The development of effective deep learning systems for cardiac disease diagnosis requires extensive and diverse electrocardiogram (ECG) databases, which remain scarce due to patient confidentiality regulations and costly medical annotation procedures. This paper introduces a novel hybrid Quantum Generative Adversarial Network (QGAN) that generates realistic 12-lead ECG signals trained on the MIT-BIH Arrhythmia Database. A parameterised quantum circuit (PQC) serves as the generative core, exploiting quantum superposition and entanglement to model biomedical signal distributions that traditional generators cannot efficiently replicate. The system pairs a PennyLane-based quantum generator with a 1D Convolutional Neural Network discriminator implemented in PyTorch, trained via the parameter-shift rule. Experimental results demonstrate that the Quantum PQC achieves a Fréchet Distance of 12.3 ±2.1, outperforming Classical CNN (15.8 ±3.2) and Classical LSTM (18.4 ±4.1) baselines while using fewer parameters. Generated 12-lead traces accurately reproduce P-waves, QRS complexes, T-waves, and inter-lead spatial relationships, validating quantum-augmented synthesis as a viable, privacy-preserving approach for cardiovascular AI training.
The development of effective deep learning systems for cardiac disease diagnosis requires extensive and diverse electrocardiogram (ECG) databases, which remain scarce due to patient confidentiality regulations and costly medical annotation procedures. This paper introduces a novel hybrid Quantum Generative Adversarial Network (QGAN) that generates realistic 12-lead ECG signals trained on the MIT-BIH Arrhythmia Database. A parameterised quantum circuit (PQC) serves as the generative core, exploiting quantum superposition and entanglement to model biomedical signal distributions that traditional generators cannot efficiently replicate. The system pairs a PennyLane-based quantum generator with a 1D Convolutional Neural Network discriminator implemented in PyTorch, trained via the parameter-shift rule. Experimental results demonstrate that the Quantum PQC achieves a Fréchet Distance of 12.3 ±2.1, outperforming Classical CNN (15.8 ±3.2) and Classical LSTM (18.4 ±4.1) baselines while using fewer parameters. Generated 12-lead traces accurately reproduce P-waves, QRS complexes, T-waves, and inter-lead spatial relationships, validating quantum-augmented synthesis as a viable, privacy-preserving approach for cardiovascular AI training.
The class imbalance of electrocardiogram (ECG) data is a serious impediment to the development of diagnostic systems for heart disease. To address this issue, this paper proposes HQ-DCGAN, a hybrid quantum deep convolutional generative adversarial network, specifically designed for the generation of ECGs. The proposed algorithm employs different quantum convolutional layers for the generator and discriminator as feature extractors and utilizes parameterized quantum circuits (PQCs) to enhance computational capabilities, along with the model-feature mapping process. Moreover, this algorithm preserves the nonlinearity and scalability inherent to classical convolutional neural networks (CNNs), thereby optimizing the utilization of quantum resources, and ensuring compatibility with contemporary quantum devices. In addition, this paper proposes a novel evaluation metric, 1D Fréchet Inception Distance (1DFID), to assess the quality of the generated ECG signals. Simulation experiments show that HQ-DCGAN exhibits strong performance in ECG signal generation. Furthermore, the generated signals achieve an average classification accuracy of 82.2%, outperforming the baseline algorithms. It has been experimentally proven that HQ-DCGAN is friendly to currently noisy intermediate-scale quantum (NISQ) computers, in terms of both number of qubits and circuit depths, while improving the stability.
Although cardiovascular disease constitutes the primary cause of death in China, the clinical deployment of artificial intelligence-assisted diagnostic tools remains constrained by the limited availability of large-scale, high-fidelity electrocardiogram datasets with adequate privacy safeguards. Conventional generative adversarial networks (GANs) encounter persistent constraints regarding temporal sequence representation, elevated-dimensional attribute extraction, and optimization stability, concurrently failing to accommodate resource limitations inherent to contemporary noisy intermediate-scale quantum (NISQ) apparatus. Drawing inspiration from ensemble methodologies typified by random forests, this investigation introduces an ECG generative information architecture designated QRFGAN-ECG, predicated upon quantum random forest generative adversarial networks. The generative component implements quantum residual recurrent networks (QResRNN) to encapsulate extended-range cardiac cycle dependencies; conversely, the discriminative module incorporates quantum random forest convolutional networks (QRF-CNN) to execute multi-resolution attribute voting, thereby augmenting discrimination robustness. Variational quantum circuit architectures formulated through matrix product states (MPS) and tree tensor networks (TTN) substantially diminish qubit requirements and circuit depth. Furthermore, this study pioneerly integrates quantum dropout mechanisms with forest ensemble strategies to mitigate quantum overfitting phenomena. Empirical evaluations conducted across four publicly accessible ECG repositories demonstrate QRFGAN-ECG superiority over classical Against BiGRU-CNN, ECG-QGAN, and DCGAN baselines, the proposed framework demonstrates superior performance across Pearson correlation coefficient, percentage root-mean-square difference, root-mean-square error, and downstream classification accuracy. Adversarial training convergence improves by roughly one-third, with markedly enhanced robustness to four dominant quantum noise regimes. The generated ECG samples carry intrinsic labels, eliminating manual annotation overhead while precluding privacy exposure. These synthetic specimens readily support training set expansion, establishing a dependable data foundation for cardiac disease diagnostic assistance systems.
To address the scarcity and class imbalance of abnormal electrocardiogram (ECG) databases, which are crucial in AI-driven diagnostic tools for potential cardiovascular disease detection, this study proposes a novel quantum conditional generative adversarial algorithm (QCGAN-ECG) for generating abnormal ECG signals. The QCGAN-ECG constructs a quantum generator based on patch method. In this method, each sub-generator generates distinct features of abnormal heartbeats in different segments. This patch-based generative algorithm conserves quantum resources and makes QCGAN-ECG practical for near-term quantum devices. Additionally, QCGAN-ECG introduces quantum registers as control conditions. It encodes information about the types and probability distributions of abnormal heartbeats into quantum registers, rendering the entire generative process controllable. Simulation experiments on Pennylane demonstrated that the QCGAN-ECG could generate completely abnormal heartbeats with an average accuracy of 88.8%. Moreover, the QCGAN-ECG can accurately fit the probability distribution of various abnormal ECG data. In the anti-noise experiments, the QCGAN-ECG showcased outstanding robustness across various levels of quantum noise interference. These results demonstrate the effectiveness and potential applicability of the QCGAN-ECG for generating abnormal ECG signals, which will further promote the development of AI-driven cardiac disease diagnosis systems. The source code is available at github.com/VanSWK/QCGAN_ECG.
The need for synthetic medical data is growing rapidly as machine learning becomes more important in healthcare, especially with strict privacy rules like HIPAA in the United States and GDPR in Europe. While Generative Adversarial Networks (GANs) have shown potential in generating realistic synthetic data, they often face problems such as mode collapse, unstable training, and difficulty handling complex, high-dimensional medical datasets.In this work, we introduce a new Quantum Generative Adversarial Network (QUGAN) that combines the strengths of quantum and classical computing. The generator is built using a Variational Quantum Circuit (VQC), while the discriminator remains a classical neural network. To our knowledge, this is the first time a QUGAN has been applied to structured tabular healthcare data and especially using a VQC circuit in the generator, which makes up most real electronic health records but is rarely explored in quantum machine learning. The model was developed step by step, starting from a simple GAN baseline, and gradually improved into a more powerful hybrid system.The QUGAN performed considerably better compared with the standard baseline when tested on actual cardiovascular dataset. While achieving synthetic precision scores of up to 92%, it decreased the Maximum Mean Discrepancy (MMD) by 50.6%, the Kullback–Leibler (KL) divergence by 65.6%, and the Jensen–Shannon (JS) divergence by 67.7%. These results indicate that the model creates diverse samples close to real data, which are a key factor for safe data sharing and dependable analysis later. In short, our findings suggest that QUGANs that uses VQC circuits can be reliable method to produce tabular medical datasets.
Cardiovascular diseases (CVD) remain a leading cause of morbidity and mortality worldwide, with arrhythmia being one of their common manifestations. Traditional diagnostic methods, such as electrocardiography (ECG), rely heavily on expert interpretation for detecting irregular heart rhythms, which can lead to inaccuracies and delays in diagnosis, particularly in resource-limited settings. In recent years, machine learning (ML) has emerged as a powerful tool for improving the accuracy and efficiency of medical diagnostics. Although deep learning models have demonstrated remarkable success in various domains, their computational complexity poses challenges for real-time applications. This study investigates various quantum-classical hybrid algorithms for improving ECG-based arrhythmia classification. In this study, we employed three hybrid quantum-classical algorithms for arrhythmia detection, estimator-based Quantum Neural Network (QNN), sampler-based QNN, and Quantum Support Vector Classifier (QSVC), to enhance the accuracy of automated arrhythmia classification. The models are evaluated using the MIT-BIH arrhythmia database. The experimental results demonstrated that these three quantum models outperform traditional Recurrent Neural Networks (RNNs) and their variants in several metrics, including accuracy, sensitivity, and specificity. The QSVC (k = 16) model achieved the highest performance, with an accuracy of 97.7%, F1-score of 97.5%, and area under the receiver operating characteristic curve of 99.4%. Additionally, we developed an interactive application using StreamLit for real-time arrhythmia risk prediction and visual feedback. Overall, the proposed quantum ML model demonstrates significant potential for arrhythmia detection with robust and highly accurate performance, providing a solid foundation for future clinical deployment and integration into wearable devices.
Cardiac arrhythmia is a leading cause of sudden cardiac death, often occurring without warning and requiring early, accurate detection for effective intervention. However, their intermittent nature and reliance on expert ECG interpretation make timely diagnosis difficult, especially at scale. This study proposes a hybrid quantum-classical deep learning algorithm for automated arrhythmia detection using ECG images. The algorithm integrates a pre-trained ResNet50 for feature extraction with a 10-qubit variational quantum circuit, enhancing classification through quantum expressiveness. A dataset of 123,998 images, derived from MIT-BIH and PTB ECG recordings, was used to train the algorithm. Each image was resized to 224×224 pixels, and 10-dimensional ResNet50 features were processed by a quantum layer via PennyLane’s lightning.kokkos simulator. The algorithm achieved 99.98% test accuracy. Training was completed in an efficient manner, demonstrating both high performance and time-conscious execution, positioning quantum-enhanced algorithms as promising tools in real-time diagnostic cardiology.