Project Chappie

BEL-based Emotional AI

Project Chappie is the name of our latest project in BELresearch. The goal of this project is to bring the power of our intelligent algorithms into diagnosis of mental illness, specifically schizophrenia (and in phase II major depressive disorder, as well as diagnosis of degenerative brain disorders, such as alzheimer’s disease (AD) and Parkinson Disease (PD)).


Schizophrenia is everywhere: You may have watched a few movies about schizophrenia, such as A Beautiful Mind, The Soloist, The Voices, Shine, Call Me Crazy, etc. It might affect everyone, even super-intelligent personas.

According to the National Institute of Mental Health, schizophrenia is a severe and chronic mental disorder that negatively alters the thoughts, feelings and behaviors of the patient. It’s estimated that 1 percent of the world’s population suffer from schizophrenia. Majority of schizophrenics may lose touch with reality. The schizophrenia’s causes are still unknown. The symptoms include delusions, hallucinations, disorganized thoughts, speech and behavior, and disability to function normally. Currently, the treatments only focus on alleviating the symptoms of the disease.

If a schizophrenic patient left untreated, it can result in severe complications in every aspect of life, that are very severe and can be life-threatening. These complications include:

  • Suicidal thoughts and behavior, such as self-injury
  • Anxiety disorders and OCD (obsessive-compulsive disorder)
  • Depression
  • Drug and alcohol abuse
  • Social isolation
  • Homelessness
  • Inability to work

Therefore, an early diagnosis and treatment is extremely important.


Electroencephalography or EEG has gained significant attention in the last few years to quantitatively study brain activity. In many mental conditions, EEGs show significant alterations compared to normal EEGs. For example schizophrenia is associated with an increase in delta waves and a decrease in the signal power of the N100/P200 activity. Interestingly, the EEGs of schizophrenics are distinguishable from those patients with other cognitive and psychological damages, such as affective disorders and personality disorders.

Abnormal EEG complexity in patients with schizophrenia and depression

We use electroencephalography (EEG) data from schizophrenic and healthy people to train our AI models. Our BEL-based models have high sensitivity and accuracy in diagnosis of schizophrenia using EEG data.