An adaptive neurofuzzy technique for determination of blood acidity

Abstract

This paper presents an adaptive neurofuzzy-based method for the determination of blood acidity (pH). The main advantage of this method in comparison with conventional ones used for blood pH measurements is that it is capable of estimating the blood pH without the need for a pH sensor, which in turn reduces the volume of blood sample required to conduct the chemical analysis. This method uses blood carbon dioxide partial pressure (pCO₂) and bicarbonate (HCO⁻₃) as input to a neurofuzzy approach to predict the value of blood pH. This method was validated using 60 test data points that had not been used during the training process. The obtained results showed that the pH values predicted using this method have good concordance with experimentally measured pH values. The high correlation coefficient (87.6%) between the measured and the predicted pH values reflects the method's ability to measure (estimate) the pH values using pCO₂ and HCO⁻₃ with accuracy satisfying clinical demands.

Design of an integrated model for diagnosis and classification of pediatric acute leukemia using machine learning

“…The particles are then manipulated by equations (23) and (24). The i-particle location of the velocity vector is changed according to equation (23) x…”

“…The exchange of information between particles is applied to the particle as a velocity vector, as stated by equation (24)…”

“…23 Another study using the adaptive neuro-fuzzy method has been done for determining the blood acidity (pH). 24 In order to classify heart sound signals, a biomedical-based system was presented that utilized 120 subjects with normal, pulmonary, and mitral stenosis heart valve diseases via a stethoscope. 25 The author applied ANFIS for diagnosis of the heart valve diseases using wavelet transform with entropy.…”