ABCDEFGHIJKLMNOPQRSTUVWXYZ
1
AImedReport
2
Updated: 2/28/2024
3
4
Purpose:
5
Our team sought to create a comprehensive documentation guidance, from which we derived an accompanying AImedReport. The AImedReport consolidates available research reporting items and guidelines, and helps multidisciplinary teams adhere to reporting measures, designate responsibility for different items, and generally provide guidance throughout stages of preparation, development, validation, translation, and maintenance. Components from the CONSORT-AI, DECIDE-AI, ML Test Score, Model Card, SPIRIT-AI, and TRIPOD, reporting guidelines were included in the AImedReport to build the reporting requirements for each phase.
6
7
8
9
10
11
12
How it works:
13
The AImedReport follows the phases Prepare, Develop, Validate, Deploy, and, Maintain and associated subphases of the AI lifecycle to build a comprehensive documentation deliverable, using categorized Reporting Items to guide evaluation and documentation throughout. Research teams can use the Reporting Item: Project column to document project specific information prompted by the Reporting Item: Description column. Each Reporting Item can be marked as complete or incomplete within the Completed? column by choosing the Yes or No option. Teams or team members responsible for each satisying each Reporting Item can be documented within the Responsible, Accountable, Consulted, and Informed columns. We recommend designated Accountable team members mark Complete once reviewed. Please be sure to make a copy of the AImedReport prior to use.
14
15
16
17
18
19
20
21
22
23
Reporting Guidelines Used:
Description:
24
CONSORT-AIAims to promote transparency and completeness in reporting clinical trials for AI interventions, helping to understand, interpret and appraise the quality of clinical trial design and risk of bias in the reported outcomes; focuses on reporting the results of clinical trials
25
DECIDE-AIAims to improve the reporting of studies describing the evaluation of AI-based decision support systems during their early, small-scale implementation in live clinical settings
26
ML Test ScoreAims to measure how ready for production a machine learning system is by offering a scoring system that focuses on assessing testing and monitoring needs
27
Model CardsAims to encourage transparent model reporting, clarifying the intended use cases of models and detailing performance characteristics
28
SPIRIT-AIAims to promote transparent prospective evaluation and completeness of clinical trial protocol reporting for AI interventions; focuses on publishing the clinical trial protocol before the trial is conducted
29
TRIPODAims to improve the transparency and reporting of studies developing, validating, or improving a prediction model
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100