Efficient Signal Conditioning for Microcontroller Based Medical Solutions

Signal conditioning is an important aspect in almost all analog measurement applications such as blood pressure monitors, heart rate monitors, digital hearing aids, magnetic resonance imaging, pulsoximeters etc. While some of the high end medical applications need digital signal processors (DSP) for processing their complex algorithms, most portable medical applications need only a microcontroller. Power consumption is of primary interest for portable battery operated Medical equipments. The use of microcontrollers boasts of lower power consumption. Filtering is the fundamental aspect of any signal conditioning scheme and the choice of digital filters over analog filters has several proven advantages. This paper proposes an efficient way to implement digital filters on a microcontroller without the need to scale the filter coefficients. The method uses an extension of Horner's method to support an integer-float multiply. This method exhibits excellent accuracy when compared to implementations using floating point math libraries.