Artificial Intelligence (AI) has been a hot topic the past decade, where KPMG projects the investments in AI, machine learning and RPA to increase over the next 5 years to a staggering 232 billion dollars (KPMG, 2019). Looking at the healthcare sector, the technology is evolving to the point AI can already identify certain diseases such as cancer. However, in the mental healthcare, trials to use AI for detection of mental disorders (e.g. depression or borderline) are trending and results are somewhat promising. Yet, one of the main complications for AI in mental health are measurements of the physiological (e.g. electrodermal activity, heart rate) and psychological (e.g. shift in tone, usage of certain words) patterns.
Over the last couple of years, more and more people own a smartwatch and a smartphone. Mental health professionals and engineers have been trying to use wearables and smartphones as an instrument for the AI to assess the persons physiological and psychological state, after which the AI notifies a health care professional if an anomaly should occur.
But what if the AI could detect these tendencies of an underlying mental disorder? For example, an elevated heart rate (picked up by the wearable) could mean exercise or a simple scare. However, if it’s accompanied by sudden elevation of respiration (a.k.a. hyperventilation), or increased speed of speech (picked up by the smartphone) this may indicate a panic reaction to a stimuli, or simple arousal. Yet, if this happens multiple times a day, this could indicate a panic disorder or such. These physiological changes can all be gathered by the wearables and smartwatches and evaluated by the AI, which a healthcare professional could then use for the assessment of the mental being of the patient.
Thus far, research has shown that AI can at least as accurate detect a mental disorder as a trained healthcare professional can. Yet, mental healthcare professionals are not keen on the implementation of AI (Chandler, Foltz & Elvevåg, 2019). Examples of reasons are the trustworthiness of AI and machine learning in mental healthcare and the accuracy of vital sign measurements by wearables (Hahnen et al., 2020). For now, the best implementation of AI in mental health care is by means of applications, with which an AI can keep track of your mental health through questionnaires. So one day AI knows how you are feeling, but for now, you just have to tell him.
But, would you allow AI to track you mental health?
References
KPMG, (2019), Khube Mag: Intelligent automation edition.
Chandler, C., Foltz, P.W. & Elvevåg, B., 2019. Using Machine Learning in Psychiatry: The Need to Establish a Framework That Nurtures Trustworthiness. Available at: https://academic.oup.com/schizophreniabulletin/article-abstract/46/1/11/5611057 [Accessed September 29, 2020].
Hahnen, C. et al., 2020. Accuracy of Vital Signs Measurements by a Smartwatch and a Portable Health Device: Validation Study. Available at: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7055753/ [Accessed September 29, 2020].
