The sessions in the evening started with focusing insufficient diagnostic potential of self-report.
The critical point was made that patients frequently do not recognize their own symptom change.
Relevant body phenomena correlate well with disease entity diagnosed with classical diagnostic criteria. For example, those who have the more severe form of PTSD show higher heart rate. Base on this finding, researchers developed wearable devices measuring physical signs and intervening the symptoms of PTSD.
It seems true that focuses are heavily endowed upon measurement in current psychiatry. Validity and reliability of psychiatric measurements have always been a major limitation in conventional epidemiology using big data.
It was truly interesting that speech became a hot biomarker for mental diseases.
Researchers have been exploring the feature of speech in correlation with clinical states; they matched disease domains with the pattern of speech.
It seems that they are utilizing EVERY source of information.
There was an interesting presentation about suicide and digital health. The suicidal rate has been so consistent from past to present; even though the number of suicide attempt seems to be reduced, the actual number of suicidal death remains constant.
We should consider that 1/3 people do not express their intention to suicide, and self-report is not enough. However, Implicit Association Test(IAT) can be one of the solutions to overcome this problem.
IAT score correlated well with suicidality.
With real-time monitoring using smartphones, patterns of suicidality can be categorized, using machine learning.
A real-time intervention of suicidal thoughts can be made by mobile apps.
It was so cool to know that there are certain apps developed for research. Beiwe is also a popular app that can measure an amount of information (you can get apps from app store, search for beiwe).
However, big data by digital assessments should always consider reproducibility. It’s not only the technology matters, but it’s the science we’re doing.