Selected work

Representative Papers

Classification of Disease from Lungs X-ray Images using VGG16, VGG19 and ResNet50 Models

Jul 29, 2026

This study addresses the urgent need for early and accurate diagnosis of pulmonary diseases amid their rising prevalence by systematically evaluating the performance of three prominent deep convolutional neural networks—VGG16, VGG19, and ResNet50—in classifying chest X-ray images. Using a large-scale public dataset, the models were trained and tested on four categories: pneumonia, tuberculosis, lung cancer, and normal lungs. Experimental results demonstrate that ResNet50 significantly outperforms the other architectures in both classification accuracy and computational efficiency, underscoring its superiority for intelligent computer-aided diagnosis of pulmonary conditions and offering a robust, efficient solution with strong potential for clinical deployment.

0 citationsRead paper

$W-δ-μ$ dual codes and LCD codes

Jun 11, 2026

This work proposes a unified $W$-$\delta$-$\mu$ inner product framework that generalizes Euclidean, Hermitian, and $\delta$-inner products to linear codes over finite fields and semisimple rings. Within this framework, the authors systematically define and analyze fundamental code structures—including dual codes, self-orthogonal codes, self-dual codes, dual-containing codes, and LCD (linear complementary dual) codes—and establish, for the first time, existence conditions for such codes over semisimple rings. The study derives explicit dual descriptions under the new inner product for classical code families such as repetition codes, binary codes, and $\lambda$-constacyclic codes, thereby extending classical duality theory and providing a theoretical foundation and constructive tools for emerging code constructions like LCD codes.

0 citationsRead paper

On concatenation of matrices for reversible linear codes over a finite commutative ring and applications to DNA codes

Jun 09, 2026

This work addresses the systematic construction of reversible linear codes, reversible self-dual codes, and reversible DNA codes by proposing a unified matrix concatenation framework based on involution matrices over finite commutative rings. By integrating the matrix product approach with DNA code mapping techniques, the study not only resolves an open problem posed by Oztas et al., but also corrects and improves upon certain of their results, thereby establishing a cohesive theory for constructing reversible codes. The proposed method yields a broad class of generator matrices with favorable parameters, enabling the successful construction of various novel reversible linear and DNA codes, significantly expanding both the scope and performance of existing constructions.

0 citationsRead paper

Recent publications

Latest Papers

Classification of Disease from Lungs X-ray Images using VGG16, VGG19 and ResNet50 Models

Jul 29, 2026

This study addresses the urgent need for early and accurate diagnosis of pulmonary diseases amid their rising prevalence by systematically evaluating the performance of three prominent deep convolutional neural networks—VGG16, VGG19, and ResNet50—in classifying chest X-ray images. Using a large-scale public dataset, the models were trained and tested on four categories: pneumonia, tuberculosis, lung cancer, and normal lungs. Experimental results demonstrate that ResNet50 significantly outperforms the other architectures in both classification accuracy and computational efficiency, underscoring its superiority for intelligent computer-aided diagnosis of pulmonary conditions and offering a robust, efficient solution with strong potential for clinical deployment.

0 citationsRead paper

$W-δ-μ$ dual codes and LCD codes

Jun 11, 2026

This work proposes a unified $W$-$\delta$-$\mu$ inner product framework that generalizes Euclidean, Hermitian, and $\delta$-inner products to linear codes over finite fields and semisimple rings. Within this framework, the authors systematically define and analyze fundamental code structures—including dual codes, self-orthogonal codes, self-dual codes, dual-containing codes, and LCD (linear complementary dual) codes—and establish, for the first time, existence conditions for such codes over semisimple rings. The study derives explicit dual descriptions under the new inner product for classical code families such as repetition codes, binary codes, and $\lambda$-constacyclic codes, thereby extending classical duality theory and providing a theoretical foundation and constructive tools for emerging code constructions like LCD codes.

0 citationsRead paper

On concatenation of matrices for reversible linear codes over a finite commutative ring and applications to DNA codes

Jun 09, 2026

This work addresses the systematic construction of reversible linear codes, reversible self-dual codes, and reversible DNA codes by proposing a unified matrix concatenation framework based on involution matrices over finite commutative rings. By integrating the matrix product approach with DNA code mapping techniques, the study not only resolves an open problem posed by Oztas et al., but also corrects and improves upon certain of their results, thereby establishing a cohesive theory for constructing reversible codes. The proposed method yields a broad class of generator matrices with favorable parameters, enabling the successful construction of various novel reversible linear and DNA codes, significantly expanding both the scope and performance of existing constructions.

0 citationsRead paper