Blog

Mental Health Awareness Month: “Affective Computing” 

In England, one in four people are affected by mental health disorders at some point in their life. One in six report a common mental health problem, such as anxiety and depression in a week. With almost two-thirds of people with mental health disorders not seeking help, there is a real need to reach these people in new ways. There are a variety of reasons why people don’t seek help, from the stigma still attached to mental health issues, to the inaccessibility of help. Inaccessibility is a global issue as there is less than one mental health worker for every hundred thousand people in most countries in the world. Especially since Covid struck, the healthcare industry as a whole has been put under immense strain and there is a need for new opportunities to help people. 

Artificial intelligence and machine learning has started to pave the way for innovative technology to help people in the comfort of their own homes. Initial results have had a positive response, however the long term efficacy has to still be thoroughly tested. 

There are two main ways that technology is helping; through measuring and tracking data, and through technology-assisted treatment. 

Affective Computing

Affective computing, or emotion AI, originated in the 1990s. One of its pioneers is Rosalind Picard, and she defined it as “computing that relates to, arises from, or deliberately influences emotions.” The technology can recognise, express and adapt to human emotions. It relies on sensors, voice analysis programs and machine learning to collect and analyse written text, physiological signals and physical cues to detect emotional changes. 

Data Measurement & Tracking

Technology can help doctors track a patient’s symptoms and can measure what’s going on in the body, including the brain. By using machine learning algorithms, the most suited medication and therapy can be predicted based on previous data. As of now, this can’t be used as a stand-alone tool, but rather as a second opinion to a clinician. 

Apps that help with mental health care can collect data which can help in emergencies. The project called “AI for precision mental health: Data-driven healthcare solutions” by the Alan Turing Institute shows that the data collected from these apps can help to personalise programs for universal use. The study aims to develop statistical models based on machine learning approaches, which can help to differentiate between low and high risk mental health related issues. Furthermore, this could enhance clinical trial efficacy, fuel pharma investment and guide tests to improve diagnosis and predictability of the patient outcome. 

Technology-assisted Treatment

In my opinion, the area that currently can help the most is technology-assisted treatment. This is where anyone who has access to a computer, smartphone, VR applications and video conferencing software (which has become all the rage since Covid) is able to have access to e-mental health services. These are much more widely available and importantly, can be anonymous, which would appeal to anyone uncomfortable with reaching out for help. 

There are sites such as “Psyberguide” which compiles hundreds of app reviews and recommends which apps to use for each relevant condition. “Togetherall” is a site that is popular among students, and has self-assessments, resources and creative tools as well as peer support. 

Mobile apps have more advanced tools such as AI and computational psychology that can act as specialised therapeutic chatbots. “Wysa” is one of the most popular mental health support apps, which consists of an AI chatbot that reacts to the user’s emotions. Users can practise cognitive behavioural therapy techniques as well as mindfulness exercises. If extra support is needed, users can get guidance from a qualified mental health professional. There are many apps that can also help users to track sleep, mood and give insights into the user’s personality. 

An interesting developing technology is AI in speech based apps. This is where a system can detect day-to-day changes in someone’s speech. “MyCoachConnect” was developed by a group of researchers at UCLA and is an AI based app that takes the user’s own words to give a personalised analysis based on shifts in tone or pace, which can be indicators for depression. It can also help track illogical sentence patterns, which could be an indicator for schizophrenia, and memory loss, which can indicate cognitive problems. 

The mental health technology and “affective computing” industry is forecasted to be an almost £30 billion industry by 2026. Facial expressions, heartbeat, eye blinks and speech are now becoming profitable sources of data as the industry expands. Also, as I mentioned previously, the Covid-19 pandemic has shifted a lot of life online, which makes affective computing a more appealing tool for governments to address the mental health crisis that we live with today. Intelligent computing in the mental health industry is still relatively in its infancy, with scientists disagreeing over the nature of emotions and how they are expressed in different populations. However, it is a rapidly growing field with more funding going towards research to aid its development. 

References;

“The incredible ways AI and technology is used in mental health”. Popovska, A. May 2021

Turing.ac.uk. Accessed May 2022

“Computing in mental health”. Calvo, R, Picard, R, Dinakar, K, Maes, P. May 2016

“The wellness industry’s risky embrace of Ai-driven mental health care”. Royer, A. October 2021

Image via source 

  • Written by Mikita Maru, May 20 2022