B.R.A.I.N.
RESEARCH
EPISODE DETECTION & FORECASTING
Every good thing in this world comes with a price. What B.R.A.I.N. is trying to do is to limit that price. Or at least shrink it, by detecting illness earlier. Every day a new obstacle, disease, hardship comes - whatever it is, B.R.A.I.N is already getting those things out of the way. And every single day, it is shrinking, shrinking, and shrinking.
PHYSICISTS + PHYSICIANS = B.R.A.I.N.
New mathematical models have the potential to provide insight into the underlying mechanisms of mental illness and revolutionize our understanding and treatment of mood disorders. However, major gaps need to be filled before we can apply these newer models to clinical practice. B.R.A.I.N. is a step forward in this direction.
The role of chaotic processes in the generation of mood fluctuations
We have shown that mathematical models, not only of groups, but also of individuals, can be used to forecast short-term changes based on current behavior (Ortiz et al, 2018). We have also shown that the underlying architecture of mood variability is in keeping with that of a chaotic system, suggesting that the window for episode prediction in BD will be inevitably short (Ortiz et al, 2021).
Using wearables for electronic (e-)monitoring in mood disorders
Our high recruitment rates, in combination with longitudinal follow-up and crisp clinical monitoring have surpasses expectations. Attrition rates are amongst the lowest in the world and adherence rates are close to 80% (Ortiz et al, 2023).
Missing data are not missing at random, and thus are informative
As discussed above, while adherence with e-monitoring in our own study has been higher than in most published studies, it was not perfect. We developed a new metric to characterize adherence with e-monitoring and showed that missing data were associated with participants’ clinical status, suggesting that these are not missing at random and can inform our models (Halabi et al, under review).