New Study: Eye-Movement Test Accurately Indicates Schizophrenia

Health Care, Psychology, Science


recent study proposes that we may be able to use current technology to identify schizophrenics without spending the copious resources for a qualified neuropsychologist to diagnose an individual case.  What implications can this have for pilot licensing, holding government office, police recruiting, and generally the overall stigma associated with individuals who are functioning and non-functioning clinical schizophrenics?

A group of scientists from Scotland, Germany, and the USA recruited schizophrenic patients from mental hospitals in Munich, Germany and Aberdeen, Scotland. The researchers confirmed schizophrenia by diagnostic procedures in the DSM-IV as well as case history.  Control group participants were recruited from the area surrounding University of Aberdeen, excluding people with a history of alcohol abuse/dependence, major head trauma involving loss of consciousness for more than 5 minutes, epilepsy or other neurological dysfunction, and first-degree family history of psychosis.

Using infrared eye-tracking technology via the EyeLink I and a 19” video screen, the study tested visual patterns in smooth pursuit of a moving object for 20 seconds, fixation stability on the same stationary object, and free-viewing of photographs including:

“Luminance-balanced natural and man made environments showing information at different spatial scales; everyday objects and food in sparse and cluttered scenes; expressive, neutral, and occluded faces; animals; and unfamiliar computer-generated images (fractal patterns, gray-scale ‘pink noise.’”

The conclusion brought by the research is that schizophrenic individuals clearly lack an ability to perform visual tests the same as control individuals.  Diagnosed schizophrenics cannot accurately pursue an object with a smooth speed and path or concentrate with normal patterns when steadily gazing at the photographs presented.

Not every individual exhibits “normal” or “abnormal” eye movements on every trial, so the study analysis combined multiple performance measures from multiple tests.  A small covariation in fixation stability was caused by advanced age in the test subjects, but aside from that no other consistent covariations were presented by the data. Pharmaceutical regimens had no bearing on the concluding classification.  There was no relationship between probability of classification as “schizophrenic” and the number of chlorpromazine equivalents prescribed, age of onset/duration of illness, or presence or absence of cigarette smoking.

98% of the participants were correctly classified; one schizophrenic patient was misclassified as control and five control subjects were misclassified as borderline.  After a retest at 9 months, the reliability of diagnosis was no different (test (SD = .843) and retest (SD = .804)).  The study purports, “This is a remarkable level of discrimination and well beyond that of other potential trait markers previously reported in schizophrenia.”


Leave a Comment