机器学习用于预测类风湿性关节炎治疗反应:系统性审查
Claudia Mendoza-Pinto1, Marcial Sánchez-Tecuatl2, Roberto Berra-Romani3
1Department of Rheumatology, Medicine School, Benemérita Universidad Autónoma de Puebla, Mexico; Rheumatology and Autoimmune Diseases Research Unit, Specialties Hospital UMAE-CIBIOR, Instituto Mexicano del Seguro Social, Puebla, Mexico.
Seminars in arthritis and rheumatism
|September 3, 2024
概括
机器学习 (ML) 模型对预测类风湿性关节炎 (RA) 治疗反应充满希望,但大多数研究存在不清楚或高偏差风险. 提高透明度和解决偏见对于在RA护理中可靠的ML预测至关重要.
科学领域:
- 类风湿病学 类风湿病学
- 人工智能的人工智能
- 生物统计学 生物统计学
背景情况:
- 机器学习 (ML) 越来越多地用于预测类风湿性关节炎 (RA) 治疗反应.
- 一项系统性审查评估了ML模型的性能和可重复性,用于预测RA治疗反应.
- 评估这些预测模型中的质量和偏差风险对于临床实用性至关重要.
结论:
- 大多数用于预测RA治疗反应的ML模型都存在不清楚或高偏差风险.
- 建议通过校准措施提高透明度,并明确报告缺失的数据处理.
- 进一步的研究应侧重于提高方法论的严格性和减少RAML模型中的偏差.
相关概念视频
01:20The JAK-STAT Signaling Pathway
8.7K
Several cytokine receptors have tightly bound Janus kinase or JAK proteins attached at their cytosolic tail. Small signaling molecules such as cytokines, growth hormones, or prolactins bind to the cytokine receptors and initiate their dimerization. The dimerization brings the cytosolic JAKs together that trans-phosphorylate and activates each other. The activated JAKs now phosphorylate cytosolic tails of the cytokine receptors, which serve as binding sites for adaptor proteins such as SH2...
8.7K
01:24T Cell Types and Functions
963
When T cells with CD4 markers are activated, they give rise to two types of effector cells: helper T cells and regulatory T cells. Meanwhile, T cells with CD8 markers differentiate into effector cytotoxic T cells. The differentiation of CD4 T cells into helper T cell subsets, such as Th1, Th2, and Th17 cells, is dependent on the antigen type, antigen-presenting cell, and regulatory cytokines.
Th1 cells stimulate dendritic cells to express necessary co-stimulatory molecules on their surfaces for...
Th1 cells stimulate dendritic cells to express necessary co-stimulatory molecules on their surfaces for...
963