This study integrates a tunicate swarm algorithm (TSA) with a local escaping operator (LEO) for overcoming the weaknesses of the original TSA. The LEO strategy in TSA–LEO prevents searching deflation in TSA and improves the convergence rate and local search efficiency of swarm agents. The efficiency of the proposed TSA–LEO was verified on the CEC’2017 test suite, and its performance was compared with seven metaheuristic algorithms (MAs). The comparisons revealed that LEO significantly helps TSA by improving the quality of its solutions and accelerating the convergence rate. TSA–LEO was further tested on a real-world problem, namely, segmentation based on the objective functions of Otsu and Kapur. A set of well-known evaluation metrics was used to validate the performance and segmentation quality of the proposed TSA–LEO. The proposed TSA-LEO outperforms other MA algorithms in terms of fitness, peak signal-to-noise ratio, structural similarity, feature similarity, and segmentation findings.

Artificial intelligence (AI) techniques have recently been used in a variety of medical applications. AI is also used to treat atrial and ventricular arrhythmias in order to reduce their mortality rate. Early detection and classification of ventricular arrhythmias may allow for more effective treatment. Convolutional neural networks (CNNs) have improved the detection and classification of ventricular arrhythmias using ECG signals when compared to traditional methods. On the other hand, metaheuristic optimization algorithms had proven their efficiency to tackle many real-world problems. The arithmetic optimization algorithm (AOA) is a recent metaheuristic optimization algorithm inspired by basic arithmetic operations such as summation, subtraction, multiplication, and division. This study proposes a new optimized CNN model based on a CNN and AOA migration that is used to optimize three network hyperparameters: learning rate (LR), number of layers, and dropout rate. The proposed model is evaluated using the BIH dataset, which was chosen because it is one of the most widely used arrhythmia databases for ECG signal processing research. To validate the proposed model's robustness, it was compared to a set of recent studies in terms of Accuracy, Precision, Recall, and F1 Score and the proposed model achieved 99.73 % Accuracy, 95.20 % Precision, 92.20 % Recall, and 93.60 % F1-Score. The findings revealed that the proposed model outperforms all other studies in terms of evaluation metrics used.