Clinical decision support system
Example apparatus and methods concern a next generation clinical decision support system (ngCDSS) for the management of neurological conditions (e.g., advanced Parkinson's disease (PD)). Conventional coupled adjustment of pharmacologic therapy and stimulation parameter settings is a time-consuming process that sometimes yields sub-optimal outcomes. Example ngCDSS use a machine learning trained function that relates deep brain stimulation (DBS) parameters, medication dosages, and patient-specific pre and post operative clinical data with actual treatment outcomes for a population of previously treated patients. Example ngCDSS incorporate image-based patient-specific computer models of the estimated stimulation volume of tissue stimulated by DBS in a multi-linear regression analysis to produce a predictor function that is highly correlated with actual outcomes. Example ngCDSS facilitate predicting the outcomes of a combined pharmacologic-DBS therapy, which in turn facilitate optimizing patient-specific treatment for improved benefits with minimal adverse effects.
1. An apparatus for selecting treatment parameters for a patient, comprising:
a first circuit that produces first electronic data that characterizes a neuroanatomical condition of the patient, where the characterization of the neuroanatomical condition is based, at least in part, on an electrode implanted in a brain of the patient;
a second circuit that produces second electronic data that characterizes the patient based on patient symptom data and patient non-symptom data;
a processor that computes a similarity metric for the patient based, at least in part, on the first electronic data and the second electronic data;
a third circuit that identifies relevant data associated with a set of other patients and their therapeutic outcomes based on the similarity metric;
a fourth circuit that produces third electronic data identifying a combination of treatment parameters for the patient based, at least in part, on the relevant data, where the combination of treatment parameters includes one or more stimulation parameters and one or more medication parameters; and
a fifth circuit that provides a visualization, to a display, of the relevant data from which the fourth circuit selects the combination of treatment parameters.
2. The apparatus of claim 1 , where the first circuit characterizes the neuroanatomical condition of the patient based on a magnetic resonance image.
3. The apparatus of claim 2 , where the first circuit determines an overlap between an estimated stimulation volume (ESV) in the brain of the patient and a target stimulation area (TSA) in the brain of the patient.
4. The apparatus of claim 3 , where the processor computes the similarity metric based on the patient symptom data, the non-symptom data, and the overlap between the ESV and the TSA.
5. The apparatus of claim 1 , where the third circuit identifies the relevant data based on a linear weighted sum function applied to data associated with the set of other patients.
6. The apparatus of claim 5 , where the linear weighted sum function is the product of machine learning associated with multi-linear regression analyses that identify correlations in data associated with the set of other patients and their therapeutic outcomes.
7. The apparatus of claim 6 , where the machine learning includes naïve Bayesian (NB) learning, random forest (RF) of trees learning, and support vector machine SVM learning.
8. The apparatus of claim 7 , where the linear weighted sum function produces an aggregate score from separate scores for different elements of the patient symptom data and the patient non-symptom data.
9. The apparatus of claim 8 , where the separate scores for different elements of the patient symptom data and the patient non-symptom data are selected from results produced by different machine learning approaches.