Sensitivity Threshold refers to the minimum signal or stimulus intensity that can be detected by a certain measuring system, instrument or sense organ. It is widely used in medical detection, acoustics, optics, electronic engineering, psychology and many other fields. Accurate determination of sensitivity threshold is of great significance to ensure the reliability of the measurement system and improve the detection accuracy. This paper will briefly introduce several common sensitivity threshold determination methods and their applications.
I. Definition and Significance
Sensitivity threshold usually refers to the lowest signal level that the system can detect stably without false alarm. In different application scenarios, this concept may have different expressions. For example, in the listening test, it is the lightest sound that the human ear can hear; In medical detection, it is the lowest concentration that can be reliably identified by detection methods.
Second, the determination method
1. Signal Detection Theory, SDT).
This method is widely used in psychology and perception research. By letting the subjects judge whether they perceive a specific stimulus, the threshold is determined by combining “Hit Rate” and “False Alarm Rate”. The advantage of SDT is that it can separate the perception ability from the decision criteria, so as to evaluate the sensitivity more accurately.
2. Method of Limits.
This is a classical threshold measurement method, which is often used in perception experiments. In the test, the intensity of stimulation gradually increases from low to high (ascending method) until the subject perceives the stimulation for the first time; Or decreasing from high to low (descending method) until the subject can no longer perceive it. After repeated for many times, the average value can be obtained.
3. The method of constant simulation.
Different from the rising and falling method, this method randomly presents different intensity stimuli, which avoids the expected error. Subjects only need to respond “perceived” or “not perceived” to each stimulus. Through statistical analysis, the perceptual probability curve is drawn, and then the stimulus intensity at 50% recognition rate is determined as the threshold.
4. Standard curve method in clinical and engineering.
In laboratory detection or engineering measurement, the standard curve method is often used to determine the sensitivity threshold of the detection system. For example, in the enzyme-linked immunosorbent assay (ELISA), by measuring the signal responses of samples with different concentrations, the concentration-response curve is drawn, and the lowest concentration of the signal significantly higher than the background noise is determined as the detection limit.
Third, the influencing factors
-Background noise level: High noise environment will raise the threshold and reduce the sensitivity.
-Individual differences: In the perception experiment, people’s physiological and psychological states will affect the threshold determination.
-Instrument accuracy and calibration status: the resolution and stability of the equipment directly affect the measurement results.
Fourth, application examples
Taking medical diagnosis as an example, the sensitivity threshold of PCR (polymerase chain reaction) technology determines whether it can detect very low concentration of viral RNA, which is very important for early diagnosis and epidemic prevention and control. For another example, in the acoustic test, the sensitivity threshold of the hearing aid determines whether it can effectively help the hearing impaired perceive bass.
V. Conclusion
The determination of sensitivity threshold is not only a technical work, but also a manifestation of scientific rigor. Different fields need to choose appropriate methods according to their own characteristics, and at the same time pay attention to controlling interference factors in order to obtain reliable and repeatable determination results. With the development of science and technology, the future determination methods will be more intelligent and automatic, which will provide higher precision guarantee for various detection systems.