Chapter8-HOMI-Biosignal Analysis

Biosignal analysis- Purpose is to derive information from the data (the signals)

Signals as a function of time  can be Stationary or NOn-stationary and waveshapes can be deterministic or statistical (stochatic)

Biological processes that are repetitive (i.e.beating heart, breathing, blinking) are called deterministic and these can be periodic, quasi periodic, or transient Periodic -- not seen really in humans Quasi-periodic -  blinking Transient - depolarizaiton/poliarization of a cell - single waveform  Point processes - impulses where it is primarily 0 but ocassional stimulus produces  a "1"  - These don't occur in biological processes - although a pacemaker is closest to one 

Stochastic Waveforms - occur when signals are generated by groups of cells in a random fashion - Wafeforms are non-determinstic and require statistical descriptions

ALL BIOLOGICAL SIGNALS ARE ANALOG!!!! - No exception 

Analog to Digital- must occur for computers to interpret biologic singals if done correctly, not information is lost analog can be restored by digital to analog conversion later Sampling yields discrete signals from analog signals 

Sampling Theorem- Nyquist and SHannon - a signal must be sampled at 2x rate of the rate of the highest frequency componentIf we sample at rate that is too low, the signal is distortedIf obey theorem, complete Syntatic information content of signal is retainedIn sampling , the amplitudes of the analog signal are measured at equidistant intervals and converted to discrete values that are expressed as binary numbers = Quantization or quantification. Sampling and quantization occurs in an Analog to Digital converterSampling frequency - FsSampling interval Delta T = 1/Fs = the more changes per second  there are in a signal, the higher the frequency content of the signal , the higher the sampling frequency must be 

Four different categories of signal analysis: Output signal only - This is most common situation = we have no or minor knowledge of the process generating the signals- EEG Evoked signals - we offer an input signal or stimullus - ideally we are then able to gain insight into the state of the biological process -i.e. evoked responses in EEG Provocative Test - Test a process under a forced or at least known condition - i.e. stress testing Modeling - sufficient knowledge exist to develop a model of the biological process

New Page