Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

Quantum Machine Learning for Speech and Audio Understanding: A PRISMA-Aligned Systematic Review

Deepak Bolleddu, Ping Yu, U. Rajendra Acharya

Published May 16, 2026• 0 citations
Publisher

Abstract

Background. Speech and audio understanding underpin affective computing, human–computer interaction, and clinical decision support. Classical deep-learning systems for tasks such as speech emotion recognition (SER) are increasingly bound by data demand, parameter scale, and cross-corpus brittleness. Quantum machine learning (QML) offers a complementary paradigm for high-dimensional, temporally structured signals but has not been systematically synthesised. Methods. We conducted a PRISMA-aligned systematic review of QML methods applied to speech and audio understanding. Six databases (IEEE Xplore, ACM, Scopus, Web of Science, PubMed, arXiv) were searched for studies published between January 2019 and April 2026. Of 2,623 deduplicated records, 55 were confirmed as primary (Tier-1) evidence, with 272 retained as adjacent or backbone evidence. Each Tier-1 study was scored against an eight-question quality instrument and entered into a comprehensive comparison matrix. Results. The corpus splits into 16 SER studies, 24 other speech tasks, and 15 general audio and music studies. Variational and parameterised quantum circuits (18 studies), quantum kernel methods (9), and quantum convolutional networks (8) dominate. Three structural weaknesses recur: aggressive classical preprocessing precedes quantum encoding in nearly every study; 26/55 score zero on NISQ-noise consideration; only 2 perform a formal significance test for the central quantum-versus-classical comparison. Conclusion. The field is in early but rapid and uneven growth (with annual outputs ranging from 2 to 16 papers across 2019–2025) but not yet ready to displace classical foundation models. We synthesise a method-by-task taxonomy and three-horizon research roadmap, and operationalise an eight-item benchmarking protocol as a community call to address recurrent reliability gaps and help raise the floor of QML–audio research.

DOI
10.17605/osf.io/wxh2s

Related papers

Quantum Computing, Quantum Machine Learning, and Quantum Mechanical Principles in Medical Physics: A PRISMA 2020-Compliant Systematic Review

Gökhan ZORLU, Cemil Çolak

Objectives: The convergence of quantum computing, quantum machine learning, and quantum mechanical principles represents an emerging paradigm in medical physics. Classical methods encounter hard limits in treatment planning, molecular-scale drug simulation, and medical image interpretation. The four quantum phenomena superposition, entanglement, quantum tunneling, and quantum coherence offer distinct routes around these barriers through both hardware sensing and algorithmic frameworks.To systematically identify, appraise, and synthesise evidence on applications of quantum computing, quantum machine learning, and quantum mechanical principles in medical physics, mapping the evidence base, assessing clinical readiness, and proposing a research roadmap.Materials and Methods: A PRISMA 2020-compliant systematic search was conducted across PubMed/MEDLINE, Scopus, IEEE Xplore, Web of Science, arXiv, and bioRxiv (January 2015 – March 2026). Two independent reviewers screened 5,248 identified records. After deduplication, screening, and eligibility assessment, 133 studies were included in the qualitative synthesis and 89 in the quantitative performance comparison. Methodological quality was assessed using a modified CLAIM checklist and QUADAS-2 framework.Results: Four application clusters emerged. Superposition underpins hyperpolarized ¹³C MRI (technology readiness level (TRL) 7) and hybrid quantum CNN architectures. Entanglement enables OPM-MEG neuroimaging (TRL 6–7) and quantum kernel classification. Quantum tunneling drives radiotherapy optimisation (TRL 4–5). Quantum coherence supports variational algorithms for drug discovery (TRL 2–3) and quantum dot theranostics (TRL 3–4). Hybrid quantum–classical methods outperformed classical baselines, though most studies showed high risk of bias.Conclusions: Quantum-enhanced sensing and hybrid quantum–classical treatment planning are credible near-term clinical translation targets. Fault-tolerant drug simulation and large-scale quantum AI await hardware maturation beyond 2030. Critical enablers include standardised benchmarking, quantum-specific regulatory guidance, and equity planning.

ODÜ Tıp Dergisi•Oct 1, 2026•0 citations

A systematic literature review of quantum machine learning for medical: trends, datasets, topics, and methods

Agustinus Eko Setiawan, Supriadi Rustad, Abdul Syukur +5 more

Quantum Machine Learning (QML) has emerged as a promising interdisciplinary paradigm that integrates quantum computing and machine learning to address limitations of classical approaches in processing complex high-dimensional heterogeneous and imbalanced medical data. This study presents a systematic literature review of QML applications in healthcare based on 94 primary studies published between 2020 and April 2025. The review follows Kitchenham’s guidelines and the PRISMA framework to ensure rigor transparency and reproducibility. Seven research questions were formulated to examine publication trends sources datasets topics quantum model categories algorithms methods and evaluation metrics. Findings reveal a sharp increase in publication activity after 2023 indicating growing academic interest in healthcare-oriented QML. Medical imaging datasets were most frequently used followed by signal-based and tabular datasets with 86.5% of studies relying on public data. Hybrid classical-quantum approaches including CNN-QSVM Quantum Neural Networks and Variational Quantum Classifiers were dominant because they combine classical feature extraction with quantum-enhanced learning. However most studies were conducted in simulation environments rather than real hardware reflecting NISQ limitations including noise decoherence limited qubits and circuit depth constraints. Many results used small curated or imbalanced datasets raising concerns regarding generalizability reproducibility and clinical validity. Although numerous studies reported high accuracy standardized evaluation protocols external validation and uncertainty assessment remain limited. Overall QML in healthcare remains exploratory with substantial potential but significant barriers. This review highlights progress gaps and future directions emphasizing clinically representative datasets robust benchmarking interpretable noise-resilient models and real-hardware implementation to support trustworthy deployment in healthcare systems worldwide settings today.

International Journal of Cognitive Computing in Engineering•May 10, 2026•3 citations

Quantum Machine Learning in Oncology: A Systematic Review of Clinical Applications, Challenges, and Future Research Directions

Khairil Imran Ghauth, Yanche Ari Kustiawan

Quantum machine learning (QML) has emerged as a promising approach for analyzing the complex, high-dimensional data encountered in oncology, yet research in this area remains fragmented. This systematic literature review synthesizes current applications of QML in cancer care. Following PRISMA guidelines, six databases were searched for peer-reviewed English-language studies published between 2020 and 2026. Of the 212 records identified, 49 studies met the inclusion criteria after screening and quality assessment. The findings show that QML research is dominated by classification and detection tasks, while segmentation is beginning to emerge. Breast cancer and brain tumors are the most frequently investigated domains. Hybrid quantum-classical models, particularly quantum kernel methods, quantum neural networks, and quantum convolutional neural networks, are the predominant approaches. The main barriers to adoption are hardware limitations, including quantum noise, limited qubit availability, and reliance on simulators. Overall, QML in oncology remains in its early stages of development, with limited clinical validation, insufficient model interpretability, and little evidence of a clear quantum advantage. Future research should prioritize evaluation on real quantum hardware, larger and more diverse clinical datasets, standardized benchmarking against classical methods, and closer collaboration between computer scientists and oncology experts to facilitate clinical translation.

Machine Learning and Knowledge Extraction•Aug 13, 2026•1 citation

A systematic review of quantum machine learning for digital health

Riddhi Swaroop Gupta, Carolyn E. Wood, Teyl Engstrom +2 more

The growth in digitization of health data provides opportunities for using algorithmic techniques for data analysis. This systematic review assesses whether quantum machine learning (QML) algorithms outperform existing classical methods for clinical decisioning or health service delivery. Included studies use electronic health/medical records, or reasonable proxy data, and QML algorithms designed for quantum computing hardware. Databases PubMed, Embase, IEEE, Scopus, and preprint server arXiv were searched for studies dated 01/01/2015-10/06/2024. Of an initial 4915 studies, 169 were eligible, with 123 then excluded for insufficient rigor. Only 16 studies consider realistic operating conditions involving quantum hardware or noisy simulations. We find nearly all encountered quantum models form a subset of general QML structures. Scalability of data encoding is partly addressed but requires restrictive hardware assumptions. Overall, performance differentials between quantum and classical algorithms show no consistent trend to support empirical quantum utility in digital health.

arXiv:2410.02446•npj Digital Medicine•May 2, 2025•117 citations

Machine Learning and Quantum Machine Learning: A Comprehensive Review of Algorithms, Applications, and Future Directions

Loveleen Kumar, Rajesh Rajaan, Nilam Choudhary +1 more

The rapid advancement of artificial intelligence has positioned Machine Learning (ML) as a cornerstone technology across scientific and industrial domains. In parallel, the emergence of quantum computing has catalyzed a new interdisciplinary field—Quantum Machine Learning (QML)—which promises to transcend the computational barriers faced by classical algorithms. This review paper systematically examines the evolution of ML from foundational statistical models to contemporary transformer architectures, and provides a structured analysis of QML paradigms including variational quantum circuits (VQCs), quantum kernel methods, and quantum neural networks (QNNs). Drawing from 95+ peer-reviewed publications indexed in Scopus and Science Citation Index (SCI) journals—with particular emphasis on 2025–2026 publications in IEEE Access, Nature Communications, Nature Computational Science, and Physical Review Letters—we analyze performance benchmarks, identify hardware constraints, and outline algorithmic innovations shaping the near-term quantum advantage landscape. Our findings indicate that while classical ML retains superiority in large-scale perception tasks, QML demonstrates significant advantages in combinatorial optimization and quantum chemistry simulation. Recent 2025 advances in error mitigation on superconducting qubits and improved VQC barren-plateau mitigation are narrowing this gap rapidly. The paper concludes with a forward-looking research agenda covering fault-tolerant QML, hybrid architectures, and quantum NLP.

International Journal of Advanced Engineering Research and Science•Jan 1, 2026•0 citations

Quantum machine learning: A comprehensive survey of models, tools, and challenges

Taha M. Mahmoud, Marwa R. Hassan, Shenson Joseph +2 more

Abstract Quantum Machine Learning (QML) is an emerging interdisciplinary field that merges quantum computing with classical machine learning (ML) to enhance computational efficiency, scalability, and accuracy in data-driven tasks. By leveraging quantum properties such as superposition, entanglement, and quantum parallelism, QML has the potential to solve complex problems that are intractable for classical models. This survey provides a comprehensive, systematic review of recent advancements in QML, conducted following the PRISMA 2020 guidelines and covering fundamental encoding techniques that enable quantum computers to process classical data, as well as essential tools, datasets, and evaluation metrics used in quantum model development. Furthermore, it explores various QML approaches, including supervised, semi-supervised, unsupervised, and reinforcement learning, alongside quantum deep learning (QDL) models such as Quantum Neural Networks (QNNs) and Variational Quantum Circuits (VQCs). The paper also discusses the challenges in QML, including hardware limitations, scalability issues, and the need for improved quantum error correction. A dual-axis taxonomy simultaneously classifies algorithms by learning paradigm and quantum-classical integration level – a structure not found in prior surveys. The survey also covers Quantum Transfer Learning, Quantum Graph Neural Networks, and hardware-aware encoding selection, and introduces the first structured adversarial robustness taxonomy and Explainable QML treatment in a comprehensive QML survey. Together, these contributions position this work as both a reference and a methodological guide for advancing QML research and applications.

Journal of King Saud University - Computer and Information Sciences•Oct 19, 2026•0 citations
Background. Speech and audio understanding underpin affective computing, human–computer interaction, and clinical decision support. Classical deep-learning systems for tasks such as speech emotion recognition (SER) are increasingly bound by data demand, parameter scale, and cross-corpus brittleness. Quantum machine learning (QML) offers a complementary paradigm for high-dimensional, temporally structured signals but has not been systematically synthesised. Methods. We conducted a PRISMA-aligned systematic review of QML methods applied to speech and audio understanding. Six databases (IEEE Xplore, ACM, Scopus, Web of Science, PubMed, arXiv) were searched for studies published between January 2019 and April 2026. Of 2,623 deduplicated records, 55 were confirmed as primary (Tier-1) evidence, with 272 retained as adjacent or backbone evidence. Each Tier-1 study was scored against an eight-question quality instrument and entered into a comprehensive comparison matrix. Results. The corpus splits into 16 SER studies, 24 other speech tasks, and 15 general audio and music studies. Variational and parameterised quantum circuits (18 studies), quantum kernel methods (9), and quantum convolutional networks (8) dominate. Three structural weaknesses recur: aggressive classical preprocessing precedes quantum encoding in nearly every study; 26/55 score zero on NISQ-noise consideration; only 2 perform a formal significance test for the central quantum-versus-classical comparison. Conclusion. The field is in early but rapid and uneven growth (with annual outputs ranging from 2 to 16 papers across 2019–2025) but not yet ready to displace classical foundation models. We synthesise a method-by-task taxonomy and three-horizon research roadmap, and operationalise an eight-item benchmarking protocol as a community call to address recurrent reliability gaps and help raise the floor of QML–audio research.
Objectives: The convergence of quantum computing, quantum machine learning, and quantum mechanical principles represents an emerging paradigm in medical physics. Classical methods encounter hard limits in treatment planning, molecular-scale drug simulation, and medical image interpretation. The four quantum phenomena superposition, entanglement, quantum tunneling, and quantum coherence offer distinct routes around these barriers through both hardware sensing and algorithmic frameworks.To systematically identify, appraise, and synthesise evidence on applications of quantum computing, quantum machine learning, and quantum mechanical principles in medical physics, mapping the evidence base, assessing clinical readiness, and proposing a research roadmap.Materials and Methods: A PRISMA 2020-compliant systematic search was conducted across PubMed/MEDLINE, Scopus, IEEE Xplore, Web of Science, arXiv, and bioRxiv (January 2015 – March 2026). Two independent reviewers screened 5,248 identified records. After deduplication, screening, and eligibility assessment, 133 studies were included in the qualitative synthesis and 89 in the quantitative performance comparison. Methodological quality was assessed using a modified CLAIM checklist and QUADAS-2 framework.Results: Four application clusters emerged. Superposition underpins hyperpolarized ¹³C MRI (technology readiness level (TRL) 7) and hybrid quantum CNN architectures. Entanglement enables OPM-MEG neuroimaging (TRL 6–7) and quantum kernel classification. Quantum tunneling drives radiotherapy optimisation (TRL 4–5). Quantum coherence supports variational algorithms for drug discovery (TRL 2–3) and quantum dot theranostics (TRL 3–4). Hybrid quantum–classical methods outperformed classical baselines, though most studies showed high risk of bias.Conclusions: Quantum-enhanced sensing and hybrid quantum–classical treatment planning are credible near-term clinical translation targets. Fault-tolerant drug simulation and large-scale quantum AI await hardware maturation beyond 2030. Critical enablers include standardised benchmarking, quantum-specific regulatory guidance, and equity planning.
Quantum Machine Learning (QML) has emerged as a promising interdisciplinary paradigm that integrates quantum computing and machine learning to address limitations of classical approaches in processing complex high-dimensional heterogeneous and imbalanced medical data. This study presents a systematic literature review of QML applications in healthcare based on 94 primary studies published between 2020 and April 2025. The review follows Kitchenham’s guidelines and the PRISMA framework to ensure rigor transparency and reproducibility. Seven research questions were formulated to examine publication trends sources datasets topics quantum model categories algorithms methods and evaluation metrics. Findings reveal a sharp increase in publication activity after 2023 indicating growing academic interest in healthcare-oriented QML. Medical imaging datasets were most frequently used followed by signal-based and tabular datasets with 86.5% of studies relying on public data. Hybrid classical-quantum approaches including CNN-QSVM Quantum Neural Networks and Variational Quantum Classifiers were dominant because they combine classical feature extraction with quantum-enhanced learning. However most studies were conducted in simulation environments rather than real hardware reflecting NISQ limitations including noise decoherence limited qubits and circuit depth constraints. Many results used small curated or imbalanced datasets raising concerns regarding generalizability reproducibility and clinical validity. Although numerous studies reported high accuracy standardized evaluation protocols external validation and uncertainty assessment remain limited. Overall QML in healthcare remains exploratory with substantial potential but significant barriers. This review highlights progress gaps and future directions emphasizing clinically representative datasets robust benchmarking interpretable noise-resilient models and real-hardware implementation to support trustworthy deployment in healthcare systems worldwide settings today.
Quantum machine learning (QML) has emerged as a promising approach for analyzing the complex, high-dimensional data encountered in oncology, yet research in this area remains fragmented. This systematic literature review synthesizes current applications of QML in cancer care. Following PRISMA guidelines, six databases were searched for peer-reviewed English-language studies published between 2020 and 2026. Of the 212 records identified, 49 studies met the inclusion criteria after screening and quality assessment. The findings show that QML research is dominated by classification and detection tasks, while segmentation is beginning to emerge. Breast cancer and brain tumors are the most frequently investigated domains. Hybrid quantum-classical models, particularly quantum kernel methods, quantum neural networks, and quantum convolutional neural networks, are the predominant approaches. The main barriers to adoption are hardware limitations, including quantum noise, limited qubit availability, and reliance on simulators. Overall, QML in oncology remains in its early stages of development, with limited clinical validation, insufficient model interpretability, and little evidence of a clear quantum advantage. Future research should prioritize evaluation on real quantum hardware, larger and more diverse clinical datasets, standardized benchmarking against classical methods, and closer collaboration between computer scientists and oncology experts to facilitate clinical translation.
The growth in digitization of health data provides opportunities for using algorithmic techniques for data analysis. This systematic review assesses whether quantum machine learning (QML) algorithms outperform existing classical methods for clinical decisioning or health service delivery. Included studies use electronic health/medical records, or reasonable proxy data, and QML algorithms designed for quantum computing hardware. Databases PubMed, Embase, IEEE, Scopus, and preprint server arXiv were searched for studies dated 01/01/2015-10/06/2024. Of an initial 4915 studies, 169 were eligible, with 123 then excluded for insufficient rigor. Only 16 studies consider realistic operating conditions involving quantum hardware or noisy simulations. We find nearly all encountered quantum models form a subset of general QML structures. Scalability of data encoding is partly addressed but requires restrictive hardware assumptions. Overall, performance differentials between quantum and classical algorithms show no consistent trend to support empirical quantum utility in digital health.
The rapid advancement of artificial intelligence has positioned Machine Learning (ML) as a cornerstone technology across scientific and industrial domains. In parallel, the emergence of quantum computing has catalyzed a new interdisciplinary field—Quantum Machine Learning (QML)—which promises to transcend the computational barriers faced by classical algorithms. This review paper systematically examines the evolution of ML from foundational statistical models to contemporary transformer architectures, and provides a structured analysis of QML paradigms including variational quantum circuits (VQCs), quantum kernel methods, and quantum neural networks (QNNs). Drawing from 95+ peer-reviewed publications indexed in Scopus and Science Citation Index (SCI) journals—with particular emphasis on 2025–2026 publications in IEEE Access, Nature Communications, Nature Computational Science, and Physical Review Letters—we analyze performance benchmarks, identify hardware constraints, and outline algorithmic innovations shaping the near-term quantum advantage landscape. Our findings indicate that while classical ML retains superiority in large-scale perception tasks, QML demonstrates significant advantages in combinatorial optimization and quantum chemistry simulation. Recent 2025 advances in error mitigation on superconducting qubits and improved VQC barren-plateau mitigation are narrowing this gap rapidly. The paper concludes with a forward-looking research agenda covering fault-tolerant QML, hybrid architectures, and quantum NLP.
Abstract Quantum Machine Learning (QML) is an emerging interdisciplinary field that merges quantum computing with classical machine learning (ML) to enhance computational efficiency, scalability, and accuracy in data-driven tasks. By leveraging quantum properties such as superposition, entanglement, and quantum parallelism, QML has the potential to solve complex problems that are intractable for classical models. This survey provides a comprehensive, systematic review of recent advancements in QML, conducted following the PRISMA 2020 guidelines and covering fundamental encoding techniques that enable quantum computers to process classical data, as well as essential tools, datasets, and evaluation metrics used in quantum model development. Furthermore, it explores various QML approaches, including supervised, semi-supervised, unsupervised, and reinforcement learning, alongside quantum deep learning (QDL) models such as Quantum Neural Networks (QNNs) and Variational Quantum Circuits (VQCs). The paper also discusses the challenges in QML, including hardware limitations, scalability issues, and the need for improved quantum error correction. A dual-axis taxonomy simultaneously classifies algorithms by learning paradigm and quantum-classical integration level – a structure not found in prior surveys. The survey also covers Quantum Transfer Learning, Quantum Graph Neural Networks, and hardware-aware encoding selection, and introduces the first structured adversarial robustness taxonomy and Explainable QML treatment in a comprehensive QML survey. Together, these contributions position this work as both a reference and a methodological guide for advancing QML research and applications.