Alphabet’s X shares Amber EEG system to expand the quest for mental health biomarkers

Amber’s final EEG pro­to­type: Head­set, sen­sor strip and bioamp

Alphabet’s X details Project Amber, a quest for a sin­gle bio­mark­er for depres­sion that fell short of its goal (TechCrunch):

Alphabet’s X (the Google-owner’s so-called “Moon­shot Fac­to­ry”) pub­lished a new blog post today about Project Amber, a project it’s been work­ing on over the past three years — the results of which it’s now mak­ing avail­able open source for the rest of the men­tal health research com­mu­ni­ty to learn from, and hope­ful­ly build upon.

The X project sought to iden­ti­fy a spe­cif­ic bio­mark­er for depres­sion — it did not accom­plish that (and the researchers now believe that a sin­gle bio­mark­er for depres­sion and anx­i­ety like­ly didn’t exist), but X is still hop­ing that its work on using elec­troen­cephalog­ra­phy (EEG) com­bined with machine learn­ing to try to find one will be of ben­e­fit to oth­ers … What is per­haps most notable about this pur­suit, and the post today that Alpha­bet released detail­ing its efforts, is that it’s essen­tial­ly a sto­ry of a years-long inves­ti­ga­tion that didn’t work out — not the side of the moon­shot sto­ry you typ­i­cal­ly hear from big tech companies.

In fact, this is per­haps one of the best exam­ples yet of what crit­ics of many of the approach­es of large tech com­pa­nies fail to under­stand — that some prob­lems are not solv­able by solu­tions with analogs in the world of soft­ware and engineering.

The announcement:

Shar­ing Project Amber with the men­tal health com­mu­ni­ty (X blog post):

Today at the Sapi­en Labs Sym­po­sium, my col­league Vlad Miskovic pre­sent­ed insights from Project Amber, an ear­ly stage men­tal health project at X. Amber’s small team of neu­ro­sci­en­tists, hard­ware and soft­ware engi­neers, machine learn­ing researchers and med-tech prod­uct experts have been devel­op­ing pro­to­type tech­nolo­gies to help tack­le the huge and grow­ing prob­lem of men­tal health. After three years of explo­ration, we recent­ly wrapped up our work at X. Now we are mak­ing our tech­nol­o­gy and research find­ings freely avail­able in the hope that the men­tal health com­mu­ni­ty can build upon our work…

Here are three key insights from our user research:

  1. Men­tal health mea­sure­ment 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 set­tings. Rea­sons range from bur­den (“I don’t have time for this”) to skep­ti­cism (“Using a scale is no bet­ter than using my clin­i­cal judge­ment”) to lack of trust (“I don’t think my client is fill­ing this in truth­ful­ly” and ”I don’t want to reveal this much to my coun­sel­lor”). These find­ings were in line with the lit­er­a­ture on mea­sure­ment-based men­tal health care. Any new mea­sure­ment tool would have to over­come these bar­ri­ers by cre­at­ing clear val­ue for both the per­son with lived expe­ri­ence and the clinician.
  2. There is val­ue in com­bin­ing sub­jec­tive and objec­tive data. Peo­ple with lived expe­ri­ence and clin­i­cians both wel­comed the intro­duc­tion of objec­tive met­rics, but not as a replace­ment for sub­jec­tive assess­ment and ask­ing peo­ple about their expe­ri­ence and feel­ings. The com­bi­na­tion of sub­jec­tive and objec­tive met­rics was seen as espe­cial­ly pow­er­ful. Objec­tive met­rics might val­i­date the sub­jec­tive expe­ri­ence; or if the two diverge, that in itself is an inter­est­ing insight which pro­vides the start­ing point for a conversation.
  3. There are mul­ti­ple use cas­es for new mea­sure­ment tech­nol­o­gy. Our ini­tial hypoth­e­sis was that clin­i­cians might use a “brain­wave test” as a diag­nos­tic aid. How­ev­er, this con­cept got a luke­warm recep­tion. Men­tal health experts such as psy­chi­a­trists and clin­i­cal psy­chol­o­gists felt con­fi­dent in their abil­i­ty to diag­nose via clin­i­cal inter­view. Pri­ma­ry care physi­cians thought an EEG test could be use­ful, but only if it was con­duct­ed by a med­ical assis­tant before their con­sul­ta­tion with the patient, sim­i­lar to a blood pres­sure test. Coun­sel­lors and social work­ers don’t do diag­no­sis in their prac­tice, so it was irrel­e­vant to them. Some peo­ple with lived expe­ri­ence did not like the idea of being labelled as depressed by a machine. By con­trast, there was a notably strong inter­est in using tech­nol­o­gy as a tool for ongo­ing mon­i­tor­ing — cap­tur­ing changes in men­tal health state 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 on their own. They were also very inter­est­ed in EEG’s poten­tial pre­dic­tive qual­i­ties, e.g. pre­dict­ing who is like­ly to get more depressed in future. More research is need­ed to deter­mine how a tool such as EEG would be best deployed in clin­i­cal and coun­sel­ing set­tings, includ­ing how it could be com­bined with oth­er mea­sure­ment tech­nolo­gies such as dig­i­tal phenotyping.

News in Context:

About SharpBrains

SHARPBRAINS is an independent think-tank and consulting firm providing services at the frontier of applied neuroscience, health, leadership and innovation.
SHARPBRAINS es un think-tank y consultoría independiente proporcionando servicios para la neurociencia aplicada, salud, liderazgo e innovación.

Top Articles on Brain Health and Neuroplasticity

Top 10 Brain Teasers and Illusions

Newsletter

Subscribe to our e-newsletter

* indicates required

Got the book?