Lucena 2021.

Study characteristics
Patient Sampling	Retrospective, case‐control study. A diagnostic pattern bank generated by a specialist physician. The database consisted of 1172 examples of corneal topography, divided into 275 spherical patterns, 302 regular symmetrical astigmatism patterns, 295 regular asymmetrical astigmatism patterns, and 300 irregular astigmatism patterns (keratoconus).
Patient characteristics and setting	This study is a registry‐based study; the registry contained healthy controls and people with keratoconus.
Index tests	Convolutional neural network. The algorithm is a semi‐automatic, manual interference model. It uses a hierarchical system that tries to represent the structure in relation to the recognition of an image, where pixels form edges, edges form patterns, patterns form objects, which in turn describe the scenes. The algorithm analyses the topographic images and decides whether it is regular or irregular astigmatism (keratoconus). The algorithm was developed with a training phase and a validation phase.
Target condition and reference standard(s)	A specialist physician developed the diagnostic pattern bank of images made by topographers. He divided the included topographies into the following 4 groups. Spherical patterns Regular symmetrical astigmatism patterns Regular asymmetrical astigmatism patterns Irregular astigmatism patterns (keratoconus) The cases were divided before the algorithm analysed the images.
Flow and timing	All topographies included in the diagnostic pattern bank were judged by the specialist physician and all images were included in the analysis.
Comparative	Not applicable
Notes	This study was supported by Government of the State of Ceará (Foundation for the Support to Scientific and Technological Development of Ceará), and Ophthalmology School of Ceará.
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient selection
Was a consecutive or random sample of patients enrolled?	Unclear
Was a case‐control design avoided?	No
Did the study avoid inappropriate exclusions?	No
Could the selection of patients have introduced bias?		High risk
Are there concerns that the included patients and setting do not match the review question?			High
DOMAIN 2: Index test (All tests)
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	Unclear
Was the model designed in an appropriate manner?	Unclear
Could the conduct or interpretation of the index test have introduced bias?		Unclear risk
Are there concerns that the index test, its conduct, or interpretation differ from the review question?			High
DOMAIN 3: Reference standard
Is the reference standard likely to correctly classify the target condition?	No
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
Could the reference standard, its conduct, or its interpretation have introduced bias?		High risk
Are there concerns that the target condition as defined by the reference standard does not match the question?			Unclear
DOMAIN 4: Flow and timing
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
Could the patient flow have introduced bias?		Low risk
DOMAIN 5: Comparative