Movement disorder recovery system and method
First Claim
1. A system for providing treatment recommendations for movement disorders comprising:
- a movement measuring apparatus for measuring external body motion of a subject with a movement disorder and for calculating a movement score representative of a score that a skilled clinician might give to a subject during a movement analysis exam using a standardized scale, the movement measuring apparatus comprising at least one sensor and a transmitter;
an algorithm trained for correlating at least in part movement scores calculated from the subject with at least in part movement scores stored in at least one central database;
at least one central database where historical movement data comprising at least in part movement data, including at least in part movement scores, from a plurality of other subjects is stored;
a processor for correlating at least in part the calculated movement scores to the at least one central database and determining a recommended treatment using the algorithm; and
a device for outputting or otherwise communicating the recommended treatment to a clinician or other user for review and administering;
wherein the at least one central database contains at least former patient disorder and treatment information; and
the algorithm is trained through a supervised learning process that uses as algorithm inputs at least any three of patient demographics, treatment history, disorder details, recorded movement data, current treatment protocols, and movement scores, and uses as matched outputs recommended treatment stimulation parameters; and
the algorithm performs a stepwise regression to determine a combination of algorithm inputs and matched outputs with a correlation to clinician treatments for subjects, wherein, when calculated by the processor or algorithm, the resultant correlation to clinician treatments for subjects with successful outcomes has a correlation coefficient of greater than 0.25.
1 Assignment
0 Petitions
Accused Products
Abstract
Most particularly, the present invention relates to a customized and adaptive movement recovery system and a method of improving the functional motor recovery of a subject with a movement disorder. The present invention provides for a system and method, which in some embodiments can accurately quantify treatment device parameters and protocols including electrical stimulation amplitude (volts/amps), frequency (Hz), and pulse width (microseconds), and medication titrations, doses, and times by utilizing accelerometric, gyroscopic or other movement related information, such as electromyography (EMG) data, or the like, and a central database, or system of databases, of patient and treatment histories. In other embodiments, the system and method provide for an adaptive central database system and automated control of movement disorder treatment devices.
-
Citations
10 Claims
-
1. A system for providing treatment recommendations for movement disorders comprising:
-
a movement measuring apparatus for measuring external body motion of a subject with a movement disorder and for calculating a movement score representative of a score that a skilled clinician might give to a subject during a movement analysis exam using a standardized scale, the movement measuring apparatus comprising at least one sensor and a transmitter; an algorithm trained for correlating at least in part movement scores calculated from the subject with at least in part movement scores stored in at least one central database; at least one central database where historical movement data comprising at least in part movement data, including at least in part movement scores, from a plurality of other subjects is stored; a processor for correlating at least in part the calculated movement scores to the at least one central database and determining a recommended treatment using the algorithm; and a device for outputting or otherwise communicating the recommended treatment to a clinician or other user for review and administering; wherein the at least one central database contains at least former patient disorder and treatment information; and the algorithm is trained through a supervised learning process that uses as algorithm inputs at least any three of patient demographics, treatment history, disorder details, recorded movement data, current treatment protocols, and movement scores, and uses as matched outputs recommended treatment stimulation parameters; and the algorithm performs a stepwise regression to determine a combination of algorithm inputs and matched outputs with a correlation to clinician treatments for subjects, wherein, when calculated by the processor or algorithm, the resultant correlation to clinician treatments for subjects with successful outcomes has a correlation coefficient of greater than 0.25. - View Dependent Claims (2, 3, 4, 5)
-
-
6. A method for providing treatment recommendations for movement disorders comprising steps of:
-
providing a movement measuring apparatus comprising at least one sensor and a transmitter, the movement measuring apparatus adapted to measure external body motion of a subject; measuring external body motion of a subject wearing the movement measuring apparatus; calculating at least one movement score representative of a score that a skilled clinician might give to a subject during a movement analysis exam using a standardized scale based at least in part on the measured external body motion of the subject; transmitting the calculated at least one movement score from the movement measuring apparatus to a processor; training an algorithm to correlate the at least one movement score with a database of movement data; correlating with the processor the at least one movement score to at least one central database comprising movement data, including movement scores, from a plurality of other subjects using the algorithm, the database further comprising recommended treatment protocols; and determining a recommended treatment based at least in part on the correlation between the at least one movement score and the at least one central database; wherein the training of the algorithm is performed using a backpropagation technique on a feedforward model comprising at least three input layers, at least one hidden layer, and at least one output layer, and wherein the at least three input layers consist at least of; clinical definitions, comprising one or more of current patient demographics, disorder history, treatment history, or clinician notes; subject data, comprising one or more of recorded movement data and diary information; and a dynamic input that accounts for changes in any other input variables with respect to time. - View Dependent Claims (7)
-
-
8. A method for providing treatment recommendations for movement disorders comprising steps of:
-
providing a movement measuring apparatus comprising at least one sensor and a transmitter, the movement measuring apparatus adapted to measure external body motion of a subject; measuring external body motion of a subject wearing the movement measuring apparatus; calculating at least one movement score representative of a score that a skilled clinician might give to a subject during a movement analysis exam using a standardized scale based at least in part on the measured external body motion of the subject; transmitting the calculated at least one movement score from the movement measuring apparatus to a processor; training an algorithm to correlate the at least one movement score with a database of movement data; correlating with the processor the at least one movement score to at least one central database comprising movement data, including movement scores, from a plurality of other subjects using the algorithm, the database further comprising recommended treatment protocols; and determining a recommended treatment based at least in part on the correlation between the at least one movement score and the at least one central database, wherein the algorithm is a neural network algorithm, and the step of training the algorithm comprises using a set of input variables and neural network layer weights to test the neural network'"'"'s ability to generalize data using a “
one left out”
technique where training data consists of all but one subject, a “
left out”
subject, and the training data is used to solve for the neural network weights, and the “
left out”
subject is then input to the neural network to generate predicted treatments. - View Dependent Claims (9)
-
-
10. A method for providing treatment recommendations for movement disorders comprising steps of:
-
providing a movement measuring apparatus comprising at least one sensor and a transmitter, the movement measuring apparatus adapted to measure external body motion of a subject; measuring external body motion of a subject wearing the movement measuring apparatus; calculating at least one movement score representative of a score that a skilled clinician might give to a subject during a movement analysis exam using a standardized scale based at least in part on the measured external body motion of the subject; transmitting the calculated at least one movement score from the movement measuring apparatus to a processor; training an algorithm to correlate the at least one movement score with a database of movement data; correlating with the processor the at least one movement score to at least one central database comprising movement data, including movement scores, from a plurality of other subjects using the algorithm, the database further comprising recommended treatment protocols; and determining a recommended treatment based at least in part on the correlation between the at least one movement score and the at least one central database, wherein the at least one central database contains at least former patient disorder and treatment information; and the algorithm is trained through a supervised learning process that uses as algorithm inputs at least any three of patient demographics, treatment history, disorder details, recorded movement data, current treatment protocols, and movement scores, and uses as matched outputs recommended treatment stimulation parameters; and the combination of algorithm inputs and matched outputs is determined using a stepwise regression wherein the resultant correlation to clinician treatments for subjects with successful outcomes has a correlation coefficient of greater than 0.25.
-
Specification