基于联合临床指标的机器学习算法,用于预测不孕症和怀孕流产
Rui Zhang1, Yuanbing Guo1, Xiaonan Zhai1
1Department of Clinical Laboratory Medicine, Xijing Hospital, Fourth Military Medical University, Xi'an, China.
Frontiers in endocrinology
|August 4, 2025
概括
这项研究开发了高效的机器学习模型,用于诊断不孕症和怀孕流产. 低维生素D水平是关键指标,有助于早期发现和治疗.
科学领域:
- 生殖医学 生殖医学
- 生物统计学 生物统计学
- 医疗保健中的机器学习
背景情况:
- 不孕不育和怀孕丧失的诊断是复杂的.
- 需要更简单,更有效的诊断系统.
研究的目的:
- 开发和验证用于诊断不育和怀孕流产的机器学习模型.
- 确定这些疾病的关键临床指标.
主要方法:
- 使用三种方法选了100多个临床指标.
- 开发了使用五种机器学习算法的诊断模型.
- 在大型患者队列上验证的模型.
主要成果:
- 在患者中,25-氧维生素D3缺乏是突出的因素.
- 不孕症模型实现AUC>0.958,灵敏度>86.52%,特异性>91.23%.
- 怀孕损失模型实现的AUC>0.972,灵敏度>92.02%,特异性>95.18%.
结论:
- 开发的模型提供了简单性和高诊断性能.
- 这些模型可以促进早期检测,治疗和预防不孕症和流产.
更多相关视频
相关概念视频
Infertility in Females
402
Female infertility is defined as the inability to conceive after a year of regular, unprotected intercourse and affects about 10–15% of couples worldwide. The primary cause of female infertility is ovulatory disorders, which hinder the release of eggs. These disorders can be classified as hypothalamic amenorrhea, polycystic ovarian syndrome (PCOS), premature ovarian failure, and hyperprolactinemic anovulation disorders.
Endometriosis, a condition characterized by abnormal growth of...
Endometriosis, a condition characterized by abnormal growth of...
402
Infertility in Males
333
Male infertility affects millions of couples worldwide, arising from various factors that impact different stages of the reproductive process. An endocrine imbalance resulting from conditions like hypogonadism, Klinefelter syndrome, or pituitary disorders can disrupt hormone levels and reduce sperm production. Testicular defects, such as tumors, cryptorchidism, atrophic testes, abnormal sperm morphology, and low sperm count or motility, may arise due to genetic factors, structural...
333


