Research

10 pages, 260 KiB

Open AccessArticle

A Machine Learning Approach for Predicting Wage Workers’ Suicidal Ideation

by Hwanjin Park and Kounseok Lee

J. Pers. Med. 2022, 12(6), 945; https://doi.org/10.3390/jpm12060945 - 09 Jun 2022

Viewed by 1267

(1) Background: Workers spend most of their days working. One’s working environment can be a risk factor for suicide. In this study, we examined whether suicidal ideation can be predicted using individual characteristics, emotional states, and working environments. (2) Methods: Nine years of [...] Read more.

(1) Background: Workers spend most of their days working. One’s working environment can be a risk factor for suicide. In this study, we examined whether suicidal ideation can be predicted using individual characteristics, emotional states, and working environments. (2) Methods: Nine years of data from the Korean National Health and Nutrition Survey were used. A total of 12,816 data points were analyzed, and 23 variables were selected. The random forest technique was used to predict suicidal thoughts. (3) Results: When suicidal ideation cases were predicted using all of the independent variables, 98.9% of cases were predicted, and 97.4% could be predicted using only work-related conditions. (4) Conclusions: It was confirmed that suicide risk could be predicted efficiently when machine learning techniques were applied using variables such as working environments. Full article

(This article belongs to the Special Issue Advances in the Use of Machine Learning for the Clinical Research of Mood Disorders)

12 pages, 518 KiB

Open AccessArticle

The Effectiveness of Predicting Suicidal Ideation through Depressive Symptoms and Social Isolation Using Machine Learning Techniques

by Sunhae Kim and Kounseok Lee

J. Pers. Med. 2022, 12(4), 516; https://doi.org/10.3390/jpm12040516 - 22 Mar 2022

Cited by 5 | Viewed by 2435

Abstract

(1) Background: Social isolation is a major risk factor for suicidal ideation. In this study, we investigated whether the evaluation of both depression and social isolation in combination could effectively predict suicidal ideation; (2) Methods: A total of 7994 data collected from community [...] Read more.

(1) Background: Social isolation is a major risk factor for suicidal ideation. In this study, we investigated whether the evaluation of both depression and social isolation in combination could effectively predict suicidal ideation; (2) Methods: A total of 7994 data collected from community residents were analyzed. Statistical analysis was performed using age, the Patient Health Questionnaire-9, and the Lubben Social Network Scale as predictors as the dependent variables for suicidal ideation; machine learning (ML) methods K-Nearest Neighbors, Random Forest, and Neural Network Classification were used; (3) Results: The prediction of suicidal ideation using depression and social isolation showed high area under the curve (0.643–0.836) and specificity (0.959–0.987) in all ML techniques. In the predictor model (model 2) that additionally evaluated social isolation, the validation accuracy consistently increased compared to the depression-only model (model 1); (4) Conclusions: It is confirmed that the machine learning technique using depression and social isolation can be an effective method when predicting suicidal ideation. Full article

(This article belongs to the Special Issue Advances in the Use of Machine Learning for the Clinical Research of Mood Disorders)

► Show Figures

Graphical abstract

25 pages, 3004 KiB

Open AccessArticle

Pathway Phenotypes Underpinning Depression, Anxiety, and Chronic Fatigue Symptoms Due to Acute Rheumatoid Arthritis: A Precision Nomothetic Psychiatry Analysis

by Hasan Najah Smesam, Hasan Abbas Qazmooz, Sinan Qayes Khayoon, Abbas F. Almulla, Hussein Kadhem Al-Hakeim and Michael Maes

J. Pers. Med. 2022, 12(3), 476; https://doi.org/10.3390/jpm12030476 - 16 Mar 2022

Cited by 7 | Viewed by 2638

Abstract

Rheumatoid arthritis (RA) is a chronic inflammatory and autoimmune disorder which affects the joints in the wrists, fingers, and knees. RA is often associated with depressive and anxiety symptoms as well as chronic fatigue syndrome (CFS)-like symptoms. This paper examines the association between [...] Read more.

Rheumatoid arthritis (RA) is a chronic inflammatory and autoimmune disorder which affects the joints in the wrists, fingers, and knees. RA is often associated with depressive and anxiety symptoms as well as chronic fatigue syndrome (CFS)-like symptoms. This paper examines the association between depressive symptoms (measured with the Beck Depression Inventory, BDI), anxiety (Hamilton Anxiety Rating Scale, HAMA), CFS-like (Fibro-fatigue Scale) symptoms and immune–inflammatory, autoimmune, and endogenous opioid system (EOS) markers, and lactosylcer-amide (CD17) in RA. The serum biomarkers were assayed in 118 RA and 50 healthy controls. Results were analyzed using the new precision nomothetic psychiatry approach. We found significant correlations between the BDI, FF, and HAMA scores and severity of RA, as assessed with the DAS28-4, clinical and disease activity indices, the number of tender and swollen joints, and patient and evaluator global assessment scores. Partial least squares analysis showed that 69.7% of the variance in this common core underpinning psychopathology and RA symptoms was explained by immune–inflammatory pathways, rheumatoid factor, anti-citrullinated protein antibodies, CD17, and mu-opioid receptor levels. We constructed a new endophenotype class comprising patients with very high immune–inflammatory markers, CD17, RA, affective and CF-like symptoms, and tobacco use disorder. We extracted a reliable and replicable latent vector (pathway phenotype) from immune data, psychopathology, and RA-severity scales. Depression, anxiety, and CFS-like symptoms due to RA are manifestations of the phenome of RA and are mediated by the effects of the same immune–inflammatory, autoimmune, and other pathways that underpin the pathophysiology of RA. Full article

(This article belongs to the Special Issue Advances in the Use of Machine Learning for the Clinical Research of Mood Disorders)

► Show Figures

Figure 1

17 pages, 3016 KiB

Open AccessEditor’s ChoiceArticle

Areas of Interest and Social Consideration of Antidepressants on English Tweets: A Natural Language Processing Classification Study

by Laura de Anta, Miguel Angel Alvarez-Mon, Miguel A. Ortega, Cristina Salazar, Carolina Donat-Vargas, Javier Santoma-Vilaclara, Maria Martin-Martinez, Guillermo Lahera, Luis Gutierrez-Rojas, Roberto Rodriguez-Jimenez, Javier Quintero and Melchor Alvarez-Mon

J. Pers. Med. 2022, 12(2), 155; https://doi.org/10.3390/jpm12020155 - 25 Jan 2022

Cited by 8 | Viewed by 2951

Abstract

Background: Antidepressants are the foundation of the treatment of major depressive disorders. Despite the scientific evidence, there is still a sustained debate and concern about the efficacy of antidepressants, with widely differing opinions among the population about their positive and negative effects, which [...] Read more.

Background: Antidepressants are the foundation of the treatment of major depressive disorders. Despite the scientific evidence, there is still a sustained debate and concern about the efficacy of antidepressants, with widely differing opinions among the population about their positive and negative effects, which may condition people’s attitudes towards such treatments. Our aim is to investigate Twitter posts about antidepressants in order to have a better understanding of the social consideration of antidepressants. Methods: We gathered public tweets mentioning antidepressants written in English, published throughout a 22-month period, between 1 January 2019 and 31 October 2020. We analysed the content of each tweet, determining in the first place whether they included medical aspects or not. Those with medical content were classified into four categories: general aspects, such as quality of life or mood, sleep-related conditions, appetite/weight issues and aspects around somatic alterations. In non-medical tweets, we distinguished three categories: commercial nature (including all economic activity, drug promotion, education or outreach), help request/offer, and drug trivialization. In addition, users were arranged into three categories according to their nature: patients and relatives, caregivers, and interactions between Twitter users. Finally, we identified the most mentioned antidepressants, including the number of retweets and likes, which allowed us to measure the impact among Twitter users. Results: The activity in Twitter concerning antidepressants is mainly focused on the effects these drugs may have on certain health-related areas, specifically sleep (20.87%) and appetite/weight (8.95%). Patients and relatives are the type of user that most frequently posts tweets with medical content (65.2%, specifically 80% when referencing sleep and 78.6% in the case of appetite/weight), whereas they are responsible for only 2.9% of tweets with non-medical content. Among tweets classified as non-medical in this study, the most common subject was drug trivialization (66.86%). Caregivers barely have any presence in conversations in Twitter about antidepressants (3.5%). However, their tweets rose more interest among other users, with a ratio 11.93 times higher than those posted by patients and their friends and family. Mirtazapine is the most mentioned antidepressant in Twitter (45.43%), with a significant difference with the rest, agomelatine (11.11%). Conclusions: This study shows that Twitter users that take antidepressants, or their friends and family, use social media to share medical information about antidepressants. However, other users that do not talk about antidepressants from a personal or close experience, frequently do so in a stigmatizing manner, by trivializing them. Our study also brings to light the scarce presence of caregivers in Twitter. Full article

(This article belongs to the Special Issue Advances in the Use of Machine Learning for the Clinical Research of Mood Disorders)

► Show Figures

Figure 1

13 pages, 4162 KiB

Open AccessArticle

Aberrant Structure MRI in Parkinson’s Disease and Comorbidity with Depression Based on Multinomial Tensor Regression Analysis

by Xuan Cao, Fang Yang, Jingyi Zheng, Xiao Wang and Qingling Huang

J. Pers. Med. 2022, 12(1), 89; https://doi.org/10.3390/jpm12010089 - 11 Jan 2022

Cited by 12 | Viewed by 1868

Abstract

Background: Depression is a prominent and highly prevalent nonmotor feature in patients with Parkinson’s disease (PD). The neural and pathophysiologic mechanisms of PD with depression (DPD) remain unclear. The current diagnosis of DPD largely depends on clinical evaluation. Methods: We proposed a new [...] Read more.

Background: Depression is a prominent and highly prevalent nonmotor feature in patients with Parkinson’s disease (PD). The neural and pathophysiologic mechanisms of PD with depression (DPD) remain unclear. The current diagnosis of DPD largely depends on clinical evaluation. Methods: We proposed a new family of multinomial tensor regressions that leveraged whole-brain structural magnetic resonance imaging (MRI) data to discriminate among 196 non-depressed PD (NDPD) patients, 84 DPD patients, 200 healthy controls (HC), and to assess the special brain microstructures in NDPD and DPD. The method of maximum likelihood estimation coupled with state-of-art gradient descent algorithms was used to predict the individual diagnosis of PD and the development of DPD in PD patients. Results: The results reveal that the proposed efficient approach not only achieved a high prediction accuracy (0.94) with a multi-class AUC (0.98) for distinguishing between NDPD, DPD, and HC on the testing set but also located the most discriminative regions for NDPD and DPD, including cortical regions, the cerebellum, the brainstem, the bilateral basal ganglia, and the thalamus and limbic regions. Conclusions: The proposed imaging technique based on tensor regression performs well without any prior feature information, facilitates a deeper understanding into the abnormalities in DPD and PD, and plays an essential role in the statistical analysis of high-dimensional complex MRI imaging data to support the radiological diagnosis of comorbidity of depression with PD. Full article

(This article belongs to the Special Issue Advances in the Use of Machine Learning for the Clinical Research of Mood Disorders)

► Show Figures

Figure 1

23 pages, 1322 KiB

Open AccessArticle

Intersections between Copper, β-Arrestin-1, Calcium, FBXW7, CD17, Insulin Resistance and Atherogenicity Mediate Depression and Anxiety Due to Type 2 Diabetes Mellitus: A Nomothetic Network Approach

by Hussein Kadhem Al-Hakeim, Hadi Hasan Hadi, Ghoufran Akeel Jawad and Michael Maes

J. Pers. Med. 2022, 12(1), 23; https://doi.org/10.3390/jpm12010023 - 01 Jan 2022

Cited by 9 | Viewed by 1962

Abstract

Type 2 diabetes mellitus (T2DM) is frequently accompanied by affective disorders with a prevalence of comorbid depression of around 25%. Nevertheless, the biomarkers of affective symptoms including depression and anxiety due to T2DM are not well established. The present study delineated the effects [...] Read more.

Type 2 diabetes mellitus (T2DM) is frequently accompanied by affective disorders with a prevalence of comorbid depression of around 25%. Nevertheless, the biomarkers of affective symptoms including depression and anxiety due to T2DM are not well established. The present study delineated the effects of serum levels of copper, zinc, β-arrestin-1, FBXW7, lactosylceramide (LacCer), serotonin, calcium, magnesium on severity of depression and anxiety in 58 men with T2DM and 30 healthy male controls beyond the effects of insulin resistance (IR) and atherogenicity. Severity of affective symptoms was assessed using the Hamilton Depression and Anxiety rating scales. We found that 61.7% of the variance in affective symptoms was explained by the multivariate regression on copper, β-arrestin-1, calcium, and IR coupled with atherogenicity. Copper and LacCer (positive) and calcium and BXW7 (inverse) had significant specific indirect effects on affective symptoms, which were mediated by IR and atherogenicity. Copper, β-arrestin-1, and calcium were associated with affective symptoms above and beyond the effects of IR and atherogenicity. T2DM and affective symptoms share common pathways, namely increased atherogenicity, IR, copper, and β-arrestin-1, and lowered calcium, whereas copper, β-arrestin-1, calcium, LacCer, and FBXW7 may modulate depression and anxiety symptoms by affecting T2DM. Full article

(This article belongs to the Special Issue Advances in the Use of Machine Learning for the Clinical Research of Mood Disorders)

► Show Figures

Figure 1

11 pages, 272 KiB

Open AccessArticle

Screening of Mood Symptoms Using MMPI-2-RF Scales: An Application of Machine Learning Techniques

by Sunhae Kim, Hye-Kyung Lee and Kounseok Lee

J. Pers. Med. 2021, 11(8), 812; https://doi.org/10.3390/jpm11080812 - 20 Aug 2021

Cited by 2 | Viewed by 2791

Abstract

(1) Background: The MMPI-2-RF is the most widely used and most researched test among the tools for assessing psychopathology, and previous studies have established its validity. Mood disorders are the most common mental disorders worldwide; they present difficulties in early detection, go undiagnosed [...] Read more.

(1) Background: The MMPI-2-RF is the most widely used and most researched test among the tools for assessing psychopathology, and previous studies have established its validity. Mood disorders are the most common mental disorders worldwide; they present difficulties in early detection, go undiagnosed in many cases, and have a poor prognosis. (2) Methods: We analyzed a total of 8645 participants. We used the PHQ-9 to evaluate depressive symptoms and the MDQ to evaluate hypomanic symptoms. We used the 10 MMPI-2 Restructured Form scales and 23 Specific Problems scales for the MMPI-2-RF as predictors. We performed machine learning analysis using the k-nearest neighbor classification, linear discriminant analysis, and random forest classification. (3) Results: Through the machine learning technique, depressive symptoms were predicted with an AUC of 0.634–0.767, and the corresponding value range for hypomanic symptoms was 0.770–0.840. When using RCd to predict depressive symptoms, the AUC was 0.807, but this value was 0.840 when using linear discriminant classification. When predicting hypomanic symptoms with RC9, the AUC was 0.704, but this value was 0.767 when using the linear discriminant method. (4) Conclusions: Using machine learning analysis, we defined that participants’ mood symptoms could be classified and predicted better than when using the Restructured Clinical scales. Full article

(This article belongs to the Special Issue Advances in the Use of Machine Learning for the Clinical Research of Mood Disorders)

15 pages, 977 KiB

Open AccessArticle

Automatic Diagnosis of Bipolar Disorder Using Optical Coherence Tomography Data and Artificial Intelligence

by Eva M. Sánchez-Morla, Juan L. Fuentes, Juan M. Miguel-Jiménez, Luciano Boquete, Miguel Ortiz, Elvira Orduna, María Satue and Elena Garcia-Martin

J. Pers. Med. 2021, 11(8), 803; https://doi.org/10.3390/jpm11080803 - 18 Aug 2021

Cited by 9 | Viewed by 3538

Abstract

Background: The aim of this study is to explore an objective approach that aids the diagnosis of bipolar disorder (BD), based on optical coherence tomography (OCT) data which are analyzed using artificial intelligence. Methods: Structural analyses of nine layers of the retina were [...] Read more.

Background: The aim of this study is to explore an objective approach that aids the diagnosis of bipolar disorder (BD), based on optical coherence tomography (OCT) data which are analyzed using artificial intelligence. Methods: Structural analyses of nine layers of the retina were analyzed in 17 type I BD patients and 42 controls, according to the areas defined by the Early Treatment Diabetic Retinopathy Study (ETDRS) chart. The most discriminating variables made up the feature vector of several automatic classifiers: Gaussian Naive Bayes, K-nearest neighbors and support vector machines. Results: BD patients presented retinal thinning affecting most layers, compared to controls. The retinal thickness of the parafoveolar area showed a high capacity to discriminate BD subjects from healthy individuals, specifically for the ganglion cell (area under the curve (AUC) = 0.82) and internal plexiform (AUC = 0.83) layers. The best classifier showed an accuracy of 0.95 for classifying BD versus controls, using as variables of the feature vector the IPL (inner nasal region) and the INL (outer nasal and inner inferior regions) thickness. Conclusions: Our patients with BD present structural alterations in the retina, and artificial intelligence seem to be a useful tool in BD diagnosis, but larger studies are needed to confirm our findings. Full article

(This article belongs to the Special Issue Advances in the Use of Machine Learning for the Clinical Research of Mood Disorders)

► Show Figures

Figure 1

13 pages, 682 KiB

Open AccessArticle

Prediction of Probable Major Depressive Disorder in the Taiwan Biobank: An Integrated Machine Learning and Genome-Wide Analysis Approach

by Eugene Lin, Po-Hsiu Kuo, Wan-Yu Lin, Yu-Li Liu, Albert C. Yang and Shih-Jen Tsai

J. Pers. Med. 2021, 11(7), 597; https://doi.org/10.3390/jpm11070597 - 24 Jun 2021

Cited by 5 | Viewed by 3144

Abstract

In light of recent advancements in machine learning, personalized medicine using predictive algorithms serves as an essential paradigmatic methodology. Our goal was to explore an integrated machine learning and genome-wide analysis approach which targets the prediction of probable major depressive disorder (MDD) using [...] Read more.

In light of recent advancements in machine learning, personalized medicine using predictive algorithms serves as an essential paradigmatic methodology. Our goal was to explore an integrated machine learning and genome-wide analysis approach which targets the prediction of probable major depressive disorder (MDD) using 9828 individuals in the Taiwan Biobank. In our analysis, we reported a genome-wide significant association with probable MDD that has not been previously identified: FBN1 on chromosome 15. Furthermore, we pinpointed 17 single nucleotide polymorphisms (SNPs) which show evidence of both associations with probable MDD and potential roles as expression quantitative trait loci (eQTLs). To predict the status of probable MDD, we established prediction models with random undersampling and synthetic minority oversampling using 17 eQTL SNPs and eight clinical variables. We utilized five state-of-the-art models: logistic ridge regression, support vector machine, C4.5 decision tree, LogitBoost, and random forests. Our data revealed that random forests had the highest performance (area under curve = 0.8905 ± 0.0088; repeated 10-fold cross-validation) among the predictive algorithms to infer complex correlations between biomarkers and probable MDD. Our study suggests that an integrated machine learning and genome-wide analysis approach may offer an advantageous method to establish bioinformatics tools for discriminating MDD patients from healthy controls. Full article

(This article belongs to the Special Issue Advances in the Use of Machine Learning for the Clinical Research of Mood Disorders)

► Show Figures

Figure 1

Review

Jump to: Research, Other

24 pages, 351 KiB

Open AccessReview

Personalized Diagnosis and Treatment for Neuroimaging in Depressive Disorders

by Jongha Lee, Suhyuk Chi and Moon-Soo Lee

J. Pers. Med. 2022, 12(9), 1403; https://doi.org/10.3390/jpm12091403 - 29 Aug 2022

Cited by 2 | Viewed by 2249

Abstract

Depressive disorders are highly heterogeneous in nature. Previous studies have not been useful for the clinical diagnosis and prediction of outcomes of major depressive disorder (MDD) at the individual level, although they provide many meaningful insights. To make inferences beyond group-level analyses, machine [...] Read more.

Depressive disorders are highly heterogeneous in nature. Previous studies have not been useful for the clinical diagnosis and prediction of outcomes of major depressive disorder (MDD) at the individual level, although they provide many meaningful insights. To make inferences beyond group-level analyses, machine learning (ML) techniques can be used for the diagnosis of subtypes of MDD and the prediction of treatment responses. We searched PubMed for relevant studies published until December 2021 that included depressive disorders and applied ML algorithms in neuroimaging fields for depressive disorders. We divided these studies into two sections, namely diagnosis and treatment outcomes, for the application of prediction using ML. Structural and functional magnetic resonance imaging studies using ML algorithms were included. Thirty studies were summarized for the prediction of an MDD diagnosis. In addition, 19 studies on the prediction of treatment outcomes for MDD were reviewed. We summarized and discussed the results of previous studies. For future research results to be useful in clinical practice, ML enabling individual inferences is important. At the same time, there are important challenges to be addressed in the future. Full article

(This article belongs to the Special Issue Advances in the Use of Machine Learning for the Clinical Research of Mood Disorders)

Other

Jump to: Research, Review

22 pages, 3042 KiB

Open AccessOpinion

Precision Nomothetic Medicine in Depression Research: A New Depression Model, and New Endophenotype Classes and Pathway Phenotypes, and A Digital Self

by Michael Maes

J. Pers. Med. 2022, 12(3), 403; https://doi.org/10.3390/jpm12030403 - 05 Mar 2022

Cited by 29 | Viewed by 3892

Abstract

Machine learning approaches, such as soft independent modeling of class analogy (SIMCA) and pathway analysis, were introduced in depression research in the 1990s (Maes et al.) to construct neuroimmune endophenotype classes. The goal of this paper is to examine the promise of precision [...] Read more.

Machine learning approaches, such as soft independent modeling of class analogy (SIMCA) and pathway analysis, were introduced in depression research in the 1990s (Maes et al.) to construct neuroimmune endophenotype classes. The goal of this paper is to examine the promise of precision psychiatry to use information about a depressed person’s own pan-omics, environmental, and lifestyle data, or to tailor preventative measures and medical treatments to endophenotype subgroups of depressed patients in order to achieve the best clinical outcome for each individual. Three steps are emerging in precision medicine: (1) the optimization and refining of classical models and constructing digital twins; (2) the use of precision medicine to construct endophenotype classes and pathway phenotypes, and (3) constructing a digital self of each patient. The root cause of why precision psychiatry cannot develop into true sciences is that there is no correct (cross-validated and reliable) model of clinical depression as a serious medical disorder discriminating it from a normal emotional distress response including sadness, grief and demoralization. Here, we explain how we used (un)supervised machine learning such as partial least squares path analysis, SIMCA and factor analysis to construct (a) a new precision depression model; (b) a new endophenotype class, namely major dysmood disorder (MDMD), which is a nosological class defined by severe symptoms and neuro-oxidative toxicity; and a new pathway phenotype, namely the reoccurrence of illness (ROI) index, which is a latent vector extracted from staging characteristics (number of depression and manic episodes and suicide attempts), and (c) an ideocratic profile with personalized scores based on all MDMD features. Full article

(This article belongs to the Special Issue Advances in the Use of Machine Learning for the Clinical Research of Mood Disorders)

► Show Figures

Figure 1

Journal Menu

Journal Browser

Advances in the Use of Machine Learning for the Clinical Research of Mood Disorders

Share This Special Issue

Special Issue Editor

Special Issue Information

Keywords

Published Papers (11 papers)

Research

Review

Other

Further Information

Guidelines

MDPI Initiatives

Follow MDPI