Assembly, method and computer program product for influencing a biological process

Abstract

The invention provides an assembly comprising: a recording assembly for recording a time-based brain-related signal; a stimulus generator for providing a stimulus, and a computer assembly, functionally coupled to said recording assembly and to said stimulus generator, said computer assembly comprising: a memory for storing at least a data segment of said time-based brain-related signal during recording of said time-based brain-related signal, and a computer program which, when running on said computer assembly, functionally real-time performs: retrieving a most-recent data segment of said stored data segment of said time-based brain-related signal; fitting at least one curve to said retrieved most-recent data segment; predicting a future continuation of said most-recent data segment using said at least one curve fitted to said most-recent data segment; detecting a predefined pattern in said predicted future continuation for predicting occurrence of said predefined pattern, and defining a predicted event time of said predefined pattern, said predicted event time being in the future with respect to said most-recent data segment, and actuating said stimulus generator for providing a stimulus within a predefined event time window of said predicted event time.

1 . An assembly comprising:
 a recording assembly for recording a time-based brain-related signal;   a stimulus generator for providing a stimulus, and   a computer assembly, functionally coupled to said recording assembly and to said stimulus generator, said computer assembly comprising:   a memory for storing at least a data segment of said time-based brain-related signal during recording of said time-based brain-related signal, and   a computer program which, when running on said computer assembly, functionally real-time performs:   retrieving a most-recent data segment of said stored data segment of said time-based brain-related signal;   fitting at least one curve to said retrieved most-recent data segment;   predicting a future continuation of said most-recent data segment using said at least one curve fitted to said most-recent data segment;   detecting a predefined pattern in said predicted future continuation for predicting occurrence of said predefined pattern, and defining a predicted event time of said predefined pattern, said predicted event time being in the future with respect to said most-recent data segment, and   actuating said stimulus generator for providing a stimulus within a predefined event time window of said predicted event time.   
     
     
         2 . The assembly of  claim 1 , wherein said prediction extending at least 0.5 seconds beyond the end time of said data segment, in particular at least 0.5 seconds past a current time of the assembly. 
     
     
         3 . The assembly of  claim 1 , wherein said stimulus generator is arranged for applying a sensory discernible stimulus, in particular a sound or light stimulus that can be perceived by a human. 
     
     
         4 . The assembly of  claim 1 , wherein said computer program is adapted for retrieving a most-recent data segment that has an end time that is less than 0.5 seconds from an assembly current time, in particular less than 0.1 second, more in particular less than 1 millisecond. 
     
     
         5 . The assembly of  claim 1 , wherein said event time is in the future with respect to a current time of said assembly, allowing said assembly to apply said stimulus within said event time window. 
     
     
         6 . The assembly of  claim 1 , wherein said curve fitting comprises applying a non-linear regression algorithm. 
     
     
         7 . The assembly of  claim 1 , wherein said computer program retrieves said most-recent data segment within a processing time from its recording, in particular said processing time is less than 0.5 seconds, more in particular less than 0.1 seconds, more in particular less than 1 millisecond. 
     
     
         8 . The assembly of  claim 1 , wherein said at least one curve comprises a series of periodic functions. 
     
     
         9 . The assembly of  claim 1 , wherein said at least one curve comprises a series of trigonometric functions, in particular sinus functions. 
     
     
         10 . The assembly of  claim 1 , wherein said computer program further evaluates a reliability of said fit, and if said reliability is within a predefined criterion, calculates said future continuation. 
     
     
         11 . The assembly of  claim 1 , wherein said computer program retrieves a most-recent data segment with a data window of which has a width of less then 1 second, in particular less then 0.5 seconds. 
     
     
         12 . The assembly of  claim 1 , wherein said computer program fits at least one periodic function to said most-recent data segment, said at least one periodic function having a period shorter then 2 seconds, in particular shorter then 0.2 seconds, more in particular shorter than 0.1 seconds. 
     
     
         13 . The  claim 1 , wherein said brain-related signal comprises an electromagnetic signal, in particular an electro encephalogram (EEG). 
     
     
         14 . The assembly of  claim 13 , wherein said EEG has a time resolution of at least 100 samples per second, more in particular at least 500 samples per second. 
     
     
         15 . The assembly of  claim 1 , wherein said event is a predefined oscillatory phase in said brain-related signal, more in particular selected from a rising and a falling slope in said brain-related signal. 
     
     
         16 . A computer program product, wherein said computer program product, when running on a computers system, performs a method comprising:
 retrieving a time-based brain-related signal;   performing real-time fitting of at least one curve to a most-recent data segment of said brain-related signal;   predicting a future continuation of the recorded signal using said curve fitted to said data segment;   detecting a predefined pattern in said predicted future continuation and defining a predicted event time of said predefined pattern, and   generating a stimulus within an event window of said predicted event time if said predefined pattern is detected in said future continuation.