Machine learning-based discrimination of panic disorder from other anxiety disorders

A propensity score-adjusted HRV analysis of social avoidance and distress in patients with panic disorders

Panic disorder (PD) may accompany elevated social avoidance and distress (SAD). Higher SAD in patients with PD predicts a poorer prognosis and response to treatment. As heart rate variability (HRV) reflects the self-regulatory capacity underlying SAD, applying HRV to distinguish PD heterogeneity would be clinically beneficial in formulating personalized treatment strategies. We hypothesized that HRV would be lower in patients with PD and severe SAD (PD-SAD group) than in those without SAD (PD group). A total of 288 patients met the eligibility criteria, with complete collection of variables of interest and HRV. The inverse probability of treatment weighting (IPTW) method was used to balance the groups for the baseline characteristics of patients. Following IPTW adjustment, the average treatment effects were computed using inverse propensity weighting with regression adjustment (IPW-RA). All characteristics were similar in both groups after IPTW adjustment. The PD-SAD group showed a decreased tendency in time domain parameters, including mean heart rate, SDNN, RMSSD, and pNN50 with a nonlinear domain of SD1. However, no significant intergroup differences were observed in the frequency domain. Higher SAD in PD was associated with reduced HRV, mainly in the time domain, which may be attributed to the shared neural networks between dysfunctional self-regulation statuses, as indexed by reduced HRV. As respiratory sinus arrhythmia is mainly reflected in the frequency domain, the time domain may be more reliable for identifying heterogeneity within patients with PD, who are frequently associated with respiratory pattern abnormalities.

Diagnosis of mental disorders using machine learning: Literature review and bibliometric mapping from 2012 to 2023

Background

Mental disorders (MDs) are becoming a leading burden in non-communicable diseases (NCDs). As per the World Health Organization's 2022 assessment report, there was a steep increase of 25 % in MDs during the COVID-19 pandemic. Early diagnosis of MDs can significantly improve treatment outcome and save disability-adjusted life years (DALYs). In recent times, the application of machine learning (ML) and deep learning (DL)) has shown promising results in the diagnosis of MDs, and the field has witnessed a huge research output in the form of research publications. Therefore, a bibliometric mapping along with a review of recent advancements is required.

Methods

This study presents a bibliometric analysis and review of the research, published over the last 10 years. Literature searches were conducted in the Scopus database for the period from January 1, 2012, to June 9, 2023. The data was filtered and screened to include only relevant and reliable publications. A total of 2811 journal articles were found. The data was exported to a comma-separated value (CSV) format for further analysis. Furthermore, a review of 40 selected studies was performed.

Results

The popularity of ML techniques in diagnosing MDs has been growing, with an annual research growth rate of 17.05 %. The Journal of Affective Disorders published the most documents (n = 97), while Wang Y. (n = 64) has published the most articles. Lotka's law is observed, with a minority of authors contributing the majority of publications. The top affiliating institutes are the West China Hospital of Sichuan University followed by the University of California, with China and the US dominating the top 10 institutes. While China has more publications, papers affiliated with the US receive more citations. Depression and schizophrenia are the primary focuses of ML and deep learning (DL) in mental disease detection. Co-occurrence network analysis reveals that ML is associated with depression, schizophrenia, autism, anxiety, ADHD, obsessive-compulsive disorder, and PTSD. Popular algorithms include support vector machine (SVM) classifier, decision tree classifier, and random forest classifier. Furthermore, DL is linked to neuroimaging techniques such as MRI, fMRI, and EEG, as well as bipolar disorder. Current research trends encompass DL, LSTM, generalized anxiety disorder, feature fusion, and convolutional neural networks.

Assessing validity of the Klenico diagnostic software system in a large psychotherapeutic inpatient sample

Introduction

Mental disorders are often underdiagnosed in routine diagnostic procedures due to the use of unstandardized assessments; this can result in people either not receiving necessary treatment or receiving ineffective treatment for their condition. Klenico is an online diagnostic software system that facilitates diagnosis of mental disorders in adults through the use of standardized procedures. The procedure encompasses two modules, self-report and clinical validation. The current study aimed to confirm the validity of the Klenico assessment in a large clinical sample.

Methods

Fully anonymized data from 495 adult inpatients were used. ICD-10 diagnoses were made during an initial interview by the clinical staff. Afterwards, patients filled out self-report questionnaires (BDI-II, BSI, EDE-Q, OCI-R, PHQ-D, and Y-BOCS) and completed the Klenico self-report module, which involves selecting and rating the severity of applicable symptoms. Finally, in the clinical validation module, mental health professionals validated the symptoms endorsed in the self-report module. Six Klenico domains were tested against patient self-reports and routine ICD-10 diagnoses by following the multitrait-multimethod approach. Internal consistency was assessed by calculating Cronbach's alpha.

Results

The Klenico depressive disorders, OCD, and somatoform disorders domains revealed high correlations with the congruent questionnaires (i.e., those pertaining to these specific disorders) and revealed low correlations with the noncongruent questionnaires (i.e., those pertaining to other disorders), therefore evidencing construct validity. For the eating disorders and psychotic disorders domains, divergent validity was demonstrated. For the anxiety disorders domain, although analysis mostly indicated construct validity, this should be further confirmed.

Discussion

Overall, the results largely confirmed the construct validity of the Klenico assessment, demonstrating its use as an easy-to-use, valid, standardized, and comprehensive instrument for diagnosing mental disorders.

Applying Self-Supervised Representation Learning for Emotion Recognition Using Physiological Signals

The use of machine learning (ML) techniques in affective computing applications focuses on improving the user experience in emotion recognition. The collection of input data (e.g., physiological signals), together with expert annotations are part of the established standard supervised learning methodology used to train human emotion recognition models. However, these models generally require large amounts of labeled data, which is expensive and impractical in the healthcare context, in which data annotation requires even more expert knowledge. To address this problem, this paper explores the use of the self-supervised learning (SSL) paradigm in the development of emotion recognition methods. This approach makes it possible to learn representations directly from unlabeled signals and subsequently use them to classify affective states. This paper presents the key concepts of emotions and how SSL methods can be applied to recognize affective states. We experimentally analyze and compare self-supervised and fully supervised training of a convolutional neural network designed to recognize emotions. The experimental results using three emotion datasets demonstrate that self-supervised representations can learn widely useful features that improve data efficiency, are widely transferable, are competitive when compared to their fully supervised counterparts, and do not require the data to be labeled for learning.

Identification of Cardiac Patients Based on the Medical Conditions Using Machine Learning Models

Chronic diseases are the most severe health concern today, and heart disease is one of them. Coronary artery disease (CAD) affects blood flow to the heart, and it is the most common type of heart disease which causes a heart attack. High blood pressure, high cholesterol, and smoking significantly increase the risk of heart disease. To estimate the risk of heart disease is a complex process because it depends on various input parameters. The linear and analytical models failed due to their assumptions and limited dataset. The existing studies have used medical data for classification purposes, which help to identify the exact condition of the patient, but no one has developed any correlation equation which can be directly used to identify the patients. In this paper, mathematical models have been developed using the medical database of patients suffering from heart disease. Curve fitting and artificial neural network (ANN) have been applied to model the condition of patients to find out whether the patient is suffering from heart disease or not. The developed curve fitting model can identify the cardiac patient with accuracy, having a coefficient of determination (R²-value) of 0.6337 and mean absolute error (MAE) of 0.293 at a root mean square error (RMSE) of 0.3688, and the ANN-based model can identify the cardiac patient with accuracy having a coefficient of determination (R²-value) of 0.8491 and MAE of 0.20 at RMSE of 0.267, it has been found that ANN provides superior mathematical modeling than curve fitting method in identifying the heart disease patients. Medical professionals can utilize this model to identify heart patients without any angiography or computed tomography angiography test.

Application of photoplethysmography signals for healthcare systems: An in-depth review

BACKGROUND AND OBJECTIVES

Photoplethysmography (PPG) is a device that measures the amount of light absorbed by the blood vessel, blood, and tissues, which can, in turn, translate into various measurements such as the variation in blood flow volume, heart rate variability, blood pressure, etc. Hence, PPG signals can produce a wide variety of biological information that can be useful for the detection and diagnosis of various health problems. In this review, we are interested in the possible health disorders that can be detected using PPG signals.

METHODS

We applied PRISMA guidelines to systematically search various journal databases and identified 43 PPG studies that fit the criteria of this review.

RESULTS

Twenty-five health issues were identified from these studies that were classified into six categories: cardiac, blood pressure, sleep health, mental health, diabetes, and miscellaneous. Various routes were employed in these PPG studies to perform the diagnosis: machine learning, deep learning, and statistical routes. The studies were reviewed and summarized.

CONCLUSIONS

We identified limitations such as poor standardization of sampling frequencies and lack of publicly available PPG databases. We urge that future work should consider creating more publicly available databases so that a wide spectrum of health problems can be covered. We also want to promote the use of PPG signals as a potential precision medicine tool in both ambulatory and hospital settings.

Online prediction model for primary aldosteronism in patients with hypertension in Chinese population: A two-center retrospective study

BACKGROUND

The prevalence of primary aldosteronism (PA) varies from 5% to 20% in patients with hypertension but is largely underdiagnosed. Expanding screening for PA to all patients with hypertension to improve diagnostic efficiency is needed. A novel and portable prediction tool that can expand screening for PA is highly desirable.

METHODS

Clinical characteristics and laboratory data of 1,314 patients with hypertension were collected for modeling and randomly divided into a training cohort (919 of 1,314, 70%) and an internal validation cohort (395 of 1,314, 30%). Additionally, an external dataset (n = 285) was used for model validation. Machine learning algorithms were applied to develop a discriminant model. Sensitivity, specificity, and accuracy were used to evaluate the performance of the model.

RESULTS

Seven independent risk factors for predicting PA were identified, including age, sex, hypokalemia, serum sodium, serum sodium-to-potassium ratio, anion gap, and alkaline urine. The prediction model showed sufficient predictive accuracy, with area under the curve (AUC) values of 0.839 (95% CI: 0.81-0.87), 0.814 (95% CI: 0.77-0.86), and 0.839 (95% CI: 0.79-0.89) in the training set, internal validation, and external validation set, respectively. The calibration curves exhibited good agreement between the predictive risk of the model and the actual risk. An online prediction model was developed to make the model more portable to use.

CONCLUSION

The online prediction model we constructed using conventional clinical characteristics and laboratory tests is portable and reliable. This allowed it to be widely used not only in the hospital but also in community health service centers and may help to improve the diagnostic efficiency of PA.

Performance Assessment of Certain Machine Learning Models for Predicting the Major Depressive Disorder among IT Professionals during Pandemic times

Major depressive disorder (MDD) is the most common mental disorder in the present day as all individuals' lives, irrespective of being employed or unemployed, is going through the depression phase at least once in their lifetime. In simple terms, it is a mood disturbance that can persist for an individual for more than a few weeks to months. In MDD, in most cases, the individuals do not consult a professional, and even if being consulted, the results are not significant as the individuals find it challenging to identify whether they are depressed or not. Depression, most of the time, cooccurs with anxiety and leads to suicide in few cases, among the employees, who are about to handle the pressure at work and home and mostly unnoticing such problems. This is why this work aims to analyze the IT employees who are mostly working with targets. The artificial neural network, which is modeled loosely like the brain, has proved in recent days that it can perform better than most of the classification algorithms. This study has implemented the multilayered neural perceptron and experimented with the backpropagation technique over the data samples collected from IT professionals. This study aims to develop a model that can classify depressed individuals from those who are not depressed effectively with the data collected from them manually and through sensors. The results show that deep-MLP with backpropagation outperforms other machine learning-based models for effective classification.