| 摘要 |
A computer implemented neural stimulation system 400, particularly for treating movement disorders, has: a first module 408 to derive neural data from measurements of movements, muscle contractions or speech; a second module 406,412 to derive a neural state assessment of a patient from the neural data, e.g. to predict onset of an episode of tremors or OCD; a third module 414 for deriving at least one neural stimulation parameter based on the neural state assessment; and a fourth module 416 for delivering a therapy such as deep brain stimulation. Preferably detected speech is converted to text and analysed using a graphical text analyser (702, figure 7) to provide word graphs (712) which are input to a learning engine (714) and a predictive engine (716) which uses Markovian modelling to derive the neural state assessment. Preferably the first module also uses other inputs 410 such as EEG measurements, heart rate variability and real world behaviours which may be derived from a cloud computing environment 50. |
