|
1
|
Pigoni A, Delvecchio G, Turtulici N, Madonna D, Pietrini P, Cecchetti L, Brambilla P. Machine learning and the prediction of suicide in psychiatric populations: a systematic review. Transl Psychiatry 2024; 14:140. [PMID: 38461283 PMCID: PMC10925059 DOI: 10.1038/s41398-024-02852-9] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 06/05/2023] [Revised: 02/22/2024] [Accepted: 02/22/2024] [Indexed: 03/11/2024] Open
Abstract
Machine learning (ML) has emerged as a promising tool to enhance suicidal prediction. However, as many large-sample studies mixed psychiatric and non-psychiatric populations, a formal psychiatric diagnosis emerged as a strong predictor of suicidal risk, overshadowing more subtle risk factors specific to distinct populations. To overcome this limitation, we conducted a systematic review of ML studies evaluating suicidal behaviors exclusively in psychiatric clinical populations. A systematic literature search was performed from inception through November 17, 2022 on PubMed, EMBASE, and Scopus following the PRISMA guidelines. Original research using ML techniques to assess the risk of suicide or predict suicide attempts in the psychiatric population were included. An assessment for bias risk was performed using the transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD) guidelines. About 1032 studies were retrieved, and 81 satisfied the inclusion criteria and were included for qualitative synthesis. Clinical and demographic features were the most frequently employed and random forest, support vector machine, and convolutional neural network performed better in terms of accuracy than other algorithms when directly compared. Despite heterogeneity in procedures, most studies reported an accuracy of 70% or greater based on features such as previous attempts, severity of the disorder, and pharmacological treatments. Although the evidence reported is promising, ML algorithms for suicidal prediction still present limitations, including the lack of neurobiological and imaging data and the lack of external validation samples. Overcoming these issues may lead to the development of models to adopt in clinical practice. Further research is warranted to boost a field that holds the potential to critically impact suicide mortality.
Collapse
Affiliation(s)
- Alessandro Pigoni
- Social and Affective Neuroscience Group, MoMiLab, IMT School for Advanced Studies Lucca, Lucca, Italy
- Department of Neurosciences and Mental Health, Fondazione IRCCS Ca' Granda, Ospedale Maggiore Policlinico, Milan, Italy
| | - Giuseppe Delvecchio
- Department of Neurosciences and Mental Health, Fondazione IRCCS Ca' Granda, Ospedale Maggiore Policlinico, Milan, Italy
| | - Nunzio Turtulici
- Department of Pathophysiology and Transplantation, University of Milan, Milan, Italy
| | - Domenico Madonna
- Department of Neurosciences and Mental Health, Fondazione IRCCS Ca' Granda, Ospedale Maggiore Policlinico, Milan, Italy
| | - Pietro Pietrini
- MoMiLab, IMT School for Advanced Studies Lucca, Lucca, Italy
| | - Luca Cecchetti
- Social and Affective Neuroscience Group, MoMiLab, IMT School for Advanced Studies Lucca, Lucca, Italy
| | - Paolo Brambilla
- Department of Neurosciences and Mental Health, Fondazione IRCCS Ca' Granda, Ospedale Maggiore Policlinico, Milan, Italy.
- Department of Pathophysiology and Transplantation, University of Milan, Milan, Italy.
| |
Collapse
|
|
2
|
Sun L, Li K, Zhang Y, Zhang L. Carbon Monoxide Poisoning was Associated With Lifetime Suicidal Ideation: Evidence From A Population-Based Cross-Sectional Study in Hebei Province, China. Int J Public Health 2022; 67:1604462. [PMID: 35783447 PMCID: PMC9240916 DOI: 10.3389/ijph.2022.1604462] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/11/2021] [Accepted: 05/24/2022] [Indexed: 11/13/2022] Open
Abstract
Objectives: We want to test the association between carbon monoxide poisoning (CMP) experiencing and lifetime suicidal ideation/suicide plan among community residents. Methods: This is a population-based cross-sectional study conducted among community residents in Hebei province, China. We analyzed a total of 21,376 valid questionnaires. CMP experience and lifetime suicidal ideation/suicide plan were assessed in this study. Logistic regression and false discovery rate correction were conducted to analyze the associations and correct the p values. Results: We found that CMP (OR = 2.56, p < 0.001, corrected-p = 0.001) was associated with lifetime suicidal ideation, and the other risk factors were female (OR = 0.53, p < 0.001, corrected-p = 0.001). The association between CMP and suicide plan was not supported after false discovery rate correction (OR = 2.15, p = 0.035, corrected-p = 0.385). For the CMP patients, experiencing ≥2 times CMP (OR = 2.76, p = 0.001, corrected-p = 0.011) was also in higher risk of lifetime suicidal ideation. The association between CMP times and lifetime suicidal plan was not supported after false discovery rate correction (OR = 4.95, p = 0.021, corrected-p = 0.231). Conclusion: CMP patients are in higher risk of lifetime suicidal ideation. For CMP patients, some strategies are needed to control their suicidal ideation.
Collapse
Affiliation(s)
- Long Sun
- Centre for Health Management and Policy Research, School of Public Health, Cheeloo College of Medicine, Shandong University, Jinan, China
- National Health Commission of China Key Lab for Health Economics and Policy Research (Shandong University), Jinan, China
| | - Keqing Li
- The Sixth People Hospital of Hebei Province, Baoding, China
| | - Yunshu Zhang
- The Sixth People Hospital of Hebei Province, Baoding, China
| | - Lili Zhang
- The Sixth People Hospital of Hebei Province, Baoding, China
| |
Collapse
|
|
3
|
Kouter K, Videtic Paska A. 'Omics' of suicidal behaviour: A path to personalised psychiatry. World J Psychiatry 2021; 11:774-790. [PMID: 34733641 PMCID: PMC8546767 DOI: 10.5498/wjp.v11.i10.774] [Citation(s) in RCA: 3] [Impact Index Per Article: 0.8] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 02/27/2021] [Revised: 07/16/2021] [Accepted: 08/30/2021] [Indexed: 02/06/2023] Open
Abstract
Psychiatric disorders, including suicide, are complex disorders that are affected by many different risk factors. It has been estimated that genetic factors contribute up to 50% to suicide risk. As the candidate gene approach has not identified a gene or set of genes that can be defined as biomarkers for suicidal behaviour, much is expected from cutting edge technological approaches that can interrogate several hundred, or even millions, of biomarkers at a time. These include the '-omic' approaches, such as genomics, transcriptomics, epigenomics, proteomics and metabolomics. Indeed, these have revealed new candidate biomarkers associated with suicidal behaviour. The most interesting of these have been implicated in inflammation and immune responses, which have been revealed through different study approaches, from genome-wide single nucleotide studies and the micro-RNA transcriptome, to the proteome and metabolome. However, the massive amounts of data that are generated by the '-omic' technologies demand the use of powerful computational analysis, and also specifically trained personnel. In this regard, machine learning approaches are beginning to pave the way towards personalized psychiatry.
Collapse
Affiliation(s)
- Katarina Kouter
- Faculty of Medicine, Institute of Biochemistry and Molecular Genetics, University of Ljubljana, Ljubljana SI-1000, Slovenia
| | - Alja Videtic Paska
- Faculty of Medicine, Institute of Biochemistry and Molecular Genetics, University of Ljubljana, Ljubljana SI-1000, Slovenia
| |
Collapse
|
|
4
|
Paska AV, Kouter K. Machine learning as the new approach in understanding biomarkers of suicidal behavior. Bosn J Basic Med Sci 2021; 21:398-408. [PMID: 33485296 PMCID: PMC8292863 DOI: 10.17305/bjbms.2020.5146] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/14/2020] [Accepted: 12/15/2020] [Indexed: 11/16/2022] Open
Abstract
In psychiatry, compared to other medical fields, the identification of biological markers that would complement current clinical interview, and enable more objective and faster clinical diagnosis, implement accurate monitoring of treatment response and remission, is grave. Current technological development enables analyses of various biological marks in high throughput scale at reasonable costs, and therefore 'omic' studies are entering the psychiatry research. However, big data demands a whole new plethora of skills in data processing, before clinically useful information can be extracted. So far the classical approach to data analysis did not really contribute to identification of biomarkers in psychiatry, but the extensive amounts of data might get to a higher level, if artificial intelligence in the shape of machine learning algorithms would be applied. Not many studies on machine learning in psychiatry have been published, but we can already see from that handful of studies that the potential to build a screening portfolio of biomarkers for different psychopathologies, including suicide, exists.
Collapse
Affiliation(s)
- Alja Videtič Paska
- Medical Centre for Molecular Biology, Institute of Biochemistry, Faculty of Medicine, University of Ljubljana, Ljubljana, Slovenia
| | - Katarina Kouter
- Medical Centre for Molecular Biology, Institute of Biochemistry, Faculty of Medicine, University of Ljubljana, Ljubljana, Slovenia
| |
Collapse
|
|
5
|
Molecular characterization of the stress network in individuals at risk for schizophrenia. Neurobiol Stress 2021; 14:100307. [PMID: 33644266 PMCID: PMC7893486 DOI: 10.1016/j.ynstr.2021.100307] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/09/2020] [Revised: 01/14/2021] [Accepted: 02/03/2021] [Indexed: 01/24/2023] Open
Abstract
The biological mechanisms underlying inter-individual differences in human stress reactivity remain poorly understood. We aimed to identify the molecular underpinning of aberrant neural stress sensitivity in individuals at risk for schizophrenia. Linking mRNA expression data from the Allen Human Brain Atlas to task-based fMRI revealed 201 differentially expressed genes in cortex-specific brain regions differentially activated by stress in individuals with low (healthy siblings of schizophrenia patients) or high (healthy controls) stress sensitivity. These genes are associated with stress-related psychiatric disorders (e.g. schizophrenia and anxiety) and include markers for specific neuronal populations (e.g. ADCYAP1, GABRB1, SSTR1, and TNFRSF12A), neurotransmitter receptors (e.g. GRIN3A, SSTR1, GABRB1, and HTR1E), and signaling factors that interact with the corticosteroid receptor and hypothalamic-pituitary-adrenal axis (e.g. ADCYAP1, IGSF11, and PKIA). Overall, the identified genes potentially underlie altered stress reactivity in individuals at risk for schizophrenia and other psychiatric disorders and play a role in mounting an adaptive stress response in at-risk individuals, making them potentially druggable targets for stress-related diseases.
Collapse
|
|
6
|
Bernert RA, Hilberg AM, Melia R, Kim JP, Shah NH, Abnousi F. Artificial Intelligence and Suicide Prevention: A Systematic Review of Machine Learning Investigations. INTERNATIONAL JOURNAL OF ENVIRONMENTAL RESEARCH AND PUBLIC HEALTH 2020; 17:E5929. [PMID: 32824149 PMCID: PMC7460360 DOI: 10.3390/ijerph17165929] [Citation(s) in RCA: 88] [Impact Index Per Article: 17.6] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Received: 07/20/2020] [Accepted: 07/28/2020] [Indexed: 12/12/2022]
Abstract
Suicide is a leading cause of death that defies prediction and challenges prevention efforts worldwide. Artificial intelligence (AI) and machine learning (ML) have emerged as a means of investigating large datasets to enhance risk detection. A systematic review of ML investigations evaluating suicidal behaviors was conducted using PubMed/MEDLINE, PsychInfo, Web-of-Science, and EMBASE, employing search strings and MeSH terms relevant to suicide and AI. Databases were supplemented by hand-search techniques and Google Scholar. Inclusion criteria: (1) journal article, available in English, (2) original investigation, (3) employment of AI/ML, (4) evaluation of a suicide risk outcome. N = 594 records were identified based on abstract search, and 25 hand-searched reports. N = 461 reports remained after duplicates were removed, n = 316 were excluded after abstract screening. Of n = 149 full-text articles assessed for eligibility, n = 87 were included for quantitative synthesis, grouped according to suicide behavior outcome. Reports varied widely in methodology and outcomes. Results suggest high levels of risk classification accuracy (>90%) and Area Under the Curve (AUC) in the prediction of suicidal behaviors. We report key findings and central limitations in the use of AI/ML frameworks to guide additional research, which hold the potential to impact suicide on broad scale.
Collapse
Affiliation(s)
- Rebecca A. Bernert
- Stanford Suicide Prevention Research Laboratory, Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Stanford, CA 94304, USA
| | - Amanda M. Hilberg
- Stanford Suicide Prevention Research Laboratory, Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Stanford, CA 94304, USA
| | - Ruth Melia
- Stanford Suicide Prevention Research Laboratory, Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Stanford, CA 94304, USA
- Department of Psychology, National University of Ireland, Galway, Ireland
| | - Jane Paik Kim
- Stanford Suicide Prevention Research Laboratory, Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Stanford, CA 94304, USA
| | - Nigam H. Shah
- Department of Medicine, Center for Biomedical Informatics Research, Stanford University School of Medicine, Stanford, CA 94304, USA
- Informatics, Stanford Center for Clinical and Translational Research, and Education (Spectrum), Stanford University, Stanford CA 94304, USA
| | - Freddy Abnousi
- Facebook, Menlo Park, CA 94025, USA
- Yale University School of Medicine, New Haven, CT 06510, USA
| |
Collapse
|
|
7
|
Lack of association of SNPs from the FADS1-FADS2 gene cluster with major depression or suicidal behavior. Psychiatr Genet 2016; 26:81-6. [PMID: 26513616 DOI: 10.1097/ypg.0000000000000111] [Citation(s) in RCA: 4] [Impact Index Per Article: 0.4] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/22/2022]
Abstract
Fatty acid desaturase genes (FADS1-FADS2) encode desaturases participating in the biosynthesis of long-chain polyunsaturated fatty acids. As long-chain polyunsaturated fatty acids are implicated in major depressive disorder (MDD) and suicide risk, and as both are partly heritable, we studied the association of FADS1-FADS2 polymorphisms with MDD (635 cases, 480 controls) and suicide attempt status (291 attempters, 344 MDD nonattempters). Eighteen FADS-related single-nucleotide polymorphisms were genotyped from Caucasians enrolled in Madrid (n=791) or New York City (n=324) and entered as predictors into logistic regression analyses with diagnostic group or suicide attempt history as outcomes and location and sex as covariates. No associations were observed between any single-nucleotide polymorphisms and diagnosis or attempt status. As statistical power was adequate, we conclude that FADS1-FADS2 genetic variants may not be a common determinant of MDD.
Collapse
|
|
8
|
Quintela I, Fernandez-Prieto M, Gomez-Guerrero L, Resches M, Eiris J, Barros F, Carracedo A. A 6q14.1-q15 microdeletion in a male patient with severe autistic disorder, lack of oral language, and dysmorphic features with concomitant presence of a maternally inherited Xp22.31 copy number gain. Clin Case Rep 2015; 3:415-23. [PMID: 26185640 PMCID: PMC4498854 DOI: 10.1002/ccr3.255] [Citation(s) in RCA: 4] [Impact Index Per Article: 0.4] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/15/2015] [Accepted: 02/22/2015] [Indexed: 12/14/2022] Open
Abstract
We report on a male patient with severe autistic disorder, lack of oral language, and dysmorphic features who carries a rare interstitial microdeletion of 4.96 Mb at chromosome 6q14.1-q15. The patient also harbors a maternally inherited copy number gain of 1.69 Mb at chromosome Xp22.31, whose pathogenicity is under debate.
Collapse
Affiliation(s)
- Ines Quintela
- Grupo de Medicina Xenomica, Centro Nacional de Genotipado - Plataforma de Recursos Biomoleculares y Bioinformaticos - Instituto de Salud Carlos III (CeGen-PRB2-ISCIII), Universidade de Santiago de Compostela Santiago de Compostela, Spain
| | - Montse Fernandez-Prieto
- Grupo de Medicina Xenomica, CIBERER, Fundacion Publica Galega de Medicina Xenomica - SERGAS Santiago de Compostela, Spain
| | - Lorena Gomez-Guerrero
- Grupo de Medicina Xenomica, CIBERER, Fundacion Publica Galega de Medicina Xenomica - SERGAS Santiago de Compostela, Spain
| | - Mariela Resches
- Departamento de Psicologia Evolutiva y de la Educacion, Universidade de Santiago de Compostela Santiago de Compostela, Spain
| | - Jesus Eiris
- Unidad de Neurologia Pediatrica, Departamento de Pediatria, Hospital Clinico Universitario de Santiago de Compostela Santiago de Compostela, Spain
| | - Francisco Barros
- Grupo de Medicina Xenomica, CIBERER, Fundacion Publica Galega de Medicina Xenomica - SERGAS Santiago de Compostela, Spain
| | - Angel Carracedo
- Grupo de Medicina Xenomica, Centro Nacional de Genotipado - Plataforma de Recursos Biomoleculares y Bioinformaticos - Instituto de Salud Carlos III (CeGen-PRB2-ISCIII), Universidade de Santiago de Compostela Santiago de Compostela, Spain ; Grupo de Medicina Xenomica, CIBERER, Fundacion Publica Galega de Medicina Xenomica - SERGAS Santiago de Compostela, Spain ; Center of Excellence in Genomic Medicine Research, King Abdulaziz University Jeddah, Saudi Arabia
| |
Collapse
|
|
9
|
Sokolowski M, Wasserman J, Wasserman D. An overview of the neurobiology of suicidal behaviors as one meta-system. Mol Psychiatry 2015; 20:56-71. [PMID: 25178164 DOI: 10.1038/mp.2014.101] [Citation(s) in RCA: 30] [Impact Index Per Article: 3.0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 03/14/2014] [Revised: 06/19/2014] [Accepted: 07/22/2014] [Indexed: 12/12/2022]
Abstract
Suicidal behaviors (SB) may be regarded as the outmost consequence of mental illnesses, or as a distinct entity per se. Regardless, the consequences of SB are very large to both society and affected individuals. The path leading to SB is clearly a complex one involving interactions between the subject's biology and environmental influences throughout life. With the aim to generate a representative and diversified overview of the different neurobiological components hypothesized or shown implicated across the entire SB field up to date by any approach, we selected and compiled a list of 212 gene symbols from the literature. An increasing number of novel gene (products) have been introduced as candidates, with half being implicated in SB in only the last 4 years. These candidates represent different neuro systems and functions and might therefore be regarded as competing or redundant explanations. We then adopted a unifying approach by treating them all as parts of the same meta-system, using bioinformatic tools. We present a network of all components connected by physical protein-protein interactions (the SB interactome). We proceeded by exploring the differences between the highly connected core (~30% of the candidate genes) and its peripheral parts, observing more functional homogeneity at the core, with multiple signal transduction pathways and actin-interacting proteins connecting a subset of receptors in nerve cell compartments as well as development/morphology phenotypes and the stress-sensitive synaptic plasticity processes of long term potentiation/depression. We suggest that SB neurobiology might also be viewed as one meta-system and perhaps be explained as intrinsic unbalances acting within the core or as imbalances arising between core and specific peripheral components.
Collapse
Affiliation(s)
- M Sokolowski
- National Centre for Suicide Research and Prevention of Mental Ill-Health (NASP), Karolinska Institute (KI), Stockholm, Sweden
| | - J Wasserman
- National Centre for Suicide Research and Prevention of Mental Ill-Health (NASP), Karolinska Institute (KI), Stockholm, Sweden
| | - D Wasserman
- 1] National Centre for Suicide Research and Prevention of Mental Ill-Health (NASP), Karolinska Institute (KI), Stockholm, Sweden [2] WHO Collaborating Centre for Research, Methods Development and Training in Suicide Prevention, Stockholm, Sweden
| |
Collapse
|
|
10
|
Stewart LR, Hall AL, Kang SHL, Shaw CA, Beaudet AL. High frequency of known copy number abnormalities and maternal duplication 15q11-q13 in patients with combined schizophrenia and epilepsy. BMC MEDICAL GENETICS 2011; 12:154. [PMID: 22118685 PMCID: PMC3239290 DOI: 10.1186/1471-2350-12-154] [Citation(s) in RCA: 51] [Impact Index Per Article: 3.6] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Received: 05/04/2011] [Accepted: 11/25/2011] [Indexed: 03/01/2023]
Abstract
Background Many copy number variants (CNVs) are documented to be associated with neuropsychiatric disorders, including intellectual disability, autism, epilepsy, schizophrenia, and bipolar disorder. Chromosomal deletions of 1q21.1, 3q29, 15q13.3, 22q11.2, and NRXN1 and duplications of 15q11-q13 (maternal), 16p11, and 16p13.3 have the strongest association with schizophrenia. We hypothesized that cases with both schizophrenia and epilepsy would have a higher frequency of disease-associated CNVs and would represent an enriched sample for detection of other mutations associated with schizophrenia. Methods We used array comparative genomic hybridization (CGH) to analyze 235 individuals with both schizophrenia and epilepsy, 80 with bipolar disorder and epilepsy, and 191 controls. Results We detected 10 schizophrenia plus epilepsy cases in 235 (4.3%) with the above mentioned CNVs compared to 0 in 191 controls (p = 0.003). Other likely pathological findings in schizophrenia plus epilepsy cases included 1 deletion 16p13 and 1 duplication 7q11.23 for a total of 12/235 (5.1%) while a possibly pathogenic duplication of 22q11.2 was found in one control for a total of 1 in 191 (0.5%) controls (p = 0.008). The rate of abnormality in the schizophrenia plus epilepsy of 10/235 for the more definite CNVs compares to a rate of 75/7336 for these same CNVs in a series of unselected schizophrenia cases (p = 0.0004). Conclusion We found a statistically significant increase in the frequency of CNVs known or likely to be associated with schizophrenia in individuals with both schizophrenia and epilepsy compared to controls. We found an overall 5.1% detection rate of likely pathological findings which is the highest frequency of such findings in a series of schizophrenia patients to date. This evidence suggests that the frequency of disease-associated CNVs in patients with both schizophrenia and epilepsy is significantly higher than for unselected schizophrenia.
Collapse
Affiliation(s)
- Larissa R Stewart
- Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, TX 77030 USA.
| | | | | | | | | |
Collapse
|