A mixed-radix FFT algorithm implementation based on Petri nets to assist the certification of bio-medical systems

Abstract

Bio-medical signal acquisition systems (or simply bio-medical systems) are commonly used during the diagnosis, monitoring, and treatment of patients. The Electrocardiograph (ECG) is an example of this type of system. It is necessary to submit bio-medical systems to regulatory agencies for certification. Regulatory agencies conduct evaluations seeking the commercialization of high-quality systems. In a previous work, we specified an ECG model by integrating the Coloured Petri Nets (CPN) and MATLAB® tools. However, this introduced an external component dependency that limited the application and re-usability of the model during the certification process. We present in this paper a mixed-radix Fast Fourier Transform (FFT) algorithm based on CPN aiming to assist the design of bio-medical systems using CPN. We improved the design of an ECG model to show how the mixed-radix FFT algorithm solves the previous limitations, and to evaluate the effectiveness of a CPN/ML implementation. The comparison of a system under development with similar certified systems is commonly accepted by regulatory agencies. The replacement of a similar system by the open source CPN/Tools software and the available reference model can decrease costs, and development and certification time.