Alphabet & # 39; s X (the Google owner's "Moonshot Factory") today released a new blog post about Project Amber, a project they have been working on for the past three years. The results are now being made available as open source to the rest of the mental health research community, from which to learn, and hopefully build upon. The X project tried to identify a specific biomarker for depression – it failed (and the researchers now believe there probably wasn't a single biomarker for depression and anxiety), but X is still hoping that its work on using the Electroencephalography (EEG) combined with machine learning to try and find one will benefit others.
The X researchers hoped that depression, like other diseases and disorders, might have a clear biomarker that would help health professionals diagnose depression more easily and more objectively, which would hopefully also make it easier and more consistent to treat them. With the EEG, there was some precedent from studies conducted in labs with purpose-built games in which people with depression consistently appeared to have lower levels of EEG activity in order to effectively "win" the games.
These studies seemed to offer a route to a potential biomarker, but to make them actually useful in real-world diagnostic settings like a clinic or public health laboratory, the team at X set out to improve the process of EEG collection and interpretation to make it more accessible to both users and technicians.
Perhaps the most notable thing about this pursuit, and the post published today, which Alphabet published extensively on its endeavors, is that it is essentially a year-long investigation that has failed – not the side of the Moonshot story, from which You usually hear about big tech companies.
In fact, this is perhaps one of the best examples of what critics of many approaches by large tech companies fail to understand – that some problems are not solvable by analogous solutions in the world of software and engineering. The team at X summarizes what they have learned from years of research into three main points of their user research, and each of them touches on in some way the inadequacy of a purely objective biomarker detection method (even if it had worked). especially when it comes to mental illness. From the researchers:
Measuring mental health remains an unsolved problem. Despite the availability of many mental health surveys and scales, they are not widely used, especially in primary care and counseling. The reasons range from stress (“I don't have time for it”) to skepticism (“Using a scale is no better than using my clinical judgment”) to a lack of trust (“I don't think my client fills this out”) honest ”and“ I don't want to tell my advisor so much ”). These results were consistent with the literature on measurement-based psychiatric care. Any new measuring instrument would have to overcome these obstacles by creating clear value for both the person with lived experience and the clinician.
It makes sense to combine subjective and objective data. Experienced people and clinicians both welcomed the introduction of objective metrics, but not as a substitute for subjective assessments and asking people about their experiences and feelings. The combination of subjective and objective metrics was seen as particularly powerful. Objective metrics could validate the subjective experience. or if the two diverge, that in itself is an interesting finding that forms the starting point for a conversation.
T.Here are several use cases for new metrology. Our initial hypothesis was that clinicians could use a "brainwave test" as a diagnostic aid. However, this concept was received only lukewarm. Mental health experts such as psychiatrists and clinical psychologists were confident they could diagnose through clinical interviews. General practitioners found an EEG test useful, but only if performed by a medical assistant prior to consulting the patient, similar to a blood pressure test. Counselors and social workers do not make a diagnosis in their practice, so this was irrelevant to them. Some people with lived experience didn't like the idea of being classified as depressed by a machine. In contrast, there has been a particularly strong interest in using technology as an ongoing monitoring tool – to track changes in mental health over time – to learn what happens between visits. Many clinicians asked if they could send the EEG system home so their patients and clients could repeat the test themselves. They were also very interested in the potential predictive qualities of the EEG, e.g. Predict who is likely to get more depressed in the future. Further research is needed to determine how an instrument like the EEG can best be used in clinical and consultative settings, including how it can be combined with other measurement technologies such as digital phenotyping.
X makes Amber & # 39; hardware and software open source on Github and also gives a & # 39; Patent promise & # 39; which ensures that by using the open source material, X will not take legal action against users of the EEG patents associated with Amber. It is unclear (though unlikely) that this would have been the result if Amber had found a single biomarker for depression, but perhaps in the hands of the wider community the team's work to make the EEG more accessible beyond specialized testing facilities, lead to other interesting discoveries.