This paper presents a novel approach for COVID-19 detection using lightweight deep learning methods (LDMs) applied to lung CT slices. The proposed method, named COVID-19 Detection in Lung CT Slices using Brownian-butterfly-algorithm optimized lightweight deep features (CBDL), aims to enhance the accuracy of COVID-19 detection through image collection, initial processing with Shannon’s thresholding, deep-feature mining, feature optimization with the Brownian Butterfly Algorithm (BBA), and binary classification through three-fold cross-validation. The performance evaluation involves assessing individual, fused, and ensemble features. The results show that the CBDL scheme achieves a detection accuracy of 93.80% with individual features, 96% with fused features, and an impressive 99.10% with ensemble features, demonstrating its effectiveness in enhancing COVID-19 detection accuracy in lung CT databases. The key contributions of the research include lung CT enhancement using Shannon’s tri-level thresholding, feature extraction and optimization using LDMs and BBA, dual-deep feature-supported infection detection, and ensemble deep features for improved accuracy. The study also discusses the implementation details, experimental results, and comparisons with existing methods, highlighting the superior performance of the proposed scheme.This paper presents a novel approach for COVID-19 detection using lightweight deep learning methods (LDMs) applied to lung CT slices. The proposed method, named COVID-19 Detection in Lung CT Slices using Brownian-butterfly-algorithm optimized lightweight deep features (CBDL), aims to enhance the accuracy of COVID-19 detection through image collection, initial processing with Shannon’s thresholding, deep-feature mining, feature optimization with the Brownian Butterfly Algorithm (BBA), and binary classification through three-fold cross-validation. The performance evaluation involves assessing individual, fused, and ensemble features. The results show that the CBDL scheme achieves a detection accuracy of 93.80% with individual features, 96% with fused features, and an impressive 99.10% with ensemble features, demonstrating its effectiveness in enhancing COVID-19 detection accuracy in lung CT databases. The key contributions of the research include lung CT enhancement using Shannon’s tri-level thresholding, feature extraction and optimization using LDMs and BBA, dual-deep feature-supported infection detection, and ensemble deep features for improved accuracy. The study also discusses the implementation details, experimental results, and comparisons with existing methods, highlighting the superior performance of the proposed scheme.