Methods and apparatus for using brain imaging to predict performance

Inventors

Gallacher, Benjamin J. A. • Cook, Douglas J. • Murray, Chris I. • Ross, Andrew N.

Assignees

Voxel AI Inc

Abstract

Methods and apparatus for predicting performance of an individual on a task, the method comprises receiving brain imaging data for the individual, wherein the brain imaging data comprises structural brain data, determining values for at least one characteristic of the structural brain data within regions of interest defined for a population of individuals having different performance levels, and predicting based on the determined values, a performance potential of the individual.

Core Innovation

The invention relates to predicting an individual’s performance using brain imaging data. A computerized system or a computer-implemented method receives brain imaging data for an individual in which the brain imaging data comprises structural brain data. Regions of interest defined for a population of individuals having different performance levels are used to determine first values for at least one characteristic of the structural brain data, and a performance potential of the individual is predicted based, at least in part, on the first values.

In representative embodiments, the structural brain data includes first data characterizing white matter structure and second data characterizing gray matter structure. The white matter structure can include diffusion tensor imaging data such as fractional anisotropy, mean diffusivity, axial diffusivity, and radial diffusivity. Physiological brain imaging data can also be received, where the physiological brain data characterizes dynamics of the individual’s brain, and second values are determined within the regions of interest to predict a current performance level of the individual.

The disclosure further frames the prediction using pipeline concepts that include selecting structural-functional units and extracting structural and physiological measures. Multi-channel voxel maps associate voxels in each region of interest with multiple values corresponding to different brain data types, and voxels can be weighted based on predictive strength. The outputs can include performance deficit and personalized training recommendations based on characteristics derived from predicted performance potential and predicted current performance level.

Claims Coverage

The provided independent claims are clm-00001 and clm-00013. Each covers predicting performance for an individual by determining values within population-defined regions of interest from structural brain data and predicting performance potential based on those values. Additional claim scope is reflected in dependents that introduce physiological/dynamic data, multi-channel voxel maps, and outputs that extend beyond performance potential to current performance level and related training guidance.

Across both independent claims provided, the core inventive coverage is predicting an individual’s performance potential by using structural brain data and computing first values for structural characteristics within regions of interest defined relative to a population with different performance levels. The described claim family further narrows and expands the input modalities and prediction targets in dependent claims, including physiological dynamics for current performance level and multi-channel representations, and can extend outputs to performance deficit and training recommendations.

Stated Advantages

Documented Applications