Overview

• Johns Hopkins researchers grew prefrontal cortex–like organoids from patients’ blood and skin cells reprogrammed into iPSCs, including multiple neural cell types and myelin.
• Multi-electrode array recordings produced EEG-like readouts that machine-learning tools used to differentiate disease-specific electrical patterns.
• Disease classification in organoids reached 83% accuracy and rose to 92% after subtle electrical stimulation revealed additional activity features.
• A support vector machine achieved 95.8% accuracy distinguishing schizophrenia from controls in two-dimensional cortical interneuron cultures.
• The APL Bioengineering study used samples from 12 donors, and the team is recruiting more patients to validate findings and to test drug responses on organoids with clinical partners.