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

arXiv:2603.00223 · International Journal of Theoretical Physics

Pretty Good Measurement for Radiomics: A Quantum-Inspired Multi-Class Classifier for Lung Cancer Subtyping and Prostate Cancer Risk Stratification

Giuseppe Sergioli, Carlo Cuccu, Giovanni Pasini, Alessandro Stefano, Giorgio Russo, Andrés Camilo Granda Arango, Roberto Giuntini

cs.CVquant-phSubmitted Feb 27, 2026• Announced Mar 2, 2026• 0 citations• Revised May 8, 2026 (v2)
v1 · Feb 27, 2026v2 · May 8, 2026
arXivHTMLPDF

Abstract

We investigate a quantum-inspired approach to supervised multi-class classification based on the Pretty Good Measurement (PGM), viewed as an operator-valued decision rule derived from quantum state discrimination. The method associates each class with an encoded mixed state and performs classification through a single POVM construction, thus providing a genuinely multi-class strategy without reduction to pairwise or one-vs-rest schemes. In this perspective, classification is reformulated as the discrimination of a finite ensemble of class-dependent density operators, with performance governed by the geometry induced by the encoding map and by the overlap structure among classes. To assess the practical scope of this framework, we apply the PGM-based classifier to two biomedical radiomics case studies: histopathological subtyping of non-small-cell lung carcinoma (NSCLC) and prostate cancer (PCa) risk stratification. The evaluation is conducted under protocols aligned with previously reported radiomics studies, enabling direct comparison with established classical baselines. The results show that the PGM-based classifier is consistently competitive and, in several settings, improves upon standard methods. In particular, the method performs especially well in the NSCLC binary and three-class tasks, while remaining competitive in the four-class case, where increased class overlap yields a more demanding discrimination geometry. In the PCa study, the PGM classifier remains close to the strongest ensemble baseline and exhibits clinically relevant sensitivity--specificity trade-offs across feature-selection scenarios.

Comments
22 pages, 9 figures, 12 table, in preparation for journal submission

Related papers

QSMOTE-PGM/kPGM: QSMOTE Based PGM and kPGM for Imbalanced Dataset Classification

Bikash K. Behera, Giuseppe Sergioli, Roberto Giuntini

Quantum-inspired machine learning (QiML) employs mathematical principles from quantum theory, such as Hilbert-space representations and quantum state discrimination, to enhance classical learning algorithms. In this work, we investigate the integration of Quantum Synthetic Minority Oversampling Technique (QSMOTE) variants with two quantum-inspired classifiers: the Pretty Good Measurement (PGM) classifier and the kernelized Pretty Good Measurement (KPGM) classifier. We propose and analyze three QSMOTE variants, namely KNN-based, Fidelity-based, and Margin-based QSMOTE, designed to improve minority-class representation in imbalanced datasets through quantum-inspired similarity and sampling mechanisms. A unified theoretical and empirical comparison of PGM and KPGM is presented under amplitude and stereo encoding strategies with multiple quantum copies. Experimental evaluations on the Telco Customer Churn dataset demonstrate that the proposed quantum-inspired approaches consistently outperform a classical Random Forest baseline, particularly in terms of recall and balanced F1-score. Among all configurations, PGM with stereo encoding and n_{copies}=2 achieves the best performance with an accuracy of 0.8512 and an F1-score of 0.8234, while KPGM exhibits competitive and more stable behavior across different QSMOTE variants, reaching accuracies of 0.8511 under stereo encoding and 0.8483 under amplitude encoding. The results further show that increasing the number of quantum copies systematically improves classification performance, especially for minority-class detection. This work highlights the effectiveness of combining quantum-inspired oversampling and classification strategies for imbalanced learning, while providing practical insights into the complementary strengths of measurement-based and kernel-based quantum-inspired machine learning frameworks.

cs.LGCross-listedarXiv:2512.16960•Soft Computing•Dec 19, 2025•2 citations

Classification of quantum correlations via quantum-inspired machine learning

Giuseppe Sergioli, Andrés Camilo Granda Arango, Carlo Cuccu +3 more

Abstract Quantum information theory, and in particular, the theory of quantum state discrimination, has enabled the development of a supervised multi-class classification algorithm. Inspired by the Pretty Good Measurement (PGM), a quantum-inspired classifier named the PGM Classifier has been designed, capable of classifying among multiple classes without resorting to the One versus One or One versus Rest strategy. In the recent past, several approaches of quantum machine learning have been introduced to classify entangled and separable states. In this article, we apply the PGM Classifier to discriminate among factorized states, separable states, and entangled states. We analyze from the simplest case ( 2 qubits) up to a system of ( 5 qubits) and offer a comparison between our quantum-inspired classifier and other classical classifiers. Finally, we propose a refined classification of non-locality based on the violation of the Mermin and Svetlichny inequalities.

Quantum Machine Intelligence•Nov 25, 2025•2 citations

A quantum-inspired classification for random mixed states

Giuseppe Sergioli, Carlo Cuccu, Carla Rieger +4 more

We present a quantum-inspired classification framework designed to identify correlation structures: product, separable, and entangled, in random mixed quantum states. Building on previous work where the Pretty-Good-Measurement (PGM) classifier demonstrated a competitive performance on pure-state ensembles, we extend this method to the more challenging domain of mixed states. We apply our quantum-inspired classifier to randomly generated ensembles of two- and three-qubit mixed states, encompassing all possible varieties of subsystem correlations while ensuring statistical neutrality. The results indicate that learning architectures inspired by quantum state discrimination can offer scalable and physically grounded tools for the characterization of entanglement and separability even in the mixed-state regime.

Scientific Reports•Mar 30, 2026•0 citations

Robust Pretty Good Measurement via Hybrid Classical-Quantum Pseudoinverse Approximation and Circuit-Level Realization

Bikash K. Behera, Andrés Camilo Granda Arango, Giuseppe Sergioli +1 more

Pretty Good Measurement (PGM) is a near-optimal strategy for quantum state discrimination, but its practical realization becomes unstable when the ensemble operator is singular or ill-conditioned. We introduce a numerically robust PGM formulation based on the Moore-Penrose pseudoinverse, replacing the standard inverse square root with a threshold-regularized variant that remains well-defined across different spectral regimes. We develop a hybrid classical-quantum framework that combines pseudoinverse-based spectral preprocessing with quantum circuit realizations using block-encoding and spectral-transformation techniques. The framework incorporates support awareness, yielding physically meaningful measurement operators even in rank-deficient cases, and employs oblivious amplitude amplification to improve circuit-level success probabilities. Extensive numerical and circuit-level simulations show close agreement between theoretical predictions and quantum circuit outputs. Experiments on synthetic and real datasets, including ill-conditioned and degenerate scenarios, demonstrate stable discrimination performance where standard PGM becomes numerically unstable. The results establish a practical hybrid classical-quantum framework for robust quantum state discrimination and extend previous circuit-based implementations of the PGM testing stage toward pseudoinverse-aware measurement design.

quant-pharXiv:2606.13150•Jun 12, 2026•0 citations

Enhancing mpMRI-Based Prostate Cancer Detection by Ensemble Quantum Machine Learning Models

O. O. Solanke, K. K. A. Abdullah, S. O. Hassan +1 more

Prostate cancer is among the most prevalent malignancies in men, and early, accurate diagnosis is critical for effective treatment. Traditional machine learning techniques have demonstrated success in analyzing multiparametric magnetic resonance imaging (MRI) and clinical biomarkers; however, their scalability and capacity to model complex feature interactions remain limited. This study proposes an Ensemble Quantum Machine Learning (QML) framework to enhance prostate cancer detection using the PROSTATEx Challenge dataset. Radiomic features were extracted from MRI modalities and clinical attributes, then standardized and reduced using principal component analysis (PCA) to match current quantum hardware constraints. Three quantum classifiers—Quantum Support Vector Machine (QSVM), Variational Quantum Classifier (VQC), and Quantum Neural Network (QNN) - were independently trained and integrated using both soft voting and stacked ensemble strategies. Results from stratified 5-fold cross-validation show that the stacked ensemble outperformed individual models and baseline classifiers, achieving an average accuracy of 88.4%, recall of 89.1%, precision of 87.9%, F1-score of 88.5%, and area under the curve - receiver operating characteristic (AUC-ROC) of 0.94 (stack ensemble). These findings highlight the potential of hybrid quantum-classical ensemble learning to improve diagnostic robustness, particularly in reducing false negatives. Furthermore, validation on real quantum hardware demonstrated consistent performance, underscoring the feasibility of QML in near-term medical applications. This work contributes to the growing intersection of quantum computing and clinical AI, offering a scalable and interpretable approach to precision oncology.

Nigerian Journal of Physics•Jul 3, 2025•0 citations

Quantum Kernel-based Machine Learning for Multi-Omic Stratification of Lung Cancer Subtypes

Jenolin J, Dhanasegar Narmadha, G. Naveen Sundar

Subtyping of lung cancer is crucial for individualized treatment and optimizing outcomes. Computational modeling allows data analysis over large scales, which helps in making precision classification in large datasets or high-dimensional datasets. Conventional machine learning methods perform well, but integration of multi-omic data (DNA, RNA, miRNA) is challenging in cancer subtyping because of high dimensionality and complex inter-omic relationships. To bypass this Quantum machine learning (QML) provides better ability to navigate high-dimensional biological data with many variables using quantum computational concepts that can identify fine patterns or small-scale patterns beyond the capability of classical approaches. Through the integration of quantum computing principles with the machine learning framework, the study highlights the improvement of computer technology. The framework shows how the system is optimized with the help of quantum principles, how it can improve accuracy, parallel processing and reducing latency. In this work, we have combined data on miRNA, RNA expression, and DNA methylation to create a complete workflow for classifying cancer. This method uses a wide variety of feature selection techniques, such as random forest-based selection, chi-square, mutual information(MI), and principal component analysis(PCA).In order to illustrate the notion of quantum advantage, we carefully compare the performance metrics of the Quantum Support Vector Classifier (QSVC) to the classical Support Vector Machine (SVM), using a ZZ feature map quantum kernel as our baseline for analysis. Using five-fold crossvalidation, we have also verified that the quantum has an advantage with consistently lower variance and higher accuracy.

International Conference on Evolutionary Computing and Mobile Sustainable Networks (ICECMSN)•Nov 24, 2025•1 citation
We investigate a quantum-inspired approach to supervised multi-class classification based on the Pretty Good Measurement (PGM), viewed as an operator-valued decision rule derived from quantum state discrimination. The method associates each class with an encoded mixed state and performs classification through a single POVM construction, thus providing a genuinely multi-class strategy without reduction to pairwise or one-vs-rest schemes. In this perspective, classification is reformulated as the discrimination of a finite ensemble of class-dependent density operators, with performance governed by the geometry induced by the encoding map and by the overlap structure among classes.
To assess the practical scope of this framework, we apply the PGM-based classifier to two biomedical radiomics case studies: histopathological subtyping of non-small-cell lung carcinoma (NSCLC) and prostate cancer (PCa) risk stratification. The evaluation is conducted under protocols aligned with previously reported radiomics studies, enabling direct comparison with established classical baselines. The results show that the PGM-based classifier is consistently competitive and, in several settings, improves upon standard methods. In particular, the method performs especially well in the NSCLC binary and three-class tasks, while remaining competitive in the four-class case, where increased class overlap yields a more demanding discrimination geometry. In the PCa study, the PGM classifier remains close to the strongest ensemble baseline and exhibits clinically relevant sensitivity--specificity trade-offs across feature-selection scenarios.
Quantum-inspired machine learning (QiML) employs mathematical principles from quantum theory, such as Hilbert-space representations and quantum state discrimination, to enhance classical learning algorithms. In this work, we investigate the integration of Quantum Synthetic Minority Oversampling Technique (QSMOTE) variants with two quantum-inspired classifiers: the Pretty Good Measurement (PGM) classifier and the kernelized Pretty Good Measurement (KPGM) classifier. We propose and analyze three QSMOTE variants, namely KNN-based, Fidelity-based, and Margin-based QSMOTE, designed to improve minority-class representation in imbalanced datasets through quantum-inspired similarity and sampling mechanisms. A unified theoretical and empirical comparison of PGM and KPGM is presented under amplitude and stereo encoding strategies with multiple quantum copies. Experimental evaluations on the Telco Customer Churn dataset demonstrate that the proposed quantum-inspired approaches consistently outperform a classical Random Forest baseline, particularly in terms of recall and balanced F1-score. Among all configurations, PGM with stereo encoding and n_{copies}=2 achieves the best performance with an accuracy of 0.8512 and an F1-score of 0.8234, while KPGM exhibits competitive and more stable behavior across different QSMOTE variants, reaching accuracies of 0.8511 under stereo encoding and 0.8483 under amplitude encoding. The results further show that increasing the number of quantum copies systematically improves classification performance, especially for minority-class detection. This work highlights the effectiveness of combining quantum-inspired oversampling and classification strategies for imbalanced learning, while providing practical insights into the complementary strengths of measurement-based and kernel-based quantum-inspired machine learning frameworks.
Abstract Quantum information theory, and in particular, the theory of quantum state discrimination, has enabled the development of a supervised multi-class classification algorithm. Inspired by the Pretty Good Measurement (PGM), a quantum-inspired classifier named the PGM Classifier has been designed, capable of classifying among multiple classes without resorting to the One versus One or One versus Rest strategy. In the recent past, several approaches of quantum machine learning have been introduced to classify entangled and separable states. In this article, we apply the PGM Classifier to discriminate among factorized states, separable states, and entangled states. We analyze from the simplest case ( 2 qubits) up to a system of ( 5 qubits) and offer a comparison between our quantum-inspired classifier and other classical classifiers. Finally, we propose a refined classification of non-locality based on the violation of the Mermin and Svetlichny inequalities.
We present a quantum-inspired classification framework designed to identify correlation structures: product, separable, and entangled, in random mixed quantum states. Building on previous work where the Pretty-Good-Measurement (PGM) classifier demonstrated a competitive performance on pure-state ensembles, we extend this method to the more challenging domain of mixed states. We apply our quantum-inspired classifier to randomly generated ensembles of two- and three-qubit mixed states, encompassing all possible varieties of subsystem correlations while ensuring statistical neutrality. The results indicate that learning architectures inspired by quantum state discrimination can offer scalable and physically grounded tools for the characterization of entanglement and separability even in the mixed-state regime.
Pretty Good Measurement (PGM) is a near-optimal strategy for quantum state discrimination, but its practical realization becomes unstable when the ensemble operator is singular or ill-conditioned. We introduce a numerically robust PGM formulation based on the Moore-Penrose pseudoinverse, replacing the standard inverse square root with a threshold-regularized variant that remains well-defined across different spectral regimes. We develop a hybrid classical-quantum framework that combines pseudoinverse-based spectral preprocessing with quantum circuit realizations using block-encoding and spectral-transformation techniques. The framework incorporates support awareness, yielding physically meaningful measurement operators even in rank-deficient cases, and employs oblivious amplitude amplification to improve circuit-level success probabilities. Extensive numerical and circuit-level simulations show close agreement between theoretical predictions and quantum circuit outputs. Experiments on synthetic and real datasets, including ill-conditioned and degenerate scenarios, demonstrate stable discrimination performance where standard PGM becomes numerically unstable. The results establish a practical hybrid classical-quantum framework for robust quantum state discrimination and extend previous circuit-based implementations of the PGM testing stage toward pseudoinverse-aware measurement design.
Prostate cancer is among the most prevalent malignancies in men, and early, accurate diagnosis is critical for effective treatment. Traditional machine learning techniques have demonstrated success in analyzing multiparametric magnetic resonance imaging (MRI) and clinical biomarkers; however, their scalability and capacity to model complex feature interactions remain limited. This study proposes an Ensemble Quantum Machine Learning (QML) framework to enhance prostate cancer detection using the PROSTATEx Challenge dataset. Radiomic features were extracted from MRI modalities and clinical attributes, then standardized and reduced using principal component analysis (PCA) to match current quantum hardware constraints. Three quantum classifiers—Quantum Support Vector Machine (QSVM), Variational Quantum Classifier (VQC), and Quantum Neural Network (QNN) - were independently trained and integrated using both soft voting and stacked ensemble strategies. Results from stratified 5-fold cross-validation show that the stacked ensemble outperformed individual models and baseline classifiers, achieving an average accuracy of 88.4%, recall of 89.1%, precision of 87.9%, F1-score of 88.5%, and area under the curve - receiver operating characteristic (AUC-ROC) of 0.94 (stack ensemble). These findings highlight the potential of hybrid quantum-classical ensemble learning to improve diagnostic robustness, particularly in reducing false negatives. Furthermore, validation on real quantum hardware demonstrated consistent performance, underscoring the feasibility of QML in near-term medical applications. This work contributes to the growing intersection of quantum computing and clinical AI, offering a scalable and interpretable approach to precision oncology.
Subtyping of lung cancer is crucial for individualized treatment and optimizing outcomes. Computational modeling allows data analysis over large scales, which helps in making precision classification in large datasets or high-dimensional datasets. Conventional machine learning methods perform well, but integration of multi-omic data (DNA, RNA, miRNA) is challenging in cancer subtyping because of high dimensionality and complex inter-omic relationships. To bypass this Quantum machine learning (QML) provides better ability to navigate high-dimensional biological data with many variables using quantum computational concepts that can identify fine patterns or small-scale patterns beyond the capability of classical approaches. Through the integration of quantum computing principles with the machine learning framework, the study highlights the improvement of computer technology. The framework shows how the system is optimized with the help of quantum principles, how it can improve accuracy, parallel processing and reducing latency. In this work, we have combined data on miRNA, RNA expression, and DNA methylation to create a complete workflow for classifying cancer. This method uses a wide variety of feature selection techniques, such as random forest-based selection, chi-square, mutual information(MI), and principal component analysis(PCA).In order to illustrate the notion of quantum advantage, we carefully compare the performance metrics of the Quantum Support Vector Classifier (QSVC) to the classical Support Vector Machine (SVM), using a ZZ feature map quantum kernel as our baseline for analysis. Using five-fold crossvalidation, we have also verified that the quantum has an advantage with consistently lower variance and higher accuracy.