Table 6.

Report standards list of machine learning in clinical medication

Section and topic	Item	Description
Title/Abstract/Keywords	1	Can be judged as a machine learning predictive research. (Keywords,such as machine learning,prediction)
Introduction	2	Introduce background, existing problems, and study targets,such as evaluating machine learning models to predict prognoses and probability of disease occurrence
Method research subject	3	Inclusion and exclusion criteria, locations where data is collected and time range
	4	Describe reasons of patients' selection, including symptoms, laboratorial results, or disease golden standard.
	5	Describe golden standard and provide references
Research data	6	Describe whether study is based on past datasets (retrospective study) or latest collection data (prospective study).
	7	Describe the data collection process.
	8	Describe the process of feature engineering. At least explain why choose this way to select features.
Results Building model	9	Provide flowchart of the including and excluding process, describe demographic and clinical characteristics (such as age, sex, height, and weight)
	10	Describe data preprocessing methods, including missing data processing, and smoothly processing sparse data.
	11	Describe the mathematical theory of the algorithm and its advantages.
	12	Describe numbers and names of finally including features
Research results	13	Describe models performance at different time points (provide at least one evaluation indicator, such as AUROC, accuracy).
Discussion	14	Discuss clinical universality of predictive models, including heterogeneity discussion and clinical prospective validation.