An expert in machine learning and a psychologist may seem like an unlikely pairing. Artificial intelligence may be able to help make mental health care more accessible to patients, according to Rosalind Picard and Paola Pedrelli of the Massachusetts General Hospital.

A machine can become very good at a task if it has access to many data and examples of good behaviour (i.e., what output it should produce when it sees a particular input). It can also aid in discovering meaningful patterns that humans might not have learned as quickly without the assistance of a computer. Picard and Pedrelli can collect detailed data on the skin conductance and temperature, heart rate, activity levels, socialisation, self-assessment of depression, sleep patterns, and more using wearable devices and smartphones of study participants. 

Ultimately, they want to create machine learning algorithms that can make sense of this data and determine when and how individuals struggle and what might assist them. Physicians and patients alike can benefit in the long run from the algorithms developed by these researchers, they say.

Psychology and artificial intelligence

Picard and Pedrelli's current work at MIT's Jameel Clinic goes a step further than previous research, which focused on determining if machine learning could use data to identify a participant's current emotional state. They want to find out if machine learning can predict the trajectory of a disorder, identify changes in an individual's behaviour, and provide data that researchers can use to tailor medical care for each patient.

Researchers must design a tool that will empower and uplift its users and develop practical machine-learning algorithms. "The question we're focusing on now is, how will machine-learning algorithms help people?" says Picard. Moreover, the team consider how machine-learning algorithms may present their findings to users: a new device, a smartphone app, or even notifying a doctor or family member.

Imagine a technology that tracks a person's sleep patterns, home habits, and heart rate. These changes may be so subtle that the person and their loved ones are unaware. Machine-learning algorithms may map these data to the individual's past experiences and other users' experiences. The technology may encourage individuals to engage in past-beneficial behaviours or contact their physician.

Conclusion

Picard notes that artificial intelligence and machine-learning algorithms can make connections and identify patterns in large datasets that humans are less adept at noticing. "I believe there is a strong case to be made for the use of technology to help people improve their understanding of others."

• https://news.mit.edu/2022/deploying-machine-learning-improve-mental-health-rosalind-picard-0126
• Image source: Unsplash