Technology takes on mental illness
Image credit: Getty Images
In times like these, technology seems to be helping us maintain our wellbeing, despite social isolation. Beyond personal use, how is it being applied to improve the state of mental health today?
According to the World Health Organisation, one in four people will be affected by mental or neurological disorders at some point in their lives. Around 450 million people currently suffer from such conditions, placing mental disorders among the leading causes of ill-health and disability.
In these coronavirus-affected times, mental health is likely to be affected, putting extra strain on already-under-pressure healthcare services. However, technology-assisted treatment is being developed that can maintain people’s wellbeing, or at least help.
Many clinical specialities benefit from technology-assisted treatments, advanced diagnostics and therapies; psychiatry – the study and treatment of mental illness, emotional disturbance and abnormal behaviour – does not.
Diagnosis of mental disorders and diseases is based on categorised symptoms laid out in the Diagnostic and Statistical Manual of Mental Disorders, formulated by mental health professionals. This is essentially a guide to ‘fitting’ symptoms into boxes, yet they can overlap.
Mental health disorders are difficult to diagnose. We still do not know causal pathways that lead to psychiatric illnesses, which makes it difficult to identify optimum treatment options. Although these disorders are diagnosed as discrete diseases (e.g. major depressive disorder), some diagnoses have variable symptoms, while similar symptoms can lead to different diagnoses.
This could be because the human brain has almost 100 billion neurons and 100 trillion connections and remains one of the greatest challenges in medicine.
Its complexity means underlying causes of neurological and psychiatric disorders like Alzheimer’s disease, epilepsy, schizophrenia and depression remain a mystery.
Research into neurotechnology and psychiatry needs to have the same level of investment as more ‘visible’ conditions, where ‘one size’ tends to fit most.
Natalia Ramsden, founder and director at SOFOS Associates, a London-based centre specialising in ‘personal cognitive optimisation’, says technology has helped us measure and track data in a way that we couldn’t before, “and this has been a significant advancement. We can track symptoms and experience, but also measure what is happening inside the body and brain. This strengthens our diagnostic capabilities and helps individuals take control and accountability of their situation.”
‘We can track symptoms and experience, but also measure what is happening inside the body and brain. This strengthens our diagnostic capabilities.’
Unfortunately, not everyone can reach out for help. The still existing stigma in society, lack of therapists and - in some countries - qualification of specialists, are all factors hampering individuals trying to access the appropriate care.
Thankfully, technology-assisted treatment delivered online through computers, smartphones, virtual-reality applications, or videoconferencing, can help. Commonly known as e-mental health services, this kind of care is accessible, and can be anonymous.
In this trying time, when mental health may take a hit, the website Psyberguide can help you find apps that may be useful.
This can be more appealing and easier to continue than traditional care, but to ensure app developers are providing beneficial content, they must adhere to scientific evidence, traditional methods and consultation with experts. Apps need to be clear about what they can help with.
However, a recent article from independent global media organisation openDemocracy says there is a lack of “robust and independent trials” to test effectiveness of healthcare apps. A study from BMJ Journals of mental-health-related apps showed 85 per cent of them provided no reliable evidence of effectiveness.
As well as providing a form of care, apps can also collect data (with a guarantee of privacy for users) such as location, movement and phone use, which could help in emergency situations, and help personalise programs for universal use.
Researchers are also using data collection to further understand the human brain and help more effectively, and sooner, when people show signs of mental health issues.
The team behind one such project, at the Alan Turing Institute, say predicting mental health disorders early and developing personalised interventions would have major implications for management and treatment. However, they note that progress in early diagnosis and personalised treatment “is compromised by heterogeneity in patient populations” – in other words, it’s hard to detect common risk factors.
The project, called ‘AI for precision mental health: Data-driven healthcare solutions’, aims to develop predictive models based on machine learning approaches, to differentiate between people without symptoms who are at high versus low risk of mental health related disease.
If successful, the resulting AI-guided patient selection could potentially improve clinical trial success rates and get more companies interested in investing in new drug discoveries for mental health disorders.
The recent developments suggest that we will soon have an AI revolution in mental health with better access, care and cost.
AI’s natural next step would be to oversee our wellbeing. So would psychologists and doctors who specialise in treating people with mental health disorders become digitalised? The answer is most likely – AI could give more opportunity for people to improve their mental wellbeing.
An article from Ukraine-based IT company Sciforce says that the most prospective domains for application of AI techniques are computational psychiatry and specialised chatbots, which could serve as therapeutic services. Computational psychiatry combines numerous levels and types of computation (including machine learning) with multiple types of data to improve understanding, diagnostics, prediction and treatment of mental disorders.
Sciforce explains that one example of using machine learning is application of algorithms to predict specific medication. Although clinicians cannot guarantee whether a patient will respond positively to a medication, effectiveness of treatment can be improved by matching patients to interventions.
Another way of using AI to help mental health could be as simple as a speech-based mobile app. Recently, a team from the University of Colorado used machine-learning AI to develop a system that can detect day-to-day changes in speech. Subtle differences can suggest a mental health decline – shifts in tone or pace can indicate mania or depression, illogical sentence patterns can suggest schizophrenia, and memory loss can be a symptom of cognitive and/or mental health problems.
Machine learning is also being used in the field of neuroscience, where the volume of data generated by neurotechnology is immense and requires big-data solutions.
Margaret C Grabb, program officer at the US-based National Institute of Mental Health (NIMH), says that researchers have reported new findings from this data. “By using a machine-learning algorithm developed for analysing EEG [electroencephalogram – electrical brain activity] recordings, the researchers were able to identify EEG-based biomarkers that can predict which patients will respond to sertraline, an antidepressant drug. The researchers have extended the findings using independent data sets. If validated, these kinds of predictive measures could be used in the clinic to better match treatments to patients and could minimise the trial and error involved in medication management.”
Grabb says research in neuroscience and mental health is rapidly expanding, with new developments in our understanding of brain structure and function and by exploring differences in psychiatric disorders.
However, she adds that fundamental advancements in neurotechnology are needed at all levels, from the microscopic to the whole brain, in order to improve imaging speed, depth/spatial/time resolution, and the automation of analytic processes.
“Likewise, electrophysiology enables researchers to study brain activity in individuals and look for aberrant signals in patients with psychiatric disorders, but there are still limits to these techniques,” she says.
Grabb adds it could be improved by ease of application, data quality, multi-electrode scale and depth, and correlation with neuroimaging and behaviour, as well as the ability to record and streamline processing and analyses, “and develop commercial-grade analytical pipelines”.
Another method gaining momentum is precision medicine, which considers genes, environment and lifestyle. It could be revolutionary in changing how we view traditional medicine. There are ongoing large-scale collaborative research efforts that look at the genome (a person’s complete set of DNA) and exome (the protein-coding portion of DNA) to better understand the role of genetics in psychiatric disorders.
Grabb says the hope is that these projects may shed light on potential neurobiological mechanisms and targets for treatment. She cites one example, published by NIMH in March 2020, where researchers examined the human exome in a large sample of people with schizophrenia. “These large-scale psychiatric genetics studies will enhance our understanding of mental disorders, but genetics alone may not provide the precision needed for tailoring treatments,” she warns.
NIMH has focused on the development of central nervous system measures (structural/functional neuroimaging, electrophysiology) combined with observable patient characteristics and unbiased genomic screening, to discover biosignatures or subgroups of patients that are biologically similar, and therefore may respond to treatments more consistently.
Brain-stimulation technologies, such as electroconvulsive therapy and transcranial magnetic stimulation, treat symptoms of major depressive disorder and can be effective for patients with treatment-resistant depression. Researchers are exploring further to better understand what makes brain-stimulation technologies effective and identify biomarkers that may predict treatment response.
Clinical researchers are using stimulation technology to target specific brain regions to selectively excite or inhibit circuit function. Investigations into combination therapies are testing whether regionally targeted brain stimulation enhances effects of behavioural therapy and pharmacologic treatment.
Grabb says neurosurgical approaches to brain stimulation have also progressed. “Deep-brain stimulation (DBS), initially developed to treat motor symptoms in movement disorders, has more recently been applied to target treatment-resistant depression.
“New research funded through the BRAIN Initiative [Brain Research through Advancing Innovative Neurotechnologies] is supporting the use of next-generation approaches to DBS for major depressive disorder, obsessive compulsive disorder, and post-traumatic stress disorder.” These new approaches measures brain activity from target areas, selectively engaging specific brain networks and adjusting stimulation based on symptom readout.
In some of these systems, stimulation is adjusted based on closed-loop stimulation, or neural readouts. The Darpa Systems-Based Neurotechnology for Emerging Therapies programme is investigating the use of intracranial multielectrode recordings to establish models that associate neural activity with mood state. Grabb says the aim is to apply these models to patients who are continuously monitored via electrophysiology to inform a brain stimulation protocol.
“Over the next few years, we will see which of these engineering approaches show promise as potential future therapeutics in psychiatry,” she adds.
A less invasive approach that you can use at home is wearables, with studies claiming it to have positive mental health effects.
Kamran Adle, early-stage investor in the Future of Health team at Octopus Ventures, says a head wearable called Flow is “a CE-certified drug-free treatment for depression which uses a wearable headset and an app therapy programme. It uses transcranial direct current stimulation, a mild electrical signal, to stimulate and rebalance neural activity.”
In 2019, the US Food and Drug Administration approved app-based Cervella, a cranial electrotherapy stimulator (CES), for treatment of anxiety, depression and insomnia. It is the first CES to be integrated into noise-cancelling Bluetooth headphones, and delivers micro pulses of electrical current across the brain.
Ramsden of SOFOS Associates says audio-visual entrainment and cranio-electro stimulation are a non-invasive way to achieve peak mental and physical wellbeing, “showing success with conditions such as anxiety, depression, ADD and ADHD. These devices are used based on recommendations of a neuropsychologist and people are not recommended to self-prescribe.”
Adle points to the increasing trend of virtual-reality therapy. “It’s another of the services provided by The Institute for Mental Health, as well as Dutch start-up Psylaris. A patient can receive personalised therapy at any time by supplementing regular face-to-face therapy with independent sessions in virtual reality.”
It seems that the human mind, the most sophisticated computer in the universe, will always have its mysteries – particularly when it comes to less tangible aspects such as moods and depression. Yet, fortunately, technology is starting to unlock its secrets.
Sign up to the E&T News e-mail to get great stories like this delivered to your inbox every day.