- Emerging technologies such as virtual reality and machine learning show enormous potential in the identification and treatment of mental health issues.
- Objective and immune to fatigue, artificial intelligence has several advantages in mental health diagnostics.
- These technologies will be able to create new kinds of patient experiences, potentially transforming mental healthcare.
Earlier in the year, I examined the role digital apps were playing in mental health diagnosis and treatment. The field has come a long way in a short time: popular meditation smartphone apps prompt people to pause and take some time out, while gamified programs encourage adolescents to share positive thoughts and fight depression by stealth.
These apps and others like them have helped provide crucial mental health services at scale to previously difficult-to-reach portions of the Australian population. Taking advantage of mature platforms such as smartphones, tablets and personal computers, individuals connect on their pre-existing devices, tapping into services and resources that would otherwise be difficult to locate.
On World Mental Health Day, it’s worth looking to the future for what the next generation of digital technologies holds for mental health diagnosis and treatment.
Spanning the technologies of artificial intelligence, machine learning and virtual reality, these cutting edge systems are a step up from the digital apps of old. Representing a new frontier of experience design, such technologies show considerable promise in assisting individuals with their wellbeing, from diagnosis through to treatment and management.
A new experience
Few technologies hold as much promise in the field of experience design and mental health treatment as virtual reality (VR). Already, it has been tested for use with many disorders and conditions.
Last year, a study¹ of past clinical trials using VR to treat phobias such as arachnophobia (fear of spiders) and acrophobia (fear of heights) found patients who underwent VR treatment performed significantly better in behavioral assessments. Another 2015 experiment² showed VR to be effective in assisting patients with post-traumatic stress disorder.
VR also holds great potential in treating persecutory delusions, a symptom of schizophrenia where an individual is suspicious of strangers or fears for their safety. In a 2016 study published in the British Journal of Psychiatry, patients with persecutory delusions engaged in two different VR therapy sessions³ simulating real-world environments like crowded trains or elevators.
The results were positive: with just 30 minutes of VR, a majority of patients in one of the sessions experienced a sharp reduction in delusions.
VR’s effectiveness begins with its ability to provide immersive experiences within a safe environment. Individuals might be more willing to undergo VR-simulated exposure therapy knowing it’s not real, yet the technology provides a comparable experience to traditional in vivo exposure therapies, where the patient must confront their fear head on.
Another advantage is portability. VR provides effective treatment for patients admitted into hospital care and logistically unable to seek external treatment. Its digital scalability also allows a single patient to undergo several different treatment scenarios within the same session.
While VR makes great strides in delivering therapeutic experiences for those with mental illnesses, another emerging digital technology, machine learning, assists in how these illnesses are diagnosed.
Machine learning uses patterns and algorithms to sort and define large bodies of raw, unstructured data. Used alongside natural language processing (which enables computers to analyse and form an understanding of sentiment in social media or contact centre transcripts), a recent experiment from the Columbia University Medical Center4 showed the potential for machine learning to predict the onset of certain mental illnesses.
In the proof-of-concept study, an AI program analysed interview transcripts from 34 clinically high-risk youths to see if any were likely to develop psychosis. In particular, the program looked out for common verbal identifiers of the condition, including truncated sentences, confused meanings, and the frequent use of words such as ‘a’, ‘this’ and ‘that’.
The program was able to successfully identify the five individuals who would later develop psychosis – a 100% success rate. In their findings, the study’s authors noted the AI outperformed classification from clinical interviews.
Machine learning has several advantages over a human mental health professional in diagnostics: it doesn’t get tired, it’s able to pick up subtle cues, and it can maintain considerable objectivity – bearing only the biases of its programming. However, it still requires a raw dataset in order to perform these analyses.
This opens up the possibility for machine learning to be paired with other digital devices such as wearables or trackers, receiving real-time data regarding a patient’s moods, sleep routines, heartbeat, rate of respiration and blood pressure.
Crunching these numbers and informed by mental health best practices, an AI-powered health assistant could know when to book you in for your next appointment, or learn which events you find the most stressful or relaxing. It could also use predictive analytics to advise you on which habits are good for you and which really aren’t.
Such a future AI assistant could provide a whole new kind of personalised healthcare experience, teaching you things you didn’t know about yourself.
A new kind
of patient experience
Mental health is a national issue, affecting 17.5% of the Australian population5 across all genders, ethnicities and generations.
While first-generation digital technologies have made some improvements in addressing the diagnosis and treatment of certain conditions, new technologies such as VR and AI can revolutionise them, delivering entirely new kinds of patient experiences.
Leveraging groundbreaking new technologies, these future digital health services are empathy machines, capable of discovering hidden symptoms within patient datasets or immersing them within carefully crafted experiences designed to treat and heal.