We apply sonification strategies and quantum computing to the analysis of an episode of seizure. We first sonify the signal from a selection of channels (from real ECoG data), obtaining a polyphonic sequence. Then, we propose two quantum approaches to simulate a similar episode of seizure, and we sonify the results. The comparison of sonifications can give hints on similarities and discrepancies between real data and simulations, helping refine the \textit{in silico} model. This is a pioneering approach, showing how the combination of quantum computing and sonification can broaden the perspective of real-data investigation, and helping define a new test bench for analysis and prediction of seizures.
Comments
Presented at ISQCMC '25: 3rd International Symposium on Quantum Computing and Musical Creativity
We apply sonification strategies and quantum computing to the analysis of an episode of seizure. We first sonify the signal from a selection of channels (from real ECoG data), obtaining a polyphonic sequence. Then, we propose two quantum approaches to simulate a similar episode of seizure, and we sonify the results. The comparison of sonifications can give hints on similarities and discrepancies between real data and simulations, helping refine the \textit{in silico} model. This is a pioneering approach, showing how the combination of quantum computing and sonification can broaden the perspective of real-data investigation, and helping define a new test bench for analysis and prediction of seizures.
Epilepsy, a neurological disorder characterized by recurrent and unpredictable seizures, demands timely and accurate detection for effective management and patient safety. While traditional machine learning approaches have shown promise in seizure detection, this study explores the potential of quantum computing, specifically utilizing a Variational Quantum Classifier. The Variational Quantum Classifier, grounded in quantum mechanics, aims to enhance computational capabilities for seizure detection. Leveraging quantum computing’s intrinsic parallelism and superposition properties, the model focuses on processing complex and high-dimensional Electroencephalogram signals commonly used in epilepsy diagnosis. The model undergoes training on a dataset containing both seizure and non-seizure instances, optimizing parameters for increased accuracy. Performance evaluations utilize metrics such as sensitivity, specificity, and overall classification accuracy.
Abstract Quantum mechanics poses significant challenges for audio-visual representation, particularly in the context of quantum entanglement. Sonification—the auditory representation of data—offers a promising complementary approach to conventional visualization techniques. This paper investigates sonification methods applied to the dynamical generation of entanglement in many-qubit systems, combining phase-space representations with entanglement measures. We study the dynamics of entanglement generation in many-qubit systems under a dynamical protocol governed by two models: the one-axis twisting model and a quantum-kicked rotor exhibiting both regular and quantum chaotic behavior. We present a procedure of entanglement dynamics sonification, allowing mapping of the phase-space representation of a many-qubit quantum state and von Neumann entanglement entropy to sound. The methodology relies on the mapping of data features to amplitude, stereo spatialization, pitch, and timbral characteristics. Results illustrate how sonification translates entanglement dynamics into distinctive, model-specific sound textures, offering a route toward artistic and exploratory engagement with quantum correlations. Potential benefits for intuitive understanding remain to be tested.
Epileptic seizures are significant challenges in a neurological environment primarily due to the non-stationary and complex nature of the electroencephalogram (EEG) signals. In this paper, we present a Hybrid Quantum-Classical Neural Framework (HQCNF) that leverages quantum computing to support taxonomy of epilepsy seizures with deep neural learning. The framework applies a Continuous Wavelet Transform (CWT) to convert EEG recordings into a time-frequency representation deploying scalograms intended to rigorously scrutinize the features describing oscillations and other elements noted when seizures occur. The model presented in this paper is built from classical architecture and uses quantum-inspired neural layers to support providing the atomic feature representations in order to make sense of discriminability behavior and other elements of interpretable behavior in epistemology and learning. The HQCNF model achieves 99% classification accuracy and provides evidence of enhanced generalization, and outperforms typical deep learning models. The research supports the methodology of hybrid quantum-classical paradigms to move beyond a conventional computing biomedical signal analysis restriction, and the ability to examine infrastructure presented by HQCNF moves us toward real-time assessments and promotes further examination of efforts towards intelligent diagnostic methodologies utilized in neurological disorder management frameworks. The work focuses on the value of quantum-enhanced learning as it relates to EEG-based examination in epilepsy management, and makes advanced hybrid-learning implementations of quantum-enhanced learning that provides for more efficient and reliable accuracy in EEG-based examination of epilepsy signal processing.
This research evaluates the performance of classical and quantum machine learning models for epileptic seizure detection using the BONN and DELHI EEG datasets. The data pipeline includes a preprocessing step with feature extraction followed by multiple dimensionality reduction techniques. Seven quantum data encoding strategies (Phase, Angle, Dense Angle, Pauli Feature Map, Efficient SU2, Z Feature Map, and ZZ Feature Map) were evaluated to analyze their impact on the expressivity of quantum circuits and model performance. Classical models, including Decision Tree and Random Forest classifiers, consistently achieved high accuracy on the BONN and Delhi datasets, with performance exceeding 90\% across most dimensionality-reduction settings. Quantum models demonstrated competitive yet generally lower performance than classical baselines; however, certain combinations of encoding methods and dimensionality reduction techniques achieved the most promising results for both datasets. Overall, the study highlights the influence of encoding strategies and dataset characteristics on quantum machine learning performance and provides a comparative analysis that supports future optimization of quantum approaches for EEG-based seizure detection.
Quantum mechanics presents unique hurdles for audio-visual representation, especially in conveying quantum entanglement. Sonification—the translation of data into sound—provides a complementary channel to visualize these inherently abstract phenomena. In this study, we develop and test a sonification framework for tracking dynamical entanglement generation in multi-qubit systems.Using the Husimi-Q function as a phase-space representation alongside the von Neumann entanglement entropy, we map a quantum states structure, generated during the One-Axis Twisting protocol, onto auditory parameters such as pitch, timbre, and spatialization. Our method encodes the instantaneous phasespace distribution and entanglement entropy values into distinct sonic motifs, yielding a real-time acoustic portrayal of entanglement growth and decay.Listening tests and qualitative analyses reveal that sonification accentuates subtle transitions in entanglement dynamics—especially the onset of chaos—more intuitively than visual plots alone. Our approach offers an artistic medium to experience quantum correlations, fostering deeper engagement and insight.
Is it possible to apply some fundamental principles of quantum-computing to time series classi?cation algorithms? This is the initial spark that became the research question I decided to chase at the very beginning of my PhD studies. The idea came accidentally after reading a note on the ability of entanglement to express the correlation between two particles, even far away from each other. The test problem was also at hand because I was investigating on possible algorithms for real time bot detection, a challenging problem at present day, by means of statistical approaches for sequential classi?cation. The quantum inspired algorithm presented in this thesis stemmed as an evolution of the statistical method mentioned above: it is a novel approach to address binary and multinomial classi?cation of an incoming data stream, inspired by the principles of Quantum Computing, in order to ensure the shortest decision time with high accuracy. The proposed approach exploits the analogy between the intrinsic correlation of two or more particles and the dependence of each item in a data stream with the preceding ones. Starting from the a-posteriori probability of each item to belong to a particular class, we can assign a Qubit state representing a combination of the aforesaid probabilities for all available observations of the time series. By leveraging superposition and entanglement on subsequences of growing length, it is possible to devise a measure of membership to each class, thus enabling the system to take a reliable decision when a suf?cient level of con?dence is met. In order to provide an extensive and thorough analysis of the problem, a well-?tting approach for bot detection was replicated on our dataset and later compared with the statistical algorithm to determine the best option. The winner was subsequently examined against the new quantum-inspired proposal, showing the superior capability of the latter in both binary and multinomial classi?cation of data streams. The validation of quantum-inspired approach in a synthetically generated use case, completes the research framework and opens new perspectives in on-the-?y time series classi?cation, that we have just started to explore. Just to name a few ones, the algorithm is currently being tested with encouraging results in predictive maintenance and prognostics for automotive, in collaboration with University of Bradford (UK), and in action recognition from video streams.