Background: Haloperidol (HP) and Risperidone (RIS) are antipsychotic drugs and the
simultaneous determination of these drugs is important. Estimation of HP and RIS alone or in
combination with other drugs has been performed in a variety of ways.
Objective: The aim of this paper was to propose a rapid, simple, accurate, and robust method for
the simultaneous determination of HP and RIS using Artificial Neural Networks (ANNs), Partial
Least Squares (PLS), and Principal Component Regression (PCR) methods along with spectrophotometry
Methods: The simultaneous spectrophotometric determination of HP and RIS in synthetic mixtures
and biological fluid was performed by applying ANNs Containing Feed-forward Backpropagation
(FFBP) and Radial Basis Function (RBF) networks as intelligent methods, as well as PLS,
and principal component regression PCR as multivariate calibration methods. The Levenberg-
Marquardt (LM), Scaled Conjugate Gradient (SCG), and Resilient Back-propagation (RP) algorithms
with different layers and neurons were used in FFBP network and obtained results were
compared with each other.
Results: Among various algorithms of the FFBP network, the LM algorithm was selected as the
best model with a lower Mean Square Error (MSE). MSE of the RBF model was 1.46×10-25 and
1.62×10-23 for HP and RIS, respectively. On the other hand, the mean recovery of PLS and PCR
was 99.91%, 100.01% and 98.60%, 101.90% for HP and RIS, respectively.
Conclusion: The proposed models and High-Performance Liquid Chromatography (HPLC) as a
reference method were compared with each other by one-way Analysis of Variance (ANOVA)
test at the 95 % confidence level for the urine sample. It was observed that the developed methods
presented comparable results for the simultaneous determination of HP and RIS.