一个集成的优化和深度学习管道,用于预测IVF活产成功,使用特征优化和基于变压器的模型
Arezoo Borji1, Hossam Haick2, Birgit Pohn3
1Austrian Center for Medical Innovation and Technology, Wiener Neustadt, Austria; Department of Medicine, Danube Private University (DPU), Krems, Austria; Department of Medical Physics and Biomedical Engineering, Medical University of Vienna, Vienna, Austria.
Computer methods and programs in biomedicine
|July 30, 2025
概括
这项研究开发了一种人工智能 (AI) 管道,以准确预测体外受精 (IVF) 治疗中活产的结果. 人工智能模型实现了97%的准确性,为个性化生育护理提供了潜力.
科学领域:
- 生殖医学 生殖医学
- 医疗保健中的人工智能
- 生物医学数据科学 生物医学数据科学
背景情况:
- 由于复杂的临床,人口和程序因素,预测体外受精 (IVF) 的成功具有挑战性.
- 准确预测活产结果对于优化辅助生殖技术治疗至关重要.
研究的目的:
- 开发一个高度准确的人工智能 (AI) 管道,用于预测试管婴儿中活产的结果.
- 提高用于生育治疗预测的AI模型的可解释性.
主要方法:
- 评估了各种特征选择方法 (PCA,PSO) 和机器学习分类器 (RF,决策树,变压器,Tab_transformer).
- 分析了混因素 (年龄,以前的周期) 和预处理技术.
- 为了模型的可解释性,使用了沙普利增量解释 (SHAP).
主要成果:
- 对特征选择的粒子优化 (PSO) 和Tab_transformer深度学习模型的组合实现了97%的准确性和98.4%的AUC.
- SHAP分析确定了不孕不育的关键预测因素,并改善了模型的解释性.
结论:
- 开发了一个强大的AI管道,以高准确性和可解释性预测试管婴儿活产的结果.
- 人工智能管道显示了增强个性化生育治疗和改善患者护理的潜力.
相关概念视频
In Vitro Fertilization
395
In vitro fertilization (IVF) is a form of assisted reproductive technology where an egg is fertilized with sperm in a controlled laboratory environment before transferring the resulting embryo into the uterus. This process is designed to help individuals and couples experiencing difficulties conceiving.
The IVF process begins with ovarian stimulation, during which reproductive endocrinologists prescribe hormonal medications to stimulate the ovaries to produce multiple eggs instead of the single...
The IVF process begins with ovarian stimulation, during which reproductive endocrinologists prescribe hormonal medications to stimulate the ovaries to produce multiple eggs instead of the single...
395
Improving Translational Accuracy
11.9K
Base complementarity between the three base pairs of mRNA codon and the tRNA anticodon is not a failsafe mechanism. Inaccuracies can range from a single mismatch to no correct base pairing at all. The free energy difference between the correct and nearly correct base pairs can be as small as 3 kcal/ mol. With complementarity being the only proofreading step, the estimated error frequency would be one wrong amino acid in every 100 amino acids incorporated. However, error frequencies observed in...
11.9K


