New Advances in Treating Mood Disorders

Big data, predictive analytics and low-field magnetic stimulation are among the tools being used to improve treatment.

Predictive Analytics

In the 2002 movie Minority Report, based loosely on a short story by Philip K. Dick,  the government uses sophisticated computer analytics to predict “pre-crimes” – ie, crimes of murder that have not yet been but are about to be committed. The predictive models are so powerful and generate such confidence that the government has the power to arrest citizens even before they do anything, thus preventing crime on an unprecedented scale and driving the murder rate to zero. Issues of free will and determinism arise. Mayhem, of course, results.

But what if a powerful analytical tool could be used in collaboration with people with mood disorders? It would not predict “pre-crimes,” but instead would detect mood episodes before they become full blown and result in distress and disability. Think about it. Our current model of care and research of people with mood disorders depend on people’s ability to observe, store, and recall how they were doing — processes that paradoxically can be impaired by being in a mood episode.

What if instead we had a system that used real-time data that no longer required recall or any additional effort? What if we had a system that could alert patients and clinicians that an episode was imminent and that it was time to change treatment to prevent a mood episode?

A collaboration between Cogito® and MoodNetwork, funded by a small business innovation grant from the National Institute of Mental Health, will be testing an app that will passively collect real-time data from smartphones (GPS, accelerometer, texts, calls, web search activity, voice physics) to predict and monitor people with mood disorders. Cogito’s platform arose from preliminary work done in Alex Pentland’s Media Lab at the Massachusetts Institute of Technology (see the book Social Physics by Pentland for a broader description).

Big Data Collaborations and MoodNetwork

MoodNetwork,1 mentioned above, is part of the national Patient-Centered Clinical Research Network2 (PCORnet), which is an organization of networks — including big data from electronic health records as well as patient-powered research networks — that collaborates with patient communities, researchers, clinicians, and other stakeholders. The ultimate goal of PCORnet is to improve the health of the nation by generating data that inform clinical decisions made by patients and clinicians. The goal of MoodNetwork is to transform the lives of people with mood disorders by conducting prospective comparative effectiveness trials and outcomes research. Clinicians can invite their patients to visit the website and easily sign on to become a partner.

MoodNetwork aims to collaborate with 50,000 patient-partners (experts by experience) and provide them with tools to monitor their mood disorders and opportunities to “donate data.” Clinicians and healthcare systems can collaborate with the network to learn the outcomes of their patients.