This Is Your Brain on Death

Newsweek International - - CONTENTS - BY JOSEPH FRANKEL @Joseph­frankel


at Carnegie Mel­lon Univer­sity, tries to take pic­tures of hu­man thoughts. He was giv­ing a talk about the way con­cepts are phys­i­cally rep­re­sented in the brain when his col­league David Brent, a psy­chi­a­trist, asked him if he could de­tect changes in the thoughts of peo­ple who are sui­ci­dal.

Just de­cided it was worth a shot. The two teamed up, and the an­swer, they con­cluded, was yes. In a pa­per pub­lished in late Oc­to­ber in Na­ture Hu­man Be­hav­iour, their team in­ves­ti­gated whether con­cepts like “death” and “life” are rep­re­sented dif­fer­ently in the brains of peo­ple who had sui­ci­dal thoughts, com­pared with the brains of those who did not.

Just’s group looked at 34 young adults: 17 who were hav­ing sui­ci­dal thoughts and 17 who were not plagued by such thoughts. The re­searchers mon­i­tored each par­tic­i­pant’s brain by func­tional mag­netic res­o­nance imag­ing (FMRI) as they were shown a se­ries of words that rep­re­sented dif­fer­ent con­cepts. Some (“care­free,” “com­fort,” “bliss”) were pos­i­tive; oth­ers (“ap­a­thy,” “death,” “des­per­ate”) could be re­lated to sui­cide. Oth­ers (“bore­dom,” “crit­i­cism,” “cru­elty”) were neg­a­tive and un­re­lated to sui­cide.

The brain “sig­na­tures” of those with sui­ci­dal thoughts looked sig­nif­i­cantly dif­fer­ent when shown the sui­ci­dal con­cept words. Of the words pre­sented, “death” evoked the most dra­mat­i­cally dif­fer­ent re­sponses.

The re­searchers then turned a ma­chine-learn­ing al­go­rithm loose on this data. The re­searchers “taught” the al­go­rithm by show­ing it the FMRI data and the clas­si­fi­ca­tion of the sub­ject (whether the in­di­vid­ual had thought about sui­cide) for all of the par­tic­i­pants ex­cept one. After the al­go­rithm was trained, the re­searchers gave it the FMRI data of the miss­ing par­tic­i­pant without telling it whether the per­son had been sui­ci­dal. The al­go­rithm was able to de­ter­mine whether that mys­tery

par­tic­i­pant had ex­pe­ri­enced sui­ci­dal thoughts with 91 per­cent ac­cu­racy.

It’s im­por­tant to note that this study doesn’t show a way of pre­dict­ing who will and won’t die by sui­cide. But the re­search, Just says, is head­ing in that di­rec­tion.

“It opens that pos­si­bil­ity,” says Glenn Saxe, a psy­chi­a­trist at New York Univer­sity’s School of Medicine who was not in­volved with the study. He notes that such a pre­dic­tor “could re­ally pow­er­fully add to the tool kits psy­chi­a­trists have.” How­ever, he cau­tions against over­in­ter­pret­ing the re­sults of this pa­per, which do not de­scribe a re­li­able means of pre­dict­ing sui­cide.

For a mat­ter as high stakes as de­ter­min­ing whether some­one will at­tempt sui­cide, ac­cu­racy is hugely im­por­tant. Paul Sa­jda, a re­searcher in neu­roimag­ing and com­put­ing at Columbia Univer­sity, noted that

The al­go­rithm was able to de­ter­mine whether a par­tic­i­pant had had sui­ci­dal thoughts with 91 per­cent ac­cu­racy.

while the al­go­rithm is 91 per­cent ac­cu­rate, the prospect of fail­ing to rec­og­nize even one per­son as sui­ci­dal, or falsely la­bel­ing the in­di­vid­ual as such, is dis­turb­ing. It also de­mands that re­searchers do a lot more work be­fore think­ing about us­ing this as a di­ag­nos­tic tool.

Sa­jda also won­ders: Why use an FMRI when cur­rent psy­cho­log­i­cal tests do about as good a job telling whether a per­son is sui­ci­dal? It was un­clear to him what added value a test with an FMRI would pro­vide.

Just and his team are aware of the po­ten­tial and the risks here. He says many steps would have to be taken be­fore this in­for­ma­tion could be safely used in as­sess­ing pa­tients. For one, he’d like to repli­cate his find­ings with a larger num­ber of peo­ple, although will­ing par­tic­i­pants who are hav­ing sui­ci­dal thoughts are hard to come by. Also, sit­ting some­one in an FMRI for 30 min­utes is nei­ther prac­ti­cal nor cheap.

He also stressed that any­thing that comes out of this re­search would be strictly an add-on to an as­sess­ment by a ther­a­pist or psy­chol­o­gist. “That’s the gold stan­dard against which we com­pare it.”

As for any fears of a techno-dystopian sce­nario gone awry, Just in­sists that given how much work this di­ag­no­sis takes, “there’s no way this can be done against a per­son’s will…. You can’t point a lit­tle laser beam at them and find out what they’re think­ing.”

THOUGHTS EX­PER­I­MENT Just, top, and Brent found that words that could be as­so­ci­ated with sui­cide trig­gered a dif­fer­ent phys­i­cal re­sponse in the brains of folks who’d had sui­ci­dal thoughts. SUI­CIDE ATTEMPTERS CON­TROL GROUP

Newspapers in English

Newspapers from UK

© PressReader. All rights reserved.