Artificial Intelligence and Mental Health

By John Glass–Early detection of mental health issues from social media content can be critical effort reducing the incidence of depression and other mental health issues. Let me summarize key points highlighting the critical need for early detection of mental issues.

Some interesting observations in a PredictView Technologies sponsored research report published by Mohammed Hasanuzzaman

Mental illnesses represent a diverse group of medical conditions of varying severity, complexity, and duration. In considering deviations from normal thoughts, feelings, and behaviors which are characteristic of mental illness, it is critical to recognize the extent to which they lead to functional impairment. To be classified as a disorder, symptoms must cause significant suffering or interfere with daily functions or life goals [1]. Depression, anxiety disorders, bipolar disorders, autism, and schizophrenia, to name but a few, are all included under the banner of mental illnesses.

Mental disorders affect individuals’ abilities to function, engage meaningfully in daily activities, and maintain relationships. For instance, a few of the symptoms and side effects of depression are insomnia, weight loss, fatigue, feelings of worthlessness, drug or alcohol abuse, and impaired ability to think, concentrate and make decisions. Anxiety disorders may cause physical symptoms, such as a fast heart rate, shakiness, muscle stiffness, chest pain, dizziness, and fatigue. Some disorders also engage in avoidance behaviors that negatively impact the disorder. As such, mental illnesses cause significant suffering to individuals but also to their families and are a significant source of socio-economic burden.1

Depression detection is a challenging problem as many of its symptoms are covert. Since depressed people socialize less, detection becomes difficult. Today, for the correct diagnosis of depression, a patient is evaluated on standard questionnaires. In literature, different tools for screening depression have been proposed, such as Personal Health Questionnaire Depression Scale (PHQ), Hamilton Depression Rating Scale (HDRS), Beck Depression Inventory (BDI), Center for Epidemiologic Studies Depression Scale (CES-D), Hospital Anxiety and Depression Scale (HADS), and the Montgomery and Asberg Depression Rating Scale (MADRS).4In particular, the eight-item PHQ-8 [3] is established as a valid diagnostic and severity measure for depressive disorders in many clinical studies [4].

The steadily increasing global burden of depression and mental illness is an impetus for developing more advanced, personalized, and automatic technologies that aid in its detection. Within computer science and artificial intelligence, different initiatives targeting mental health have been emerging for the past ten years [5].

Many mobile applications exist that use cognitive behavioral therapy (CBT), mindfulness training, mood monitoring, and cognitive skills training to treat depressive symptoms [6]. Such applications are gaining momentum as it has been shown that enabling users to.

Another trend targets the early detection of mental disorders [9]. Indeed, it is vital to identify subjects suffering from mental illnesses early to minimize the impact on public health. Within this scope, most efforts have been focusing on depression over social media analysis [10], as platforms such as Facebook, Twitter, or Reddit have become the place where people share their thoughts, feelings, and overall emotional status. Thus, techniques from machine learning, natural language processing, and clinical psychology are capable of inferring with reasonable accuracy whether an individual shows a change in behavior that could be related to mental disorders.

Over the last few years, many research studies in Computer Science have been proposed to deal with mental health disorders [12, 13]. Within this context, the automatic detection of depression has received major focus. Some initial initiatives have targeted the understanding of relevant descriptors that could be used in machine learning frameworks. [14] investigate the capabilities of automatic non-verbal behavior descriptors to identify indicators of psychological disorders such as depression.

Cloud-based solutions, such as Predictview’s first of its kind Mental Health Predictive Intelligence Platform (mPIP) and service, applies a sophisticated Artificial Intelligence engine capable of ingesting social media posts of those who ask to be a part of this service. Trillions of post elements for millions of employees are processed in real-time.  Posts are classified by advanced AI models that analyze post images, emojis and text (including text-embedded images).  Using Joint Representation, these models can recognize the contextual intent of the combined text and image in a post as potential signals of depression, anxiety, or substance use. Therapists now have tools for preventive outreach to those who may not know they need care.

1National Institute of Mental Health: https://tinyurl.com/y4t87njv

 2A statistic reported by the World Health Organization available at https://bit.ly/2rsqQoP.
3A study by Hannah Ritchie and Max Roser in 2018 available at https://bit.ly/2mnyVZ6.

4Recommandation of the French Haute Autorit ́e de la Sant ́e available at https://bit.ly/2EaOs92

[5] E. Aboujaoude and V. Starcevic, Mental health in the digital age: Grave dangers, great promise. Oxford University Press, 2015.

[6] P. Chandrashekar, “Do mental health mobile apps work: evidence and recommendations for designing high-efficacy mental health mobile apps,” Mhealth, vol. 4, 2018.

[7] S. D. Kauer, S. C. Reid, A. H. D. Crooke, A. Khor, S. J. C. Hearps, A. F. Jorm, L. Sanci, and G. Patton, “Self-monitoring using mobile phones in the early stages of adolescent depression: randomized

controlled trial,” Journal of medical Internet research, vol. 14, no. 3,

p. e1858, 2012.
[8] J. Firth, J. Torous, J. Nicholas, R. Carney, A. Pratap, S. Rosenbaum,

and J. Sarris, “The efficacy of smartphone-based mental health interventions for depressive symptoms: a meta-analysis of randomized controlled trials,” World Psychiatry, vol. 16, no. 3, pp. 287–298, 2017.

[9] Y. Li, R. Mihalcea, and S. R. Wilson, “Text-based detection and understanding of changes in mental health,” in International Conference on Social Informatics. Springer, 2018, pp. 176–188.

[10] F. Cacheda, D. Fernandez, F. J. Novoa, V. Carneiro, et al., “Early detection of depression: social network analysis and random forest techniques,” Journal of medical Internet research, vol. 21, no. 6, p. e12554, 2019.

[11] S. Chancellor, M. L. Birnbaum, E. D. Caine, V. M. Silenzio, and M. De Choudhury, “A taxonomy of ethical tensions in inferring mental health states from social media,” in Proceedings of the conference on fairness, accountability, and transparency, 2019, pp. 79–88.

12] G. Andersson and N. Titov, “Advantages and limitations of internet-based interventions for common mental disorders,” World Psychiatry, vol. 13, pp. 4–11, 02 2014.

[13] N. A. Dewan, J. S. Luo, and N. M. Lorenzi, Mental Health Practice in a Digital World: A Clinicians Guide. Springer Publishing Company, Incorporated, 2015.

[14] S. Scherer, G. Stratou, M. Mahmoud, J. Boberg, J. Gratch, A. Rizzo, and L.-P. Morency, “Automatic behavior descriptors for psychological disorder analysis,” in 10th IEEE International Conference and Workshops on Automatic Face and Gesture Recognition, 2013, pp. 1–8.

Category: Uncategorized