Schizophrenia is a complex mental disorder that can negatively affect social behaviour and the capacity to distinguish fact from fiction.
To determine whether a person has schizophrenia, a mental health professional gathers evidence from self-reported symptoms, behavioural observations, and interviews with their family and friends. Because this approach is so subjective, it opens the door to misdiagnosis, especially in people with different cultural backgrounds or religious beliefs.
The search for a more objective way to diagnose schizophrenia has been going on for decades. It has focused on finding a “biomarker” for the disease, which is a biological characteristic that can be independently measured and quantified.
Dr Marta Garrido and colleagues at the University of Queensland, the University of Newcastle, and University College London found such a biomarker with the potential to improve diagnosis.
The team recorded brain responses to irregular sounds, which your brain detects even when you’re distracted, using electroencephalography (EEG). The EEG recordings were made in a group of people who had been diagnosed with schizophrenia and a control group of people without the disorder.
The team then trained a computer to classify these brain responses as coming from either the group with schizophrenia or the control group.
When presented with new EEG recordings that it had not already received, the computer was able to correctly identify – with up to 80 per cent accuracy – whether they belonged to a person in the schizophrenia group or to someone in the control group.
The use of biomarkers such as the one found in this study could help to reduce the rate of misdiagnosis in schizophrenia, which is currently estimated to be around 10 per cent.
The team will study the use of this technique in people at risk of developing a mental disorder, to try and predict how likely they are to develop schizophrenia