A Deep Learning Algorithm to Predict Hazardous Drinkers and the Severity of Alcohol-Related Problems Using K-NHANES

Identifying depression in the United States veterans using deep learning algorithms, NHANES 2005-2018

BACKGROUND

Depression is a common mental health problem among veterans, with high mortality. Despite the numerous conducted investigations, the prediction and identification of risk factors for depression are still severely limited. This study used a deep learning algorithm to identify depression in veterans and its factors associated with clinical manifestations.

METHODS

Our data originated from the National Health and Nutrition Examination Survey (2005-2018). A dataset of 2,546 veterans was identified using deep learning and five traditional machine learning algorithms with 10-fold cross-validation. Model performance was assessed by examining the area under the subject operating characteristic curve (AUC), accuracy, recall, specificity, precision, and F1 score.

RESULTS

Deep learning had the highest AUC (0.891, 95%CI 0.869-0.914) and specificity (0.906) in identifying depression in veterans. Further study on depression among veterans of different ages showed that the AUC values for deep learning were 0.929 (95%CI 0.904-0.955) in the middle-aged group and 0.924(95%CI 0.900-0.948) in the older age group. In addition to general health conditions, sleep difficulties, memory impairment, work incapacity, income, BMI, and chronic diseases, factors such as vitamins E and C, and palmitic acid were also identified as important influencing factors.

CONCLUSIONS

Compared with traditional machine learning methods, deep learning algorithms achieved optimal performance, making it conducive for identifying depression and its risk factors among veterans.

The use of artificial intelligence for delivery of essential health services across WHO regions: a scoping review

Background

Artificial intelligence (AI) is a broad outlet of computer science aimed at constructing machines capable of simulating and performing tasks usually done by human beings. The aim of this scoping review is to map existing evidence on the use of AI in the delivery of medical care.

Methods

We searched PubMed and Scopus in March 2022, screened identified records for eligibility, assessed full texts of potentially eligible publications, and extracted data from included studies in duplicate, resolving differences through discussion, arbitration, and consensus. We then conducted a narrative synthesis of extracted data.

Results

Several AI methods have been used to detect, diagnose, classify, manage, treat, and monitor the prognosis of various health issues. These AI models have been used in various health conditions, including communicable diseases, non-communicable diseases, and mental health.

Conclusions

Presently available evidence shows that AI models, predominantly deep learning, and machine learning, can significantly advance medical care delivery regarding the detection, diagnosis, management, and monitoring the prognosis of different illnesses.

Time perspective and family history of alcohol dependence moderate the effect of depression on alcohol dependence: A study in Chinese psychiatric clinics

BACKGROUND

Depression and alcohol dependence (AD) are among the most prevalent psychiatric disorders that commonly co-occur. Therefore, gaining a better grasp of factors related to this comorbidity is particularly interesting for clinicians. Past research has highlighted the significant role that time perspective and family history of alcohol dependence (FH) play in the occurrence of depression and AD. However, much remains unexplored in the understanding of the association between them. This study explored how temporal profile and other sociodemographic characteristics of patients diagnosed with AD impact the severity of depression and AD in them.

METHODS

This study was multi-centered, including 381 patients. Cross-sectional information was collected from both inpatient and outpatient psychiatric clinics in China. Data were acquired using validated self-report scales, including Michigan Alcoholism Screening Test, Zung Self-Rating Depression Scale, and Zimbardo Time Perspective Inventory-Chinese version. Multiple linear regression analyzes were conducted to control social demographic variables and construct prediction models to inspect the influence factors of variables. Moderation models were constructed to inspect further interplay between variables using hierarchical regression and PROCESS Macro.

RESULTS

Results showed that of all the patients in Chinese psychiatry clinics diagnosed with AD according to the International Classification of Diseases-10, 59.9% met the criteria of depression according to the questionnaire, and time perspective was correlated with the severity of depression. Furthermore, using regression analysis, we found that time perspective and depression could predict AD severity. The moderating role of a past negative time perspective and FH was confirmed between depression and AD. We found that, in our study, only in patients with FH and relatively moderate to high scores of past negative time perspective could the severity of depression predict the severity of AD. Therefore, during the treatment and care of patients with AD, their depression level, time perspective score, and FH should be considered.

Machine learning model for depression based on heavy metals among aging people: A study with National Health and Nutrition Examination Survey 2017–2018

Objective

To explore the association between depression and blood metal elements, we conducted this machine learning model fitting research.

Methods

Datasets from the National Health and Nutrition Examination Survey (NHANES) in 2017–2018 were downloaded (https://www.cdc.gov/nchs/nhanes). After screening, 3,247 aging samples with 10 different metals [lead (Pb), mercury (Hg), cadmium (Cd), manganese (Mn), selenium (Se), chromium (Cr), cobalt (Co), inorganic mercury (InHg), methylmercury (MeHg) and ethyl mercury (EtHg)] were included. Eight machine learning algorithms were compared for analyzing metal and depression. After comparison, XGBoost showed optimal effects. Poisson regression and XGBoost model (a kind of decision tree algorithm) were conducted to find the risk factors and prediction for depression.

Results

A total of 344 individuals out of 3247 participants were diagnosed with depression. In the Poisson model, we found Cd (β = 0.22, P = 0.00000941), EtHg (β = 3.43, P = 0.003216), and Hg (β=-0.15, P = 0.001524) were related with depression. XGBoost model was the suitable algorithm for the evaluation of depression, the accuracy was 0.89 with 95%CI (0.87, 0.92) and Kappa value was 0.006. Area under the curve (AUC) was 0.88. After that, an online XGBoost application for depression prediction was developed.

Conclusion

Blood heavy metals, especially Cd, EtHg, and Hg were significantly associated with depression and the prediction of depression was imperative.