologists currently employ ovarian ultrasonography to count the number of follicles and assess their volume in the ovaries. In addition to the patient’s medical complaints, these healthcare professionals must do additional tests and examinations for biochemical and clinical signs in order to identify PCOS. Two adapted deep learning architectures like a feature-sequence convolutional network and a gated feature-learning network are employed as base learners to automatically extract discriminative patterns from tabular clinical variables. After that, this work created a deep learning model with the TextNet (TxNet) and GateNet (GNet) with a 97% accuracy rate in identifying the PCOS from the provided. Then, this work suggested a hybrid model that uses Clinical data to ascertain if a patient has polycystic ovary syndrome (PCOS). By utilizing the TxNet and GNet architecture to extract the clinical features and combining them with PCOS information, the best model that has been constructed attained an accuracy of 96%.