Canada

Depression and suicidal thoughts are significant mental health concerns frequently expressed on social media platforms like Reddit. This study uses advanced machine learning techniques to analyze and classify posts from the Reddit dataset to identify and categorize high-risk content. By employing Latent Dirichlet Allocation (LDA) for topic modeling, we identify common themes in the posts, such as expressions of despair and cries for help [1]. We then apply clustering algorithms to group similar posts, uncovering underlying patterns and providing insights into frequent triggers and shared experiences [2]. We use classification models to accurately categorize posts and detect those indicating severe distress, including Logistic Regression, Support Vector Machines (SVM), and Neural Networks [3]. This methodology is implemented in a web application designed to provide real-time categorization of new posts, aiding mental health professionals and social media moderators in quickly identifying and responding to high-risk content. The application invites users to interact, processing their inputs via a text box. Visualization tools, such as 2D and 3D scatter plots, clearly display the data, allowing users to understand how posts are categorized and fit into broader trends [4]. This research improves our understanding of how mental health issues are expressed on social media, highlighting the importance of advanced data analytics in creating practical intervention tools. Our approach demonstrates the potential of combining psychological assessment with sophisticated computational methods to develop a comprehensive system that benefits research and individual users.

Leveraging Open Data Analytics and Machine Learning to Improve Mental Health Research and Innovation

Advanced Machine Learning Techniques for Early Detection and Classification of Depression and Suicidal Tendencies on Social Media

technical paper

The University Big Data Challenge (IUBDC) successfully concluded its latest edition, bringing together national health industry leaders, policymakers, and dignitaries for a conference centered on mental health research and innovation. Participants had the opportunity to showcase their projects in a poster session, engaging with representatives from industry and academia. The event provided a rich network of industry and academic professionals, offering pathways for scholarly publication, access to curated datasets, learning materials materials, and workshops.

Held at the Microsoft Headquarters in Toronto on July 26, 2024, the challenge featured hands-on data science and AI research projects focused on mental health. For over seven years, IUBDC has fostered collaboration between data science and health professionals and motivated university students. This year’s projects, which spanned topics from dementia to addictions, aimed to enhance our understanding of mental health, improve patient lifestyles, and reduce disease risk. Abstracts from teams were published in the NRC Research Press STEM Fellowship Journal, with all team's works featured by Underline publications.

2024 Inter-University Big Data And AI Challenge -

Thee challenge featured hands-on data science and AI research projects focused on mental health. 

Mental health issues are increasingly being addressed through machine learning techniques, which offer new possibilities for analyzing and understanding psychological conditions. This study focuses on detecting depression severity using advanced natural language processing (NLP) and sentiment analysis algorithms. Leveraging the capabilities of machine learning, we aim to provide a robust framework for assessing mental health by analyzing social media data. Our approach involves the use of three distinct models: EMOTION ENGLISH DISTILROBERTA-BASE for extracting emotional features, MULTIWD for assessing various dimensions of wellness, and Vader for sentiment analysis. These models generate a comprehensive feature set, which is then used to compute depression severity, categorized into five levels—Minimal, Mild, Moderate, Moderately Severe, and Severe—based on the PHQ-9 questionnaire. To evaluate the effectiveness of our approach, we compared the calculated severity classes with results from Latent Dirichlet Allocation (LDA) for topic modeling and K-Means clustering for numerical feature-based classification. Our findings indicate that K-Means clustering achieved a 76% accuracy in aligning severity classes, which surpasses the LDA accuracy 63%. This suggests that numerical features may offer more reliable clustering for mental health severity than topic-based methods. This study highlights the potential of combining NLP and sentiment analysis techniques in mental health assessments and suggests directions for further refinement and research.


Model Based Severity Analysis of Mental Health Issues Using NLP and Sentiment Analysis On Social Media Data (SENTI-SEV)

Depression is popularly known as major depressive disorder (MDD), which is a common but serious medical sickness that negatively affects how an individual feels, the way they think, and possibly how they act. It potentially leads to many emotional and physical breakdowns and can reduce a person’s capability to perform at work as well as at home. The project employs a robust methodology to analyze the relationships between depression and various risk factors using the National Health and Nutrition Examination Survey (NHANES) dataset. The project involves data acquisition and preprocessing, depression score calculation using the PHQ-9 questionnaire, and technical infrastructure setup. Data aggregation, preprocessing, and exploratory data analysis (EDA) are performed using Python and its libraries. Hypotheses are formulated and tested using statistical methods, and machine learning techniques are applied for predictive modeling. Feature importance and hyperparameter tuning are used to improve the models.The project identifies key factors associated with depression, including socioeconomic features, alcohol and drug consumption, and mental health measures. Predictive models are being developed to predict depression levels based on multiple variables. A web interface with interactive dashboards is being designed to visualize key insights and model predictions. The findings have the potential to inform and shape public health policies and interventions. Also, successfully identifying and analyzing these complex associations will provide better insights into the whole picture of mental health treatment, particularly depression, which ultimately improves the effectiveness of public health strategies and clinical practices. Every finding in this research project might be beneficial not only to patients with depression but also to the academic community and public health policymakers.

Deciphering Depression: A Multivariate Analysis of Influential Factors and Their Relationships

Depression is a widespread mental health disorder affecting millions worldwide. In the age of social media, it is crucial to leverage technology to detect, help, and support individuals displaying symptoms online. Our investigation revealed that social media, particularly Reddit, has a wide demographic of people and hosts a variety of discussion topics, including a significant number of posts on mental health. Using the large amounts of data on Reddit, the goal is to train and develop various machine learning models to recognize manifestations of depression and evaluating their performance. These models include machine learning and deep learning methods, such as Support Vector Machines (SVMs), transformers, and various Recurrent Neural Network (RNN) architectures — simple RNNs, Gated Recurrent Units (GRUs) and Long Short-Term Memory (LSTM) units — for sentiment analysis and depression detection on Reddit. The objective is to evaluate and enhance the models’ accuracy. Early detection facilitated by such models could potentially mitigate the impact of mental health disorders through timely intervention and support mechanisms.

Using Machine Learning and Deep Learning for Depression Detection on Social Media

Anxiety disorder (AD) is a condition associated with feelings of dread and an incapability to perform daily tasks. Symptoms can be physical, such as a rapid pulse and sweating, and emotional, including constant stress and difficulty concentrating. Anxiety disorders are subjective, variable and often overlap with underlying conditions, complicating detection and treatment. This study focuses on three factors, implicit memory, reaction time, and reaction to a perceived threat—all measured in previous studies and correlated with anxiety levels. Our goal is to determine if anxiety presence can be identified, using three games and one simulation. The first game centers around implicit memory, enabling us to perform tasks subconsciously. Anxious individuals exhibit an implicit memory bias for threat-related information. The second game will associate a level of anxiety to reaction times. The third game first makes the user complete a regular task and progressively introduces a threat to the successful completion of the task. Attentional bias shows that anxious people are more likely to respond to threat-related stimuli rather than stick to their task. Finally, the user will take part in an artificial anxious situation. In the first stage, the user plays a regular game. In the second, distracting stimuli appear. In the third, the user’s available resources decrease. Throughout the stages, both the user’s effectiveness and efficiency will be evaluated. If the user experiences anxiety, efficiency should decrease in stage 2 and effectiveness should decrease in stage 3. Based on their results, they will receive an assessment of their optimal stress level for highest performance. The results from all of three games will be compiled to provide an anxiety assessment. The player will then be prompted to complete the Hamilton Anxiety Rating Scale (HAM-A) which will be used to verify whether the game results match their true anxiety levels.


Using Interactive Games as Tools for Gauging Anxiety

Impostor syndrome is defined as a combination of self-perceived fraudulence, achievement pressure, and negative emotions [3], which creates doubts in one's capacity, leading to unfavorable comparisons with others. Our study represents an approach to evaluate individuals based on a set of questions from the Young Impostor Syndrome (YIS) scale [4] built within a web application. These questions assist users in identifying impostor syndrome symptoms, such as self-doubt, and use Likert-scale responses ranging from ‘agree’ to ‘disagree’ to calculate a more accurate impostor syndrome metric. Individuals are encouraged to provide personal information such as age and years of experience in the field, enabling users to attribute their accomplishments to their own ability. Furthermore, our initial data, from a dataset predicting impostor syndrome in medical students [2], is processed as the basis for our analysis. We then expand this data with online user responses, using a cloud solution to store their inputs. The data is then presented as clusters to the user so they can observe how many other individuals are dealing with similar kinds of issues. This enables the investigation of relationships between impostor syndrome traits and demographics. We use clustering algorithms [1] to place users into groups according to their profiles. Users can see where those clusters fit in relation to others by viewing them in 2D and 3D scatter plots. Visualizations help to better understand the frequency and the variability of this mental health phenomenon among demographic groups, in addition to quantifying it. This study adds to our understanding of the intricacy and effects of impostor syndrome. Users’ experience can be contextualized by the visual representation, which promotes awareness and debate of the psychological phenomenon. Our method emphasizes how sophisticated data analytics and psychological testing may be used to create a comprehensive tool that helps research and individuals. Keywords: impostor syndrome, data analysis, clustering algorithms, visualization, web application, mental health, Likert scale

Imposter Syndrome Comparison Map

With the advent of machine learning becoming more and more integrated into healthcare, this study explores the efficacy of machine learning models in classifying and predicting various brain disorders and diseases such as schizophrenia, Alzheimer’s disease and Parkinson’s disease. Our review highlights the potential that ML has in identifying effective biomarkers, with cortical thickness emerging as a more accurate indicator for schizophrenia. Additionally, the connection between Alzheimer's disease and metabolic syndrome was made much more clear through ML algorithms, identifying eight diagnostic genes with high AUC values. The versatility of ML is further demonstrated through the application of natural language processing to analyze social media data, uncovering mental health issues in patients with diabetes. The challenges associated with ML use in a healthcare setting were also analyzed; the performance of ML models can drop substantially when encountering new data significantly different from the training set. While ML has great promise, it is recommended to use as a supplementary tool in collaboration with clinicians for the greatest efficacy. Continuous improvement in algorithms, data quality, and model validation will be essential to fully realize the potential of ML in improving patient care and outcomes.

Machine Learning in Neurological Disorders: A Comprehensive Review of Diagnostic and Predictive Applications

Stress affects over 79% of individuals globally, presenting a significant challenge for timely and accurate identification within mental health care. Traditional methods, which rely on self-reported data and sporadic clinical assessments, often result in delayed diagnosis and inadequate stress management. To address these challenges, this project employs artificial intelligence (AI) to develop a real-time stress level detection model using unsupervised learning algorithms. By integrating data from wearable devices, such as Apple Watches, with self-reported mood and activity logs, our model provides continuous and comprehensive stress analysis. Achieving an accuracy of 60% upon validation, the model demonstrates a significant step toward more effective stress management. This AI-driven approach shows promise in enabling early stress detection and personalized interventions. The findings underscore the potential for AI to enhance stress monitoring accuracy, thereby contributing to improved individual well-being and broader accessibility to quality mental health care. Future research will focus on refining the model to improve its accuracy and exploring additional data sources to further enhance its predictive capabilities.

Enhancing Wellbeing: An AI-Driven Stress Level Identifier

Music therapy is a recognized field of therapeutic intervention used in a diverse variety of contexts, including hospital settings, retirement homes, palliative care, and community programs. It leverages intrinsic qualities of music to address mental health issues and enhance overall mental health. A challenge for music therapists working in hospice is lack of suitable musical repertoires, called “music unpreparedness”. This study aims to use a machine learning model to evaluate repertoires and predict their effects on an individual’s mental health given their age and mental health condition, if any. Streaming services such as Spotify can also implement this model to improve the overall effects of services on listeners’ mental health–for instance, promoting playlists that will likely improve, rather than worsen, the user’s mental health on their home page. To train the model, an open-access dataset was used containing features such as favorite genre, age, and mental health condition. Categorical variables including ‘Fav genre’ and ‘Frequency [genre]’ were prepared using one-hot encoding and the dataset was split into training and testing sets, using a stratified approach to maintain the distribution of mental health conditions and music preferences. A decision tree model was chosen due to its interpretability and ability to handle categorical data. The model was trained with the entropy criterion and a maximum depth of 6, achieving a training accuracy of 82% and a validation accuracy of 75%. Key features influencing the model included listening hours/frequency, favorite genres, and mental health condition levels. The model provides insights into how these factors contribute to the perceived effect of music on mental health, offering valuable predictions that can enhance therapeutic interventions and user experiences on streaming platforms. Future improvements include incorporating more diverse datasets to improve the model's generalizability. Additionally, other machine learning algorithms can be explored to enhance prediction accuracy.

Predictive Models on the Therapeutic Effects of Diverse Music Repertoires on Mental Health

Mental Health is a subjective phenomenon for which there is only one way to diagnose, and that is knowing the root cause of this health problem. In many cases, people hesitate to consult for treatment due to discrimination, stigma, or fear of the unknown. This study shows how a Natural Language Processing (NLP) based Chatbot can be an interactive medium for patients to seek guidance for their mental health-related issues without hesitation and in a healthy and safe environment that will protect their privacy. NLP-based chatbots are governed by an artificial neural network (ANN) that is deployed as an app. The model is trained on a dataset with various questions, patterns, and responses. Based on the trained data, it provides 86.04% accurate responses after the training and could increase if its knowledge base is increased. This Chatbot can address issues like anxiety, depression, phobia, post-traumatic stress disorder (PTSD), and Schizophrenia. However, there is a need to expand the volume and variety of data and include a learning curve in the model to improve the responses. Long Short-Term Memory (LSTM) or GPT-3 can make the NLP-based Chatbot play a crucial role in elevating efficiency as it can foster personalized support with tailored needs of individuals.

AI For Emotional Wellbeing: Designing an NLP-based Mental Health Chatbot

Stress is a common experience reported to impact 30% of the global population with significant negative mental health effects and has sparked an increase in research aimed at stress detection. This study explores the effectiveness of using electrocardiogram (ECG) and electroencephalography (EEG) signals as predictors for stress detection using a multi-perceptron neural network. The model is trained and validated using a dataset from a study with 40 participants, 21 female and 19 male, at the Prince of Songkla University in Thailand. The participants were between the ages of 18 and 25 and performed mental arithmetic tasks (MAT) while their EEG and ECG signals were measured. The data is divided into 3 classes: EO, A1, A2 representing normal stress, high stress, and low stress levels in participants respectively. In addition to the participant’s gender included in both the EEG and ECG data, the EEG data contains the Fp1, Fp2, F3, F4, T3, T4, P3, P4 while the ECG data contains mean heart rate, AVNN, SDNN, NN50, pNN50, RMSSD, LF, LF Norm, HF, HF Norm, and L/F ratio. The model’s architecture including its learning rate, activation function, dropout rate, epoch, and layers were adjusted and evaluated to currently achieve 62% accuracy. In comparison to existing studies using the same dataset with various machine learning algorithms such as Adaptive Boosting, RF, SVM, k-Nearest Neighbor, LR, and NB, this model achieves similar accuracy, indicating competitive performance of the multi-perceptron with existing stress detection models. Future iterations of the model aim to improve the test accuracy and decrease the loss of the model by creating a training set using the nearest centroid neighbor (NCN) approach. The scarcity of data on EEG, ECG, and stress while training the model suggest a need for further data collection to further improve the model’s training set and accuracy.

Advanced Machine Learning Techniques for Early Detection and Classification of Depression and Suicidal Tendencies on Social Media

Downloads

Next from Leveraging Open Data Analytics and Machine Learning to Improve Mental Health Research and Innovation

Model Based Severity Analysis of Mental Health Issues Using NLP and Sentiment Analysis On Social Media Data (SENTI-SEV)

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES