Synthetic simulation system for testing of physiological monitoring devices
System and method for validating algorithms used in medical devices is disclosed. As disclosed, the system and method include generating a synthetic waveform, inputting the generated synthetic waveform to the testing device, capturing an output values for the input values generating synthetic waveform from the testing device; and comparing the output values to the input values to determine performance metrics of the testing device to validate the algorithm on the testing device.
1. A method for validating an algorithm on a testing device, the method comprising:
generating at least one synthetic waveform from at least two input values;
inputting the generated at least one synthetic waveform to the testing device;
capturing output values from the testing device; and
comparing the output values to the at least two input values generating the at least one synthetic waveform to determine performance metrics of the testing device to validate the algorithm on the testing device,
wherein the performance output metrics comprise of statistical measures of error distribution including bias, variance of 95% limits of agreement.
2. The method of claim 1 , wherein the generating the at least one synthetic waveform includes adding a random variability and noise.
3. The method of claim 1 , wherein a synthetic model is used in the generating the at least one synthetic waveform.
4. The method of claim 3 , wherein the synthetic model includes generating a synthetic Electrocardiography (ECG) waveforms and applying at least one of a scaling factor, random variability and random noise for synthetic ECG waveforms.
5. The method of claim 4 , wherein the scaling factor transposes a dimensionless ECG representation z to mV/(rad×s).
6. The method of claim 3 , wherein the synthetic model includes generating synthetic triaxial accelerometer waveforms of the testing device comprises a tri-axial accelerometer and applying a random noise and random variability to the synthetic acceleration waveforms.
7. A method for validating an algorithm on a testing device, the method comprising:
generating at least one synthetic waveform from at least two input values, wherein a synthetic model is used in the generating the at least one synthetic waveform;
inputting the generated at least one synthetic: waveform to the testing device;
capturing output values from the testing device;
comparing the output values to the at least two input values generating the at least one synthetic waveform to determine performance metrics of the testing device to validate the algorithm on the testing device; and
generating a synthetic waveform for a constellation of all combinations of input parameters fundamental frequency and modulating frequency.
8. The method of claim 1 , wherein the comparing the output values to the at least two input values generating the at least one synthetic waveform comprises selecting a region of interest in the input constellation of parameters based on a ratio of at least two input independent frequencies, including fundamental frequency fc and modulating frequency fr.
9. A system for validating an algorithm, the system comprising:
a testing device; and
a simulation platform, comprising:
generating at least one synthetic waveform from at least two input values;
inputting the at least one generated synthetic waveform to the testing device;
capturing output values from the testing device; and
comparing the output values to the at least two input values generating the at least one synthetic waveform to determine performance metrics for the testing device to validate the algorithm on the testing device,
wherein a synthetic model is used in generating the at least one synthetic waveform,
wherein the synthetic model includes generating synthetic ECG waveforms and applying at least one of a scaling factor, random variability and random noise for synthetic Electrocardiography (ECG) waveforms, and
wherein the scaling factor transposes a dimensionless ECG representation z to mV/(rad×s).
10. The system of claim 9 , wherein the at least one generating a synthetic waveform includes adding a random variable.
11. The system of claim 9 , wherein the synthetic model includes generating synthetic triaxial accelerometer waveforms of a testing device comprises a tri-axial accelerometer and applying a random noise and random variability to the synthetic acceleration waveforms.
12. The system of claim 9 , comprising generating synthetic waveform for a constellation of all combinations of input parameters such as fundamental frequency and modulating frequency.
13. The system of claim 9 , wherein the comparing the output values to the at least wo input values generating the at least one synthetic waveform comprises selecting R, a region of interest in the input constellation of parameters based on a ratio of at least two input independent frequencies, including fundamental frequency fc and modulating frequency fr.
14. A computer program product stored on a non-transitory computer readable medium, comprising computer readable programming for causing a computer to control an execution of an application for validating an algorithm on a testing device comprising:
generating at least one synthetic waveform from at least two input values;
inputting the generated at least one synthetic waveform to the testing device;
capturing output values from the testing device; and
comparing the output values to the at least two input values generating the at least one synthetic waveform to determine performance metrics of the testing device to validate the algorithm on the testing device,
wherein the comparing the output values to the at least two input values generating the at least one synthetic waveform comprises selecting R, a region of interest in the in input constellation of parameters based on a ratio of at least two input independent frequencies, including fundamental frequency fc and modulating frequency fr.
15. The computer program product of claim 14 , wherein the generating the at least one synthetic waveform includes adding at least one of a scaling factor, random variability and noise.