Serag Ahmed, Abduljabbar Maram H, Althobaiti Yusuf S, Almutairi Farooq M, Alsharif Shaker T, Alzhrani Rami M, Ahmed Marwa F, Almalki Atiah H
Pharmaceutical Analytical Chemistry Department, Faculty of Pharmacy, Al-Azhar University, Nasr City, Cairo, 11751, Egypt.
Department of Pharmacology and Toxicology, College of Pharmacy, Taif University, P.O. Box 11099, 21944, Taif, Saudi Arabia.
Sci Rep. 2025 Mar 29;15(1):10838. doi: 10.1038/s41598-025-89187-7.
In the present study, a simple, rapid and cost-effective analytical method was developed for the simultaneous determination of three commonly prescribed cardiovascular drugs: propranolol, rosuvastatin and valsartan. The method employed artificial neural networks (ANN) to model the relation between the UV absorption spectra of the drugs and their concentrations. An experimental design of 25 samples was employed as a calibration set, and a central composite design of 20 samples was used as a validation set. The firefly algorithm (FA) was evaluated as a variable selection procedure to optimize the developed ANN models resulting in simpler models with improved predictive performance as evident by lower relative root mean square error of prediction (RRMSEP) values compared to the full spectrum ANN models. Validation of the developed FA-ANN models demonstrated excellent accuracy, precision and selectivity for the quantification of the target analytes as per international conference on harmonisation (ICH) guidelines. Additionally, the greenness, analytical practicality and sustainability of the developed models were assessed using the analytical greenness (AGREE), blue applicability grade index (BAGI) and the red-green-blue (RGB) tools, confirming their environmentally friendly, practical and sustainable nature. This research shed the light on the potential of ANN coupled with UV fingerprinting for the rapid and simultaneous determination of critical cardiovascular drugs posing a significant impact on pharmaceutical quality control and patient monitoring.
在本研究中，开发了一种简单、快速且经济高效的分析方法，用于同时测定三种常用的心血管药物：普萘洛尔、瑞舒伐他汀和缬沙坦。该方法采用人工神经网络（ANN）对药物的紫外吸收光谱与其浓度之间的关系进行建模。将25个样本的实验设计用作校准集，将20个样本的中心复合设计用作验证集。评估了萤火虫算法（FA）作为变量选择程序，以优化所开发的ANN模型，从而得到更简单的模型，其预测性能得到改善，这从与全光谱ANN模型相比更低的预测相对均方根误差（RRMSEP）值中可以明显看出。根据国际协调会议（ICH）指南，对所开发的FA-ANN模型进行验证，结果表明该模型在定量目标分析物方面具有出色的准确性、精密度和选择性。此外，使用分析绿色度（AGREE）、蓝色适用性等级指数（BAGI）和红绿蓝（RGB）工具评估了所开发模型的绿色度、分析实用性和可持续性，证实了它们对环境友好、实用和可持续的性质。本研究揭示了ANN结合紫外指纹图谱在快速同时测定对药物质量控制和患者监测有重大影响的关键心血管药物方面的潜力。
