Accurately Diagnosing Prosopagnosia

New testing methods for diagnosing face blindness.

Posted Dec 01, 2010

The first three doctors I saw here in Michigan said there was simply no way could I have prosopagnosia. "Too rare." And, I passed all of their basic screening tests for prosopagnosia. I argued that the tests didn't really measure my deficit, but my complaints fell on deaf ears. "The Benton works," one doctor told me. "It's what we use." Frustrated with how little was known about the disorder, and out of patience with my therapist, primary care doctor, and the local neurologists and their dusty tests, I wondered if there were sophisticated research studies I could partake in. I started researching rare diseases and eventually found the NINDS, the National Institute of Neurological Disorder and Stroke. I wrote them an old-fashioned letter.

After many months, the NINDS wrote me back. Prosopagnosia research was being done at Harvard University in a lab run by neuropsychologist Ken Nakayama. As it happened, I had an upcoming business trip to Boston, and I arranged to stay on for several days, at my own expense, to participate in the research being conducted by two post docs, Galit Yovel (of MIT) and Brad Duchaine at Harvard.

When I arrived, The first thing Brad asked me was, "Were you given the Benton Test and told you were fine?"

Yes! I said. And for the first time in my life, I felt there was some hope. I felt believed, and finally, seen. Eagerly, I went through the battery of tests Brad had compiled. In his tests, all conducted on computer, the images of faces were just faces: the hair, ears, and necks-what I used for identification-were elided. When it came to just the face, I failed. Miserably. I was at Harvard University, failing every test and I was the happiest woman in the world.

I underwent extensive testing in via functional MRI (fMRI). And I failed, spectacularly. I got the proof I needed: my brain doesn't remember the features of the human face. Ever. At all. And there's nothing I can do about it.

Why was it so hard to get diagnosed?

Well, the research is very new-we know now that 2% of the population is affected by face recognition deficits, to some degree. But that information is not widely known. And, developmental prosopagnosia is a particular syndrome, only recently delineated; it requires precise testing for the actual problem: the inability to remember the features of the human face. Yet many clinicians haven't knowingly encountered developmental or congenital cases; their experience is limited to the rare acquired case-stroke victims or epileptics who suffer brain damage and dramatically lose their ability to read faces. (It's much harder to present your problem if you've had prosopagnosia your whole life, as I have-I simply didn't know faces could be recognized in any special way.)

However, there are accurate, simple diagnostic tools available. For all types of prosopagnosia, the best single measure for prosopagnosia is the Cambridge Face Memory Test. The Cambridge, unlike the Benton, has good reliability and effectively discriminates between people with and without prosopagnosia. But prosopagnosia is a complex disorder; face recognition is a superbly complex process in the brain. No single test provides a "gold standard." So, one can suffer from prosopagnosia and still pass the Cambridge.

"Face recognition in every day life," Brad Duchaine explains, "is much more complicated than face recognition in the simple tests devised so far, plus prosopagnosia is the name for a lot of different disorders. Imagine for example that a subject can hold newly learned faces for 15 minutes but has trouble retrieving them the next day or the next month. That person would do fine on the Cambridge Face Memory Test. In every day life, one isn't able to scrutinize faces like you can in a test. All tests designed so far present still faces but faces in the real world move. That said, I don't want to be negative about the CFMT-it works well for most people, just not everyone." Like any single test for prosopagnosia, the Cambridge test is not perfect-it's merely a good indicator of prosopagnosia. Some people will score normally on the CFMT and report stories that sound very typical of prosopagnosia. Therefore, it's best to diagnose prosopagnosia using multiple measures and the patient's reports from daily life.

Here is a link to an online version of the Cambridge Face Memory Test (it's not the original test, but it's the same paradigm with different faces):

Why isn't the Cambridge test widely used? This test is relatively new and it's challenging for the researchers who develop new instruments to find pathways to inform clinicians about tests and disorders. Some clinicians are familiar with the Cambridge Face Memory Test and the limited number of other good tests available but the disorder is not listed in the DSM and won't be in the new version. Good news: the APA has a new test database website [].

The most important part of diagnosis is, ironically, recognition: we need to recognize the disorder for what it is. And to keep in mind that the testing for complicated disorders and syndromes is frought, always, with substantial limitations. Diagnosis is hardly ever an end point. It's often simply one way of seeing something complex a little more clearly. Diagnosis isn't something a doctor on high passes down to the patient. Diagnosis is an ongoing conversation, sometimes messy, wandering, frustrating, and layered.


Next: The Gifts of Prosopagnosia.


More Posts