Schizophrenia affects 2.4 million Americans and causes significant individual and societal costs. Cognitive deficits including poor working memory arise early in the course of illness, account for poor long-term outcomes and have been difficult to treat with available treatments. The investigators are proposing to develop a novel, computer-based brain training to improve working memory in schizophrenia patients, which, if successful could have significant personal, societal, and economic impact.

Schizophrenia (SCZ) is a chronic debilitating mental disorder that affects 2.4 million Americans and leads to considerable individual and societal costs. In patients with SCZ, cognitive deficits (CD) occur early in the course of the illness, are associated with more severe illness, and are the best predictor of functional outcomes. Nonetheless, to date, CD have been difficult to treat using available treatments. Recent studies suggest CD in patients with SCZ may arise from abnormal synchronization of distributed neural networks. Synchronization or synchronous firing of neurons, binds cortical areas into functional networks in a task and state-dependent manner. Thus novel therapies that improve abnormal neural synchrony may improve previously refractory symptoms arising from disordered brain networks.

Neural synchrony or coherence in the gamma band (GBR, 30-45Hz) plays a central role in top-down attention, multisensory processing, perceptual binding and working memory (WM). Patients with SCZ exhibit abnormal GBR, and the magnitude of impairment is associated with the severity of cognitive disorganization. Given these results, improving GBR should improve CD, including WM in SCZ. This hypothesis has been tested and confirmed using repetitive transcranial magnetic stimulation (rTMS). Additionally, EEG-based neurofeedback (NFB) is hypothesized to improve GBR and cognitive function in patients with SCZ. NFB is a low-cost, easily administered and well-tolerated treatment. In healthy controls, Gamma-NFB improves GBR and cognitive function including WM. Thus, the investigators propose testing the feasibility and effectiveness of improving GBR using gamma-NFB in patients with SCZ using the framework of the R61/R33 mechanism.

The first trial (R61) is a proof-of-concept study designed to assess target engagement and dose response curve. Thirty SCZ patients will receive G-NFB training for 12 weeks (2 weekly sessions of 30minute duration) and be assessed for 1) evidence of training, 2) Change in GBR, 3) Change in WM and 4) Change in community functioning. The second trial (R33) aims to confirm target engagement based on training parameters obtained from R61, and to assess whether G-NFB is superior to an active-placebo neurofeedback intervention in improving GBR, WM and community functioning.

Inclusion Criteria:

• Diagnosis of schizophrenia/schizoaffective disorder

Exclusion Criteria:

• Active substance use disorders in the last 30 days
• Medication changes in the last 30 days
• History of traumatic brain injury
• seizure disorder
• pregnancy