Published : 2025-10-03

Liquisolid technique for solubility enhancement of a poorly soluble thrombin inhibitor: optimization using design of experiments and artificial neural networks

Rama Devi Korni

Thanmaisree Bora

Akhil Majji

Jagadeesh Panda

Abstract

The current work uses liquisolid (LS) technology to increase the absorption efficiency and dissolution rate of the weakly water-soluble drug dabigatran etexilate (DE) by formulating it using the design of experiments (DoE) and comparing it with the artificial neural network (ANN). D-optimal design in Stat-Ease software was used to formulate and optimize DE liquisolid capsules. The drug-to-nonvolatile solvent ratio (X1 – 25 to 50), carrier-to-coating ratio (X2 – 10 to 30), type of nonvolatile solvent (X3 – PEG 400 or Kolliphor EL) and type of carrier (X4 – Neusilin or Fujicalin) were taken as independent variables and percentage drug release (Y1) and angle of repose (Y2) were taken as dependent variables. The results were compared to artificial neural networks (ANN) utilizing JMP software to improve the prediction of the chosen output variables. The optimized formulation was developed and evaluated. The R2 value of the D-optimal design for percentage drug release was 0.914, whereas for ANN, it was 0.943. The mean square error (MSE) value of the quadratic model obtained in the D-optimal design was 90.76, and in ANN, it was only 2.392. The R2 value for the quadratic model in the D-optimal design for the angle of repose was 0.723, whereas for ANN, it was 0.751. The MSE for the D‑optimal design was 28.11, whereas for ANN, it was 11.04. Based on the analysis of results, PEG400 was selected as the nonvolatile solvent and Neusilin as a carrier for optimized formulation. The percentage drug release and angle of repose of optimized DE liquisolid capsules were found to be 86.23 ± 1.37 and 36.08 ± 0.63, respectively. XRD studies indicated a reduction in the crystallinity of the drug in liquisolid formulations. The findings suggested that dabigatran etexilate's solubility rate could be increased by using liquisolid capsules. ANN gave better predictability than the design of experiments.

Keywords:

Liquisolid, design of experiments, D-optimal design, artificial neural networks, JMP


Similar Articles

1 2 3 > >> 

You may also start an advanced similarity search for this article.


Details

References

Statistics

Authors

Download files

PDF

Citation rules

Korni, R. D., Bora, T., Majji, A., & Panda, J. (2025). Liquisolid technique for solubility enhancement of a poorly soluble thrombin inhibitor: optimization using design of experiments and artificial neural networks. Prospects in Pharmaceutical Sciences, 23(3), 97–109. https://doi.org/10.56782/pps.340

Altmetric indicators


Cited by / Share


Licence


Editorial Team
Stefana Banacha 1
02-097 Warsaw, Poland
biuletynfarmacji@wum.edu.pl
Publisher:
Medical University of Warsaw
ul. Żwirki i Wigury 61
02-091 Warszawa

About:
Copyright 2021 by
OJS Support and Customization by LIBCOM
Platform & workfow by OJS/PKP