Methods and systems are described for monitoring patient speech to determine compliance of the patient with a prescribed regimen for treating for a brain-related disorder. Patient speech is detected with an audio sensor at the patient location, and speech data is transmitted to a monitoring location. The audio sensor and other components at the patient location may be incorporated into, or associated with, a cell phone, computing system, or stand-alone microprocessor-based device, for example. Patient speech is processed at the patient location and/or monitoring location to identify speech parameters and/or patterns that indicate whether the patient has complied with the prescribed treatment regimen. Patient identity may be determined through biometric identification or other authentication techniques. The system may provide a report to an interested party, for example a medical care provider, based on whether (and/or the extent to which) the patient has complied with the prescribed treatment regimen.