Abstract: Blood pressure signals are reconstructed from PhotoPlethysmoGraphy (PPG) signals by: receiving PPG signals including systolic, diastolic and dicrotic phases; and determining first and second derivatives of the PPG signals and: a first set of values indicative of lengths of the signal paths of the PPG signal, the first derivative and the second derivative thereof in the systolic, diastolic and dicrotic phases; a second set of values indicative of relative durations of the PPG signal and the first and second derivatives thereof in the systolic, diastolic and dicrotic phases; and a third set of values indicative of the time separation of peaks and/or valleys in subsequent waveforms of the PPG signal. Reconstruction also includes applying artificial neural network processing to the first, second and third set of values. The artificial neural network processing includes artificial neural network training as a function of blood pressure signals to produce reconstructed blood pressure signals.

Abstract: Blood pressure signals are reconstructed from PhotoPlethysmoGraphy (PPG) signals by: receiving PPG signals including systolic, diastolic and dicrotic phases; and determining first and second derivatives of the PPG signals and: a first set of values indicative of lengths of the signal paths of the PPG signal, the first derivative and the second derivative thereof in the systolic, diastolic and dicrotic phases; a second set of values indicative of relative durations of the PPG signal and the first and second derivatives thereof in the systolic, diastolic and dicrotic phases; and a third set of values indicative of the time separation of peaks and/or valleys in subsequent waveforms of the PPG signal. Reconstruction also includes applying artificial neural network processing to the first, second and third set of values. The artificial neural network processing includes artificial neural network training as a function of blood pressure signals to produce reconstructed blood pressure signals.