New DARPA initiative aims to harness cognitive science, sensors and machine learning to detect early brain signs of depression, anxiety, and suicidal ideation

New Cog­ni­tive Sci­ence Tool to Shed Light on Men­tal Health (DARPA press release):

Since Sept. 11, 2001, more than 30,000 active duty mem­bers and vet­er­ans have tak­en their own lives – four times as many as those killed in post-911 mil­i­tary oper­a­tions. Cur­rent meth­ods to detect ear­ly signs of behav­ioral and men­tal health risk fac­tors rely on self-report­ing and screen­ing ques­tion­naires, which can’t reli­ably pre­dict sui­ci­dal­i­ty. Effec­tive behav­ioral health assess­ment is a mis­sion-crit­i­cal capa­bil­i­ty requir­ing nov­el tools to iden­ti­fy and help those at risk.

Today, DARPA announced the Neur­al Evi­dence Aggre­ga­tion Tool (NEAT) pro­gram. NEAT aims to devel­op a new cog­ni­tive sci­ence tool that iden­ti­fies peo­ple at risk of sui­cide by using pre­con­scious brain sig­nals rather than ask­ing ques­tions and wait­ing for con­scious­ly fil­tered responses.

“NEAT is a proof-of-con­cept effort attempt­ing to devel­op a new tool for men­tal and behav­ioral health screen­ing that moves us beyond his­tor­i­cal and cur­rent meth­ods of ques­tions and con­scious­ly fil­tered respons­es,” said Greg Witkop, a for­mer Army sur­geon and cur­rent pro­gram man­ag­er in DARPA’s Defense Sci­ences Office. “Using the pre­con­scious will hope­ful­ly enable us to detect signs of depres­sion, anx­i­ety, or sui­ci­dal ideation ear­li­er and more reli­ably than ever before. If suc­cess­ful, NEAT will not only sig­nif­i­cant­ly aug­ment behav­ioral health screen­ing, but it could also serve as a new way to assess ulti­mate treat­ment effi­ca­cy, since patients will often tell their clin­i­cians what they think the clin­i­cian wants to hear rather than how they are tru­ly feeling.”

By way of anal­o­gy, NEAT is envi­sioned to be for men­tal health what an MRI is for the phys­i­cal body: a way to assess injury. As an MRI can detect an ear­ly menis­cal tear pri­or to a more seri­ous injury devel­op­ing that can impact a soldier’s readi­ness, NEAT would iden­ti­fy psy­cho­log­i­cal and behav­ioral changes before they impact readi­ness … “Ulti­mate­ly, NEAT intends to aug­ment cur­rent behav­ioral health screen­ing pro­grams by pro­vid­ing clin­i­cians with pre­vi­ous­ly unavail­able infor­ma­tion to enable ear­li­er inter­ven­tions and more reli­able mea­sures of suc­cess­ful treat­ment,” Witkop said. “Just as the objec­tive evi­dence of an X‑ray or MRI is some­times nec­es­sary to help a mil­i­tary mem­ber not feel like they are let­ting their unit down because of a vis­i­ble injury, NEAT will attempt to pro­vide objec­tive evi­dence of invis­i­ble injuries for help to be pro­vid­ed in time.”

More information and registration details:

A vir­tu­al Pro­posers Day for poten­tial pro­posers is sched­uled for March 15, 2022.

Neur­al Evi­dence Aggre­ga­tion Tool (NEAT) Pro­posers Day

• Pro­gram Objec­tive and Descrip­tion: NEAT will devel­op a nov­el cog­ni­tive sci­ence tool that could be used to aug­ment behav­ioral health screen­ings by accu­rate­ly detect­ing what some­one believes to be true. By bring­ing togeth­er recent advances in cog­ni­tive sci­ence, neu­ro­science, phys­i­o­log­i­cal sen­sors, data sci­ence and machine learn­ing, the NEAT pro­gram will devel­op process­es that can mea­sure what a per­son believes to be true by: 1) pre­sent­ing care­ful­ly craft­ed stim­uli that are designed to evoke spe­cif­ic pre­con­scious men­tal process­es; 2) detect­ing the result­ing pre­con­scious process­es using cur­rent phys­i­o­log­i­cal sen­sors com­bined with state-of-the-art sig­nal pro­cess­ing and neur­al ana­lyt­ics; and 3) using advances in machine learn­ing and data sci­ence to aggre­gate the pre­con­scious respons­es col­lect­ed across a set of stim­uli into a final mea­sure­ment that quan­ti­fies what a per­son believes to be true for a spe­cif­ic topic.

News in Context:

• DARPA-fund­ed non­sur­gi­cal neu­rotech­nolo­gies push the fron­tier of brain-machine interfaces
• Eight research teams work­ing with DARPA to dis­cov­er best ways to acti­vate neu­ro­plas­tic­i­ty and accel­er­ate learning
• Five rea­sons the future of brain enhance­ment is dig­i­tal, per­va­sive and (hope­ful­ly) bright
• 10 neu­rotech­nolo­gies about to trans­form brain enhance­ment and brain health