Focusing on AI hematology prediction model and paying close attention to process quality and efficiency
10.3760/cma.j.cn114452-20250217-00086
- VernacularTitle:聚焦AI血液病预测模型 关注模型构建与效能
- Author:
Linlin QU
1
;
Wei XU
1
Author Information
1. 吉林大学第一医院检验科，长春 130021
- Publication Type:Journal Article
- Keywords:
Artificial intelligence;
Machine learning;
Data;
Hematological disease;
Model;
Medical lab
- From:
Chinese Journal of Laboratory Medicine
2025;48(5):543-548
- CountryChina
- Language:Chinese
-
Abstract:
Many hematological disease prediction models built by big data analysis in medical laboratory and artificial intelligence (AI) technology, especially machine learning (ML) methods are reported more and more widely in researches on disease alerting, screening, diagnosis, discrimination, treatment and prediction. Because of a lack of guidance of standards, guidelines and normative documents in standardizing the construction and validation of these models, there is deficiency in model function, data preparation, screening and labeling, algorithm selection, model evaluation, as well as local and overall interpretation after model construction, which also limit the clinical application. Therefore, with integration of the literature review, combining the practice on blood disease prediction models, this article proposes that some keypoints and efficiency should be noticed, in order to build hematological diseases related models with clinical practical application value.
