US 12,220,581 B2
User-weighted closed loop adjustment of neuromodulation treatment
Matthew Lee McDonald, Glendale, CA (US)
Assigned to Boston Scientific Neuromodulation Corporation, Valencia, CA (US)
Filed by Boston Scientific Neuromodulation Corporation, Valencia, CA (US)
Filed on Oct. 25, 2023, as Appl. No. 18/383,792.
Application 18/383,792 is a continuation of application No. 17/895,835, filed on Aug. 25, 2022, granted, now 11,833,355.
Application 17/895,835 is a continuation of application No. 16/778,367, filed on Jan. 31, 2020, granted, now 11,471,684.
Claims priority of provisional application 62/813,262, filed on Mar. 4, 2019.
Prior Publication US 2024/0050752 A1, Feb. 15, 2024
This patent is subject to a terminal disclaimer.
Int. Cl. A61N 1/36 (2006.01); G06N 3/02 (2006.01); G16H 40/67 (2018.01)
CPC A61N 1/36132 (2013.01) [A61N 1/36071 (2013.01); A61N 1/36078 (2013.01); A61N 1/36096 (2013.01); A61N 1/36175 (2013.01); G06N 3/02 (2013.01); G16H 40/67 (2018.01)] 20 Claims
OG exemplary drawing
 
1. A non-transitory machine-readable storage medium comprising instructions for programming for a neurostimulation device, wherein the instructions, when executed by circuitry of a programming system, cause the programming system to:
identify multiple therapy objectives and corresponding multiple rating values from user input data, the multiple therapy objectives relating to respective objectives of neurostimulation treatment to be performed from the neurostimulation device;
use a trained model to identify composite output values for the neurostimulation treatment, the trained model configured to receive the multiple therapy objectives and the corresponding multiple rating values as input, and the trained model configured to provide the composite output values as an output, wherein the trained model identifies the composite output values based on applying the multiple rating values as weights, and wherein the weights are applied in the trained model to modify output values identified by processing layers of the trained model;
generate programming parameters for the neurostimulation device, based on the composite output values identified with the trained model, wherein the programming parameters configure the neurostimulation device to perform the neurostimulation treatment according to the multiple therapy objectives; and
cause the neurostimulation device to implement the programming parameters in at least one program of the neurostimulation device, wherein in response to implementing the programming parameters, the neurostimulation device is configured to output electrical neurostimulation to a patient using the at least one program.