Methods and apparatus for monitoring consciousness

A method of processing a non-stationary signal including segments having increasing and decreasing amplitude representing physiological characteristics of a sentient subject, said segments including portions in which said signal changes from increasing to decreasing amplitude or vice versa, said method comprising:

• (i) detecting each segment by determining time instants when a time derivative of said signal is substantially equal to zero;
• (ii) performing syntactic analysis for each segment including assigning height, width and error parameters;
• (iii) identifying noise segments present in said signal by comparing said width parameter to a preset threshold and said error parameter to said height parameter;
• (iv) removing said noise segments by replacing each identified noise segment with a substantially straight line;
• (v) sorting the remaining segments into a plurality of wavebands based on their width parameters;
• and (vi) classifying said signal as belonging to one of predefined sleep states based on relative frequency of occurrence of said segments in said wavebands.

A method of monitoring physiological characteristics of a sentient subject comprising:

• applying a first surface electrode to said subject to provide a first electrical signal to a remote monitoring apparatus;
• applying a second surface electrode to said subject to provide a second electrical signal to said remote monitoring apparatus;
• monitoring quality of said first electrical signal;
• in the event of a degradation in said quality of first signal, automatically substituting said second electrical signal for said first electrical signal;
• in the event of a degradation in said quality of said second electrical signal and in said quality of said first electrical signal, providing a warning signal.

An apparatus for processing a non-stationary signal including segments having increasing and decreasing amplitude representing physiological characteristics of a sentient subject, said segments including portions in which said signal changes from increasing to decreasing amplitude or vice versa, said apparatus comprising:

• (i) means for detecting each segment by determining time instants when a time derivative of said signal is substantially equal to zero;
• (ii) means for dividing said signal into said segments including data over three consecutive time instants when said time derivative is equal to zero;
• (iii) means for assigning to each segment, height, width and error parameters;
• (iv) means for identifying noise segments in said signal including means for comparing for each segment said width parameter to a preset threshold and said error parameter to said height parameter;
• (v) means for removing said noise segments including means for substituting a straight line connecting first and third time instants when the time derivative of said signal is substantially equal to zero and means for reassigning segments and their parameters after the substitution;
• (vi) means for sorting the remaining segments into a plurality of wave bands based on the value of their width parameter, each wave band being defined by upper and lower frequencies corresponding to lower and upper values for the width parameter respectively;
• and (vii) means for classifying a time interval of the signal data as belonging to one of predefined sleep states based on relative frequency of occurrence of said segments in said wave bands.

Apparatus for processing a non-stationary signal including segments having increasing and decreasing amplitude representing physiological characteristics of a sentient subject, said segments including portions in which said signal changes from increasing to decreasing amplitude or vice versa, said apparatus including:

• (i) a detector which detects each segment by determining time instants when a time derivative of said signal is substantially equal to zero;
• (ii) a divider which dividing said signal into said segments including data over three consecutive time instants when said time derivative is equal to zero;
• (iii) an assigning component which assigns to each segment, height, width and error parameters;
• (iv) an identifier which identifies noise segments in said signal including a comparing component which compares which each segment said width parameter to a preset threshold and said error parameter to said height parameter;
• (v) a removing component which removes said noise segments including a substituting component which substitutes a straight line connecting first and third time instants when the time derivative of said signal is substantially equal to zero and a reassigning component which reassigning segments and their parameters after the substitution;
• (vi) a sorter which sorts the remaining segments into a plurality of wave bands based on the value of their width parameter, each wave band being defined by upper and lower frequencies corresponding to lower and upper values which the width parameter respectively;
• and
• (vii) a classifier which classifies a time interval of the signal data as belonging to one of predefined sleep states based on relative frequency of occurrence of said segments in said wave bands.