When treating anxiety, depression and other mood related conditions it's critical for psychologists to get a picture of a patient's condition over the day in order to tweak treatment and identify behavioural triggers. "Mood diaries" are helpful for the doctor but are often difficult for patients to keep up with. A new smartphone health app that records patients' emotions and generates a daily report for practitioners, could be the solution.
Once activated on the user's phone, the app called Xpression, records changes in a person's voice and the times the changes occurs and emails their psychologist each day. The app can recognise five distinct emotional states: calm, happy, sad, angry, anxious or frightened so a fairly complete emotional data record can assist their treatment.
The app must always be on to take an acoustic snapshot every second as the patient talks to others (and pets). Changes like loudness, pitch, intensity and pace of speech are all measured. The program sends 200-millisecond long acoustic snapshots to a server that has a machine-learning system that determines the emotion before sending the data back to the phone.
While having an app on and constantly recording sounds like a huge privacy risk, the creators have built it so that it doesn't record actual words - it's the acoustics of the app owner's voice that count and helps the program assess their emotional state.
Having such a tool could be a breakthrough for many patients who aren't able to keep accurate diaries. The learning gathered could help perfect critical treatment protocols.
The Xpression app is the work of Matt Dobson and Duncan Barclay, founders of UK-based speech recognition firm El Technologies. The company is a finalist in a UK government competition to identify the nation's top mobile tech firms, and insurance companies are already expressing interest.
While speech and emotion recognition technology is already a known technology, the app would be a first of its kind and clinical trials are set to take place later this year.
If successful, it could provide a simple and effective tool for mental health care and help create a more effective link between patient and provider.
Via New Scientist