Table 2.

Key summary of artificial intelligence studies (N=153) analyzed (n=129 articles).

Mental health disorder (ICD-11a) section number: category	Artificial intelligence model family	Data type
6A0: Neurodevelopmental disorders	Regression [14]	Mixedb [14]
6A2: Schizophrenia or other primary psychotic disorders	Competing modelsc [15-17] Ensembled models [18,19] Regression [20-23] Statistical clustering [24] SVMd [25-28] Trees [29-31] Regression and statistical clustering [32] Regression and hierarchical clustering [33]	Genomic data [29] Medical image [15-19,21-23,27,28,33] Mixed [20,26,30,32] Questionnaires and scales [24,25,31]
6A6: Bipolar or related disorders	Bayesian [34] Manifold [35] Regression [36] SVM [37]	Medical image [35,37] Mixed [36] Questionnaires and scales [34]
6A7: Depressive disorders	Bayesian [38-40] Competing models [41-53] Deep learning [54] Ensembled models [55-57] Hierarchical clustering [58,59] Markov model [60] Mixture model [61] Mixture model, regression, and trees [62] Regression [63-74] Relevance vector machine [75] Statistical learning [76,77] SVM [78-86] Trees [87-93] Trees and hierarchical clustering [94] Trees and statistical learning [95]	Audio recording [64] Biomarkers [51,54,95] Biosignal [41] Ie [80,81,87,93] Genomic data [44,70,74] Medical image [38,40,48,67-69,75,82,85] Mixed [42,43,45,46,49,50,52,53,55-57,59, 61,63,65,66,71-73,77,79,83,86,88-92,94] Questionnaires and scales [47,58,60,62,76] Text [39] Video image [78,84]
6A6, 6A7: Mood episodes	Competing models [96] SVM [97] Regression [98]	Medical image [97,98] Questionnaires and scales [96]
6B0: Anxiety or fear-related disorders	Competing models [99-101] Regression [102] SVM [103] Trees [104]	Biomarkers [101] Biosignal [100] Mixed [99,103,104] Text [102]
6B2: Obsessive-compulsive or related disorders	Competing models [105]	I [105]
6B4: Disorders specifically associated with stress	Competing models [106] Ensembled models [107,108] Hierarchical clustering [109] Mixture model and regression [110] SVM [111-115] Trees [116]	Audio recording [116] Biosignal [115] Medical image [112-114] Mixed [106-108,110,111] Questionnaires and scales [109]
6B8: Feeding or eating disorders	Competing models [117]	Mixed [117]
6C2: Bodily distress disorder	Regression [118]	Mixed [118]
6C4: Disorders due to substance use	Competing models [119] Regression [120] Trees [121,122]	Medical image [120] Mixed [119,121,122]
6D1: Personality disorders	Trees [123]	Mixed [123]
6D7, 6D8: Neurocognitive disorders and dementia	Competing models [124] Ensembled models [125,126] Trees [127]	Mixed [124-127]
6E2: Mental or behavioral disorders associated with pregnancy, childbirth, or the puerperium, without psychotic symptoms	Competing models [128]	Mixed [128]
6E6: Secondary mental or behavioral syndromes associated with disorders or diseases classified elsewhere	Bayesian [129] Competing models [130] Regression [131] Trees [132,133]	Mixed [130-133] Questionnaires and scales [129]
Combination of some ICD-11 categories in mental health	Bayesian [134] Ensembled models [135-137] Regression [138,139] Regression and support vector machines [140] Trees [141]	Biosignal [140] Mixed [135-137,139] Questionnaires and scales [134,138,141]
Unspecifiedf	Competing models [142]	Mixed [142]

^aICD-11: International Classification of Diseases 11th Revision.

^bMixed: combination of type of data and predictors.

^cCompeting models: the study was designed for evaluate several types of artificial intelligence model families without assessing a priori adequacy.

^dSVM: support vector machine.

^eI: electronic health records.

^fUnspecified: The outcome of the study is “mental health problems,” therefore, it could not be classified in any specific category.