System and method for brain modelling

Abstract

Brain modelling includes receiving time-coded bio-signal data associated with a user; receiving time-coded stimulus event data; projecting the time-coded bio-signal data into a lower dimensioned feature space; extracting features from the lower dimensioned feature space that correspond to time codes of the time-coded stimulus event data to identify a brain response; generating a training data set for the brain response using the features; training a brain model using the training set, the brain model unique to the user; generating a brain state prediction for the user output from the trained brain model, and automatically computing similarity metrics of the brain model as compared to other user data; and inputting the brain state prediction to a feedback model to determine a feedback stimulus for the user, wherein the feedback model is associated with a target brain state.

What is claimed is: 
     
         1 . A computer-implemented method for generating a treatment protocol, the method comprising:
 receiving time-coded bio-signal data associated with a user from a bio-signal sensor;   projecting the time-coded bio-signal data into a lower dimensioned feature space;   extracting features from the lower dimensioned feature space to identify a brain response;   generating a training data set for the brain response using the features;   generating a brain model by transfer learning from one or more similar users by: 
 comparing attributes of users to identify the one or more similar users; 
 pooling data of the one or more similar users stored in a memory; and 
 training the brain model from a base brain model using the pooled data using a processor that modifies parameters of the base brain model; and 
   training the brain model using the training set using a processor that modifies the parameters of the brain model stored on the memory, the trained brain model unique to the user;   generating a treatment protocol for the user based in part on the trained brain model, using a processor that accesses the trained brain model stored in the memory.   
     
     
         2 . The computer-implemented method of  claim 1 , further comprising:
 receiving time-coded stimulus event data; and   wherein the features correspond to time codes of the time-coded stimulus event data.   
     
     
         3 . The computer-implemented method of  claim 1 , wherein the treatment protocol is generated by:
 generating a brain state prediction for the user output from the trained brain model, using a processor that accesses the trained brain model stored in the memory; and   inputting the brain state prediction to a feedback model to determine the treatment protocol for the user, wherein the feedback model is associated with a target brain state; and   wherein the method further comprises causing the treatment protocol to be provided to the user using a user effector.   
     
     
         4 . The computer-implemented method of  claim 1 , wherein the treatment protocol comprises at least one of administering an intervention, sensory inputs, dosing inputs, duration of therapeutic setting, frequency and intensity of interaction, set, setting, and intensification of experience. 
     
     
         5 . The computer-implemented method of  claim 1 , further comprising:
 receiving additional time-coded bio-signal data associated with the user after an interval of time;   updating the trained brain model based on the additional time-coded bio-signal data; and   updating the treatment protocol based on the updated trained brain model.   
     
     
         6 . The computer-implemented method of  claim 5 , wherein the updated treatment protocol comprises continuation, alteration, deviation, cessation, adjustment, titration, modulation, or variation of a given treatment. 
     
     
         7 . The computer-implemented method of  claim 5 , wherein the additional time-coded bio-signal data is received in real-time or over a longitudinal progression of time. 
     
     
         8 . The computer-implemented method of  claim 1 , wherein the one or more similar users are selected based on at least one of time of day, food and chemical consumptions, medical conditions, emotional states, sleepiness, age, gender, and geographical location. 
     
     
         9 . The computer-implemented method of  claim 1 , wherein the treatment protocol is configured to account for at least one of default mode network and functional connectivity changes. 
     
     
         10 . The computer-implemented method of  claim 9 , wherein the changes arise from sedatives or antidepressants. 
     
     
         11 . The computer-implemented method of  claim 1 , further comprising determining a response to a therapeutic or medical parameter based on the trained brain model. 
     
     
         12 . The method of  claim 1 , further comprising comparing attributes of users to determine covariance, and based on the covariance, identifying the one or more similar users. 
     
     
         13 . The method of  claim 1 , wherein the time-coded bio-signal data includes time-coded EEG data generated by repeated measures. 
     
     
         14 . The method of  claim 1 , wherein the training of the brain model includes at least one of supervised, semi-supervised, unsupervised learning, or reinforcement learning. 
     
     
         15 . The method of  claim 1 , wherein the brain model is at least one of a logistic regression, linear discriminant analysis, a random forest, gradient boosted trees, support vector machines, or ensemble learning. 
     
     
         16 . The method of  claim 1 , wherein the brain model comprises one or more models selected from the group of linear model, logistic regression, linear discriminant, Gaussian naive Bayes classifier, linear support vector machines, nonlinear classifiers, nonlinear support vector machines, random forests, gradient boosted trees, k-nearest neighbours classifier, neural network, fully connected neural network, convolutional neural network, recurrent neural network, long short term memory neural network, residual neural network, autoencoder, restricted Boltzmann machines, generative adversarial network, capsule network, histogram, standard parametrized distribution, multivariate Gaussian, Wishart, expert system, including combinations thereof. 
     
     
         17 . The method of  claim 1 , wherein the brain model comprises a neural network, wherein the method further comprises training the brain model by training parameters of the neural network. 
     
     
         18 . The method of  claim 16 , wherein the neural network is at least one of or a combination of a convolutional neural network, recurrent neural network or long short term memory network. 
     
     
         19 . A system for generating a treatment protocol, the system comprising:
 a client computing device;   a bio-signal sensor, wherein the bio-signal sensor is in communication with the client computing device;   a server comprising a memory storing a plurality of user brain models, wherein the server is in communication with the client computing device;   the client computing device configured to: 
 generate time-coded bio-signal data associated with a user using the bio-signal sensor; 
 transmit the time-coded bio-signal data to the server; 
   the server configured to: 
 receive the time-coded bio-signal data; 
 project the time-coded bio-signal data into a lower dimensioned feature space; 
 extract features from the lower dimensioned feature space to identify a brain response; 
 generate a training data set for the brain response using the features; 
 generate a brain model by transfer learning from one or more similar users by: 
 comparing attributes of users to identify the one or more similar users; 
 pooling data of the one or more similar users stored in the memory; and 
 training the brain model from a base brain model using the pooled data by modifying the parameters of the base brain model; and 
 
 train the brain model using the training set by modifying the parameters of the brain model, the trained brain model unique to the user; and 
   at least one of the client computing device and the server configured to: 
 generate a treatment protocol for the user based in part on the trained brain model. 
   
     
     
         20 . A non-transitory computer readable medium comprising a computer readable memory storing computer executable instructions thereon that when executed by a computer cause the computer to perform the method of  claim 1 .