Predicting Therapy Outcome in a Digital Mental Health Intervention for Depression and Anxiety: A Machine Learning Approach (Preprint)

2020 ◽  
Author(s):  
Silvan Hornstein ◽  
Valerie Forman-Hoffman ◽  
Albert Nazander ◽  
Kristian Ranta ◽  
Kevin Hilbert
Keyword(s):  
Mental Health ◽  
Machine Learning ◽  
Work Productivity ◽  
Mental Healthcare ◽  
Predictive Performance ◽  
Additional Treatment ◽  
Biological Data ◽  
Depression And Anxiety ◽  
Activity Impairment ◽  
Large Patient

BACKGROUND Predicting the outcomes of individual patients for treatment interventions appears central for making mental healthcare more tailored and effective. Machine Learning (ML) has been proven to be able to make such predictions with notable accuracy. However, little work has been done to investigate the performance of such ML-based predictions within digital mental health (DMH) interventions. Implementing ML approaches in such a context would be quite easy as data is readily available for large patient populations. OBJECTIVE This study evaluates the performance of ML in predicting treatment outcomes in a DMH intervention designed for treating depression and anxiety. METHODS Several algorithms were trained based on the data of 970 patients to predict significant reduction in depression and anxiety symptoms, by using clinical and sociodemographic variables. As a Random Forest Classifier (RF) performed best over cross-validation, it was used to predict the outcomes of 279 new patients. RESULTS The RF achieved an accuracy of 0.71 for the testset (base-rate: 0.67, AUC: 0.60, P = .001, balanced accuracy: 0.60). Additionally, predicted non-responders showed less average reduction of their PHQ-9 (-2.7 , P = .004) and GAD-7 values (-3.7, P < .001) compared to responders. Besides pre-treatment PHQ and GAD values, the self-reported motivation, type of referral into the program (self versus healthcare provider) as well as Work Productivity and Activity Impairment Questionnaire (WPAI) items contributed most to the predictions. CONCLUSIONS This study highlights that, also within DMH, social-demographic and clinical variables can be used for ML to predict therapy outcomes. Despite the overall moderate performance, this appears promising as these predictions can potentially improve the outcomes of nonresponders by monitoring their progress or by offering alternative or additional treatment. Behavioural patterns measured by smartphone-based interventions, such as app-usage, as well as biological data from wearable devices in DMH interventions are highlighted as paths towards improved predictive performance.

Optimizing existing mental health screening methods in a dementia screening and risk-factor app: An Observational, machine learning study (Preprint)

2021 ◽  
Author(s):  
Narayan Kuleindiren ◽  
Raphael Paul Rifkin-Zybutz ◽  
Monika Johal ◽  
Hamza Selim ◽  
Itai Palmon ◽  
...  
Keyword(s):  
Mental Health ◽  
Machine Learning ◽  
Risk Factors ◽  
Regression Models ◽  
Superior Performance ◽  
Training Dataset ◽  
Screening Methods ◽  
Depression And Anxiety ◽  
Dementia Screening ◽  
Fold Cross Validation

BACKGROUND Mindset4Dementia is an app that aims to improve dementia screening by assessing cognition and risk factors. It considers important clinical risk factors, including prodromal symptoms, mental health disorders, and differential diagnoses of dementia. The PHQ-9 and GAD-7 are widely validated, and commonly used scales used in screening for depression and anxiety disorders respectively. Shortened versions of both (PHQ-2/GAD-2) have been produced. OBJECTIVE We sought to develop a method that maintained the brevity of these shorter questionnaires while maintaining the better precision of the original questionnaires METHODS Single questions were designed to encompass symptoms covered in the original questionnaires. Answers to these questions were combined with the PHQ-2/GAD-2 and anonymized risk factors collected by Mindset4Dementia. Machine learning models were trained to use these single questions in combination with data already collected by the app - age, response to a joke and reporting of functional impairment to predict binary and continuous outcomes as measured by the PHQ-9/GAD-7. Our model was developed with a training dataset using ten-fold cross-validation and a hold-out testing datasets and compared to results from using the shorter questionnaires (PHQ-2/GAD-2) alone to benchmark performance. RESULTS We were able to achieve superior performance in predicting PHQ-9/GAD-7 screening cut-offs than the PHQ-2 (difference In AUC 0.04, 95% CI 0.00 – 0.08, P = 0.02) but not to GAD-2 (difference in AUC 0.00, 95% CI -0.02 – 0.03, P = 0.42). In regression models we were able to accurately predict total questionnaire scores; PHQ-9 (R2 = 0.655, MAE = 2.267), GAD-7 (R2 = 0.837, MAE = 1.780). CONCLUSIONS We have developed a short screening tool for affective disorders with superior or equivalent performance to well established methods.

scholarly journals Machine Learning in Clinical Psychology and Psychotherapy Education: A Survey of Postgraduate Students at a Swiss University

2020 ◽  
Author(s):  
Charlotte Blease ◽  
Anna Kharko ◽  
Marco Annoni ◽  
Jens Gaab ◽  
Cosima Locher
Keyword(s):  
Mental Health ◽  
Machine Learning ◽  
Health Care ◽  
Mental Health Care ◽  
Clinical Psychology ◽  
Big Data Analytics ◽  
Mental Healthcare ◽  
Convenience Sample ◽  
Postgraduate Students ◽  
Psychotherapy Education

AbstractBackgroundThere is increasing use of for machine learning-enabled tools (e.g., psychotherapy apps) in mental health care.ObjectiveThis study aimed to explore postgraduate clinical psychology and psychotherapy students’ familiarity and formal exposure to topics related to artificial intelligence and machine learning (AI/ML) during their studies.MethodsIn April-June 2020, we conducted a mixed-methods web-based survey using a convenience sample of 120 clinical psychology and psychotherapy enrolled in a two-year Masters’ program students at a Swiss university.ResultsIn total 37 students responded (response rate: 37/120, 31%). Among the respondents, 73% (n=27) intended to enter a mental health profession. Among the students 97% reported that they had heard of the term ‘machine learning,’ and 78% reported that they were familiar with the concept of ‘big data analytics’. Students estimated 18.61/3600 hours, or 0.52% of their program would be spent on AI/ML education. Around half (46%) reported that they intended to learn about AI/ML as it pertained to mental health care. On 5-point Likert scale, students moderately agreed (median=4) that AI/M should be part of clinical psychology/psychotherapy education.ConclusionsEducation programs in clinical psychology/psychotherapy may lag developments in AI/ML-enabled tools in mental healthcare. This survey of postgraduate clinical psychology and psychotherapy students raises questions about how curricula could be enhanced to better prepare clinical psychology/psychotherapy trainees to engage in constructive debate about ethical and evidence-based issues pertaining to AI/ML tools, and in guiding patients on the use of online mental health services and apps.

scholarly journals Mental Health Prediction Models Using Machine Learning in Higher Education Institution

2021 ◽  
Vol 12 (5) ◽  
pp. 1782-1792
Author(s):  
Sofianita Mutalib, Et. al.
Keyword(s):  
Neural Network ◽  
Mental Health ◽  
Higher Education ◽  
Machine Learning ◽  
Support Vector Machine ◽  
Decision Tree ◽  
Mental Health Problems ◽  
Support Vector ◽  
Depression And Anxiety ◽  
Education Students

Today, mental health problem has become a grave concern in Malaysia. According to the National Health and Morbidity Survey (NHMS) 2017, one in five people in Malaysia suffers from depression, two in five from anxiety, and one in ten from stress. Higher education students are also at risk of being part of the affected community. The increased data size without proper management and analysis, and the lack of counsellors, are compounding the issue. Therefore, this paper presents on identifying factors in mental health problems among selected higher education students. This study aims to classify students into different categories of mental health problems, which are stress, depression, and anxiety, using machine learning algorithms. The data is collected from students in a higher education institute in Kuala Terengganu. The algorithms applied are Decision Tree, Neural Network, Support Vector Machine, Naïve Bayes, and logistic regression. The most accurate model for stress, depression, and anxiety is Decision Tree, Support Vector Machine, and Neural Network, respectively.

scholarly journals A web survey to assess the perceived mental healthcare among Medical Imaging Professionals during COVID-19 pandemic

2020 ◽  
Author(s):  
Rahul P Kotian ◽  
Disha Faujdar ◽  
Brayal D'souza ◽  
Sneha P Kotian ◽  
Sindhura Kunaparaju ◽  
...  
Keyword(s):  
Mental Health ◽  
Medical Imaging ◽  
Demographic Data ◽  
Mental Healthcare ◽  
Web Survey ◽  
Depression And Anxiety ◽  
Inclusion Criteria ◽  
Online Questionnaire ◽  
Cross Sectional ◽  
Using Data

Abstract Importance: Medical Imaging Professionals (MIP’s) providing imaging services exposed to coronavirus disease 2019 (COVID-19) could be psychologically stressed.Objective: To assess the magnitude of the perceived mental health outcomes among MIP’s providing imaging services to patients exposed to COVID-19. We examined the psychological stress, depression and anxiety, experienced by MIP’s in the midst of the outbreak. Background: During the first week of March,2020 the surge of coronavirus disease (COVID-19) cases reached all over the globe with more than 150,000 cases. Healthcare national and international authorities have already initiated awareness and lockdown activities. Design, Settings, and Participants: This cross-sectional, web survey-based study collected demographic data and mental health measurements from 250 MIP’s from April 29, 2020, to May 15, 2020. MIP’s working during the pandemic in hospitals for patients with COVID-19 were eligible. An online sample of MIPs was successfully recruited via the authors' networks in India using data collection tool – write google forms. A DASS21 online questionnaire was completed by the participants and then their mental health was assessed.Results: Of 400 invited MIP’s, 314 (78.5%) participated in the study; and 187 (59.5%)were included as per inclusion criteria. Hundred and three (55.08%) participants screened positive for depression, 105 (56.14%) for anxiety, and 80 (42.78%) for stress. However, 25 (13.36%), 18 (9.62%) and 16 (8.55%) screened positive for extremely severe for depression, anxiety and stress respectively.Conclusion and Relevance: In this web survey of MIP’s during COVID-19 pandemic, participants reported experiencing high rates of psychological depression, anxiety and stress, especially frontline MIP’s directly engaged in the imaging procedures for patients with COVID-19.

Comparing enrichment analysis and machine learning for identifying gene properties that discriminate between gene classes

2019 ◽  
Vol 21 (3) ◽  
pp. 803-814 ◽  
Author(s):  
Fabio Fabris ◽  
Daniel Palmer ◽  
João Pedro de Magalhães ◽  
Alex A Freitas
Keyword(s):  
Machine Learning ◽  
Enrichment Analysis ◽  
Predictive Performance ◽  
Biological Data ◽  
Statistical Hypothesis ◽  
Classification Algorithms ◽  
Classification Methods ◽  
Bioinformatics Analyses ◽  
Advantages And Disadvantages ◽  
Enrichment Methods

Abstract Biologists very often use enrichment methods based on statistical hypothesis tests to identify gene properties that are significantly over-represented in a given set of genes of interest, by comparison with a ‘background’ set of genes. These enrichment methods, although based on rigorous statistical foundations, are not always the best single option to identify patterns in biological data. In many cases, one can also use classification algorithms from the machine-learning field. Unlike enrichment methods, classification algorithms are designed to maximize measures of predictive performance and are capable of analysing combinations of gene properties, instead of one property at a time. In practice, however, the majority of studies use either enrichment or classification methods (rather than both), and there is a lack of literature discussing the pros and cons of both types of method. The goal of this paper is to compare and contrast enrichment and classification methods, offering two contributions. First, we discuss the (to some extent complementary) advantages and disadvantages of both types of methods for identifying gene properties that discriminate between gene classes. Second, we provide a set of high-level recommendations for using enrichment and classification methods. Overall, by highlighting the strengths and the weaknesses of both types of methods we argue that both should be used in bioinformatics analyses.

scholarly journals Machine Learning in Clinical Psychology and Psychotherapy Education: A Mixed Methods Pilot Survey of Postgraduate Students at a Swiss University

2021 ◽  
Vol 9 ◽  
Author(s):  
Charlotte Blease ◽  
Anna Kharko ◽  
Marco Annoni ◽  
Jens Gaab ◽  
Cosima Locher
Keyword(s):  
Mental Health ◽  
Machine Learning ◽  
Health Care ◽  
Mixed Methods ◽  
Pilot Study ◽  
Clinical Psychology ◽  
Mental Healthcare ◽  
Psychology Students ◽  
Clinical Psychology Students ◽  
Psychotherapy Education

Background: There is increasing use of psychotherapy apps in mental health care.Objective: This mixed methods pilot study aimed to explore postgraduate clinical psychology students' familiarity and formal exposure to topics related to artificial intelligence and machine learning (AI/ML) during their studies.Methods: In April-June 2020, we conducted a mixed-methods online survey using a convenience sample of 120 clinical psychology students enrolled in a two-year Masters' program at a Swiss University.Results: In total 37 students responded (response rate: 37/120, 31%). Among respondents, 73% (n = 27) intended to enter a mental health profession, and 97% reported that they had heard of the term “machine learning.” Students estimated 0.52% of their program would be spent on AI/ML education. Around half (46%) reported that they intended to learn about AI/ML as it pertained to mental health care. On 5-point Likert scale, students “moderately agreed” (median = 4) that AI/M should be part of clinical psychology/psychotherapy education. Qualitative analysis of students' comments resulted in four major themes on the impact of AI/ML on mental healthcare: (1) Changes in the quality and understanding of psychotherapy care; (2) Impact on patient-therapist interactions; (3) Impact on the psychotherapy profession; (4) Data management and ethical issues.Conclusions: This pilot study found that postgraduate clinical psychology students held a wide range of opinions but had limited formal education on how AI/ML-enabled tools might impact psychotherapy. The survey raises questions about how curricula could be enhanced to educate clinical psychology/psychotherapy trainees about the scope of AI/ML in mental healthcare.

Schoolchildren’ Depression and Anxiety Prediction Using Machine Learning Algorithms (Preprint)

2021 ◽  
Author(s):  
Radwan Qasrawi ◽  
Stephanny Vicuna Polo ◽  
Diala Abu Al-Halawah ◽  
Sameh Hallaq ◽  
Ziad Abdeen
Keyword(s):  
Mental Health ◽  
Machine Learning ◽  
Risk Factors ◽  
Cognitive Development ◽  
Random Forest ◽  
Machine Learning Techniques ◽  
Support Vector ◽  
Depression And Anxiety ◽  
Factors Associated ◽  
Learning Techniques

BACKGROUND : Depression and anxiety symptoms in early childhood have a major effect on children's mental health growth and cognitive development. Studying the effect of mental health problems on cognitive development has gained researchers' attention for the last two decades OBJECTIVE In this paper, we seek to use machine learning techniques to predict the risk factors associated with school children's depression and anxiety METHODS The study data consisted of 5685 students in grades 5-9, aged 10-17 years, studying at public and refugee schools in the West Bank. The data were collected using the health behaviors school children questionnaire in the 2012-2013 academic year and analyzed using machine learning to predict the risk factors associated with student mental health symptoms. Five machine learning techniques (Random Forest, Neural Network, Decision Tree, Support Vector Machine, and Naïve Bayes) were used for the prediction. RESULTS The results indicated that the Random Forest model had the highest accuracy levels (72.6%, 68.5%) for depression and anxiety respectively. Thus, the Random Forest had the best performance in classifying and predicting the student's depression and anxiety. The results showed that school violence and bullying, home violence, academic performance, and family income were the most important factors affecting depression and anxiety scales CONCLUSIONS Overall, machine learning proved to be an efficient tool for identifying and predicting the associated factors that influence student depression and anxiety. The deployment of machine learning within the school information systems might facilitate the development of health prevention and intervention programs that will enhance students’ mental health and cognitive development.

scholarly journals Predicting therapy outcome in a digital mental health intervention for depression and anxiety: A machine learning approach

2021 ◽  
Vol 7 ◽  
pp. 205520762110606
Author(s):  
Silvan Hornstein ◽  
Valerie Forman-Hoffman ◽  
Albert Nazander ◽  
Kristian Ranta ◽  
Kevin Hilbert
Keyword(s):  
Mental Health ◽  
Machine Learning ◽  
Anxiety Disorder ◽  
Health Intervention ◽  
Mental Health Intervention ◽  
Depression And Anxiety ◽  
Health Questionnaire ◽  
General Anxiety Disorder ◽  
General Anxiety ◽  
Patient Health

Objective Predicting the outcomes of individual participants for treatment interventions appears central to making mental healthcare more tailored and effective. However, little work has been done to investigate the performance of machine learning-based predictions within digital mental health interventions. Therefore, this study evaluates the performance of machine learning in predicting treatment response in a digital mental health intervention designed for treating depression and anxiety. Methods Several algorithms were trained based on the data of 970 participants to predict a significant reduction in depression and anxiety symptoms using clinical and sociodemographic variables. As a random forest classifier performed best over cross-validation, it was used to predict the outcomes of 279 new participants. Results The random forest achieved an accuracy of 0.71 for the test set (base rate: 0.67, area under curve (AUC): 0.60, p = 0.001, balanced accuracy: 0.60). Additionally, predicted non-responders showed less average reduction of their Patient Health Questionnaire-9 (PHQ-9) (−2.7, p = 0.004) and General Anxiety Disorder Screener-7 values (−3.7, p < 0.001) compared to responders. Besides pre-treatment Patient Health Questionnaire-9 and General Anxiety Disorder Screener-7 values, the self-reported motivation, type of referral into the programme (self vs. healthcare provider) as well as Work Productivity and Activity Impairment Questionnaire items contributed most to the predictions. Conclusions This study provides evidence that social-demographic and clinical variables can be used for machine learning to predict therapy outcomes within the context of a therapist-supported digital mental health intervention. Despite the overall moderate performance, this appears promising as these predictions can potentially improve the outcomes of non-responders by monitoring their progress or by offering alternative or additional treatment.

scholarly journals Scaling Up Maternal Mental healthcare by Increasing access to Treatment (SUMMIT) through non-specialist providers and telemedicine: a study protocol for a non-inferiority randomized controlled trial

Trials ◽  
2021 ◽  
Vol 22 (1) ◽  
Author(s):  
D. R. Singla ◽  
S. E. Meltzer-Brody ◽  
R. K. Silver ◽  
S. N. Vigod ◽  
J. J. Kim ◽  
...  
Keyword(s):  
Mental Health ◽  
Depressive Symptoms ◽  
Study Protocol ◽  
Controlled Trial ◽  
Depression Scale ◽  
Mental Healthcare ◽  
Scaling Up ◽  
Postpartum Women ◽  
Depression And Anxiety ◽  
Access To Treatment

Abstract Background Depression and anxiety impact up to 1 in 5 pregnant and postpartum women worldwide. Yet, as few as 20% of these women are treated with frontline interventions such as evidence-based psychological treatments. Major barriers to uptake are the limited number of specialized mental health treatment providers in most settings, and problems with accessing in-person care, such as childcare or transportation. Task sharing of treatment to non-specialist providers with delivery on telemedicine platforms could address such barriers. However, the equivalence of these strategies to specialist and in-person models remains unproven. Methods This study protocol outlines the Scaling Up Maternal Mental healthcare by Increasing access to Treatment (SUMMIT) randomized trial. SUMMIT is a pragmatic, non-inferiority test of the comparable effectiveness of two types of providers (specialist vs. non-specialist) and delivery modes (telemedicine vs. in-person) of a brief, behavioral activation (BA) treatment for perinatal depressive and anxiety symptoms. Specialists (psychologists, psychiatrists, and social workers with ≥ 5 years of therapy experience) and non-specialists (nurses and midwives with no formal training in mental health care) were trained in the BA protocol, with the latter supervised by a BA expert during treatment delivery. Consenting pregnant and postpartum women with Edinburgh Postnatal Depression Scale (EPDS) score of ≥ 10 (N = 1368) will be randomized to one of four arms (telemedicine specialist, telemedicine non-specialist, in-person specialist, in-person non-specialist), stratified by pregnancy status (antenatal/postnatal) and study site. The primary outcome is participant-reported depressive symptoms (EPDS) at 3 months post-randomization. Secondary outcomes are maternal symptoms of anxiety and trauma symptoms, perceived social support, activation levels and quality of life at 3-, 6-, and 12-month post-randomization, and depressive symptoms at 6- and 12-month post-randomization. Primary analyses are per-protocol and intent-to-treat. The study has successfully continued despite the COVID-19 pandemic, with needed adaptations, including temporary suspension of the in-person arms and ongoing randomization to telemedicine arms. Discussion The SUMMIT trial is expected to generate evidence on the non-inferiority of BA delivered by a non-specialist provider compared to specialist and telemedicine compared to in-person. If confirmed, results could pave the way to a dramatic increase in access to treatment for perinatal depression and anxiety. Trial registration ClinicalTrials.gov NCT 04153864. Registered on November 6, 2019.

Sign in / Sign up

Export Citation Format

Share Document