Hitherto, not only the instruments and apparatuses supplied from engineering field, but theories for engineering also must be brought into the medical field. Among all, theories for signal detecting problem is perhaps some of the most useful tools to be applied for medical purpose.
These theories have been developed mainly in the communication engineering field and have made great strides recently by the aid of the information theory.
The methods of time series analysis, for example, the correlation function method and the periodgram method are useful for detecting any periodic components hidden by noises. The matched filter can be used when the shape of signal pulse is known.
Sometimes, the average response computer searches very delicate reactions induced any periodic stimulus.
The concept of pattern recognition has important connection with the signal detection, because any signals, generally speaking, are recognized as patterns.
Quick detection makes use of the process of the adaptive control. A direction finder capable of learning is shown as an example.
The rational criterions for judging whether signal is present or not are given by the decision theories, for example, by the stochastic decision theory.
Signals must be displayed from the human engineering point of view. An example of display hereof is presented.
Finally, the importance of the operations research is emphasized for any proiect of signal detection.
View full abstract