Technology / How machine learning is leading the fight against depression
How machine learning is leading the fight against depression
9 February 2018
Tomo, a chatbot that helps people with depression understand their mental health and state of mind, is the brain child of Fahad al Saud and Gus Booth-Clibborn, who developed the app after personal experiences with mental illness.
Al Saud said: “The chatbot was really born out of some of my own experiences. I have been through depression a number of times. There is a high rate of relapse among [sufferers of] depression and I thought I should make myself a little tool to just keep track of myself and keep on top of things.”
He and Booth-Clibborn wanted to create a chatbot that people could interact with as if they were having a natural conversation. They decided that a good way for the chatbot to communicate would be through text conversations, given that most people were familiar and comfortable with the format.
The app is based on evidence-based behavioural activation therapy, which is an established form of treatment for depression. Tomo helps users schedule activities and develop habits that can drive behavioural change. Each time a user completes a habit, Tomo asks them to share a photo. This is all done anonymously.
Tomo is still in its testing stages – the pair are currently undertaking research with mental health professionals and universities in developing the app. “The principle of behavioural activation is helping someone understand the patterns of their life, the habits that their depression creates and to unwind the cycle that feeds their depression,” says al Saud. “It is about identifying those cycles and retraining yourself. This is not an easy thing to do – it is a very gradual process, which needs to be broken down into little steps and supported.”
He explains that the app makes therapy a lot more user-friendly. “The way behavioural activation and a lot of cognitive behavioural tools are delivered are through worksheets and through homework,” he explains. “But when someone is in a very depressed state, the last thing they want to do is look at an Excel spreadsheet.”
To combat this problem, each interaction on Tomo lasts only a minute, and the average amount of time people check it is two or three times a day. The app then uses machine learning to track the user’s behaviours and can alert them when they show signs that they might be relapsing.
Al Saud says the chatbot will alert you if you haven’t seen your friend for three weeks, and that this may not be a problem now, but it could be in three months. “These little daily nudges is something AI can do very well,” he says.
When people are depressed, al Saud explains, it can also be difficult for them to know when to visit a GP, so giving them information about their state of mind can give them a nudge in the right direction. “There is a preventive care angle in this,” says al Saud. “It can alert people and help them course-correct early on, before things become a problem.”
Booth-Clibborn believes the app will extend the point of care for the user. He says: “We have the basis of a machine-learning system. These systems allow us to extend the treatment into someone’s life without them having to see a doctor once a week.
“By collecting data you begin to get a fix on how every factor of someone’s life affects their health. It is something which machine-learning technologies can really drive. It can lead to enormous changes in the way we treat people.”
This article was published in our Business Reporter Online: Revolutionising Healthcare.