The relationship between the perfusion index and precision of noninvasive blood component measurement based on dynamic spectroscopy

Noninvasive blood component measurement is a relevant topic in biomedical engineering. Noninvasive blood component measurement based on dynamic spectroscopy can effectively avoid the influence of individual differences and improve its prediction precision. The relationship between the perfusion index and precision of hemoglobin is explored in this paper. The method perfusion index grouping optimization modeling is put forward to restrain the influence of the perfusion index and to improve the prediction precision of hemoglobin. The study is based on clinical trial data from 340 volunteers, and the age of volunteers ranges from 22 to 82. The extreme learning machine modeling approach is used in our study. By considering the influence of the perfusion index, the root mean square error of the prediction set has been improved from 12.8753 to 11.3039 and the determination coefficient of the prediction set has been improved from 0.4728 to 0.5107. The result indicates that the perfusion index has an effect on the precision of noninvasive blood component measurement based on dynamic spectroscopy. By using the perfusion index grouping optimization modeling method, the root mean square error of the prediction set has been improved to 10.6274 and the determination coefficient of the prediction set has been improved to 0.6931. The result means that this method can effectively restrain the influences of the perfusion index and improve the prediction precision of noninvasive blood component measurement based on dynamic spectroscopy. In addition, the interference factors of this method are discussed. The prediction precision of blood components can be further improved by controlling these factors.