ResQ: A hybrid classical-quantum model for efficient breast cancer image classification

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F61989100%3A27240%2F25%3A10260359" target="_blank" >RIV/61989100:27240/25:10260359 - isvavai.cz</a>

Výsledek na webu

<a href="https://www.sciencedirect.com/science/article/pii/S1568494625009421?pes=vor&utm_source=clarivate&getft_integrator=clarivate" target="_blank" >https://www.sciencedirect.com/science/article/pii/S1568494625009421?pes=vor&utm_source=clarivate&getft_integrator=clarivate</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1016/j.asoc.2025.113631" target="_blank" >10.1016/j.asoc.2025.113631</a>

Jazyk výsledku

angličtina

Název v původním jazyce

ResQ: A hybrid classical-quantum model for efficient breast cancer image classification

Popis výsledku v původním jazyce

Breast cancer remains one of the leading causes of mortality worldwide, necessitating accurate and efficient diagnostic systems to improve treatment outcomes. While deep learning-based Computer-Aided Diagnostic (CAD) tools have demonstrated promise in analyzing histopathological images, they face challenges in handling high-dimensional data and computational inefficiencies. Simultaneously, quantum computing has emerged as a transformative technology, offering unparalleled capabilities in modeling complex data distributions and accelerating computations. This paper introduces ResQ, a hybrid classical-quantum framework designed for breast cancer classification. ResQ integrates a ResNet-based feature extraction module with a Variational Quantum Circuit (VQC) classifier, leveraging the complementary strengths of classical deep learning and quantum computing. Evaluations on two publicly available datasets, BreakHis and Bioimaging (BI), reveal significant performance improvements, achieving accuracies of 98.49% and 88.96%, respectively, compared to 97.36% and 87.96% achieved by its classical counterparts. Quantum circuit evaluations have been conducted on a quantum simulator (Aer simulator) as well as a noise-affected real quantum processor (ibm_brisbane). Additionally, to gain deeper insights into the effectiveness of the ResQ model over classical models, a non-parametric statistical test, viz., the Friedman test, followed by the Nemenyi test for post hoc analysis, is performed. Furthermore, a detailed circuit-level analysis explores critical trade-offs, such as circuit depth, gate count, and qubit usage, providing unique insights into the practical deployment of hybrid classical-quantum models in medical imaging. The one-qubit shallow architecture of ResQ renders lower circuit complexities, making it amenable to Noisy-Intermediate Scale Quantum (NISQ) devices. These findings underscore the potential of quantum computing to revolutionize cancer diagnostics by enhancing both accuracy and computational efficiency. The relevant codes of the proposed architecture, ResQ, are publicly available on https://github.com/DVLP-CMATERJU/ResQ-Hybrid-Classical-Quantum.

Název v anglickém jazyce

ResQ: A hybrid classical-quantum model for efficient breast cancer image classification

Popis výsledku anglicky

Breast cancer remains one of the leading causes of mortality worldwide, necessitating accurate and efficient diagnostic systems to improve treatment outcomes. While deep learning-based Computer-Aided Diagnostic (CAD) tools have demonstrated promise in analyzing histopathological images, they face challenges in handling high-dimensional data and computational inefficiencies. Simultaneously, quantum computing has emerged as a transformative technology, offering unparalleled capabilities in modeling complex data distributions and accelerating computations. This paper introduces ResQ, a hybrid classical-quantum framework designed for breast cancer classification. ResQ integrates a ResNet-based feature extraction module with a Variational Quantum Circuit (VQC) classifier, leveraging the complementary strengths of classical deep learning and quantum computing. Evaluations on two publicly available datasets, BreakHis and Bioimaging (BI), reveal significant performance improvements, achieving accuracies of 98.49% and 88.96%, respectively, compared to 97.36% and 87.96% achieved by its classical counterparts. Quantum circuit evaluations have been conducted on a quantum simulator (Aer simulator) as well as a noise-affected real quantum processor (ibm_brisbane). Additionally, to gain deeper insights into the effectiveness of the ResQ model over classical models, a non-parametric statistical test, viz., the Friedman test, followed by the Nemenyi test for post hoc analysis, is performed. Furthermore, a detailed circuit-level analysis explores critical trade-offs, such as circuit depth, gate count, and qubit usage, providing unique insights into the practical deployment of hybrid classical-quantum models in medical imaging. The one-qubit shallow architecture of ResQ renders lower circuit complexities, making it amenable to Noisy-Intermediate Scale Quantum (NISQ) devices. These findings underscore the potential of quantum computing to revolutionize cancer diagnostics by enhancing both accuracy and computational efficiency. The relevant codes of the proposed architecture, ResQ, are publicly available on https://github.com/DVLP-CMATERJU/ResQ-Hybrid-Classical-Quantum.

Druh

J_imp - Článek v periodiku v databázi Web of Science

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

—

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2025

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název periodika

Applied Soft Computing

ISSN

1568-4946

e-ISSN

1872-9681

Svazek periodika

183

Číslo periodika v rámci svazku

November

Stát vydavatele periodika

NL - Nizozemsko

Počet stran výsledku

25

Strana od-do

nestránkováno

Kód UT WoS článku

001541575100002

EID výsledku v databázi Scopus

—