Implementing a Hybrid Quantum-Classical Neural Network by Utilizing a Variational Quantum Circuit for Detection of Dementia

Abstract: Magnetic resonance imaging (MRI) is a common technique to scan brains for strokes, tumors, and other abnormalities that cause forms of dementia. However, correctly diagnosing forms of dementia from MRIs is difficult, as nearly 1 in 3 patients with Alzheimer's were misdiagnosed in 2019, an issue neural networks can rectify. Quantum computing applications This proposed novel neural network architecture uses a fully-connected (FC) layer, which reduces the number of features to obtain an expectation value by implementing a variational quantum circuit (VQC). The VQC created in this study utilizes a layer of Hadamard gates, Rotation-Y gates that are parameterized by tanh(intensity) * (pi/2) of a pixel, controlled-not (CNOT) gates, and measurement operators to obtain the expected values. This study found that the proposed hybrid quantum-classical convolutional neural network (QCCNN) provided 97.5% and 95.1% testing and validation accuracies, respectively, which was considerably higher than the classical neural network (CNN) testing and validation accuracies of 91.5% and 89.2%. Additionally, using a testing set of 100 normal and 100 dementia MRI images, the QCCNN detected normal and demented images correctly 95% and 98% of the time, compared to the CNN accuracies of 89% and 91%. With hospitals like Massachusetts General Hospital beginning to adopt machine learning applications for biomedical image detection, this proposed architecture would approve accuracies and potentially save more lives. Furthermore, the proposed architecture is generally flexible, and can be used for transfer-learning tasks, saving time and resources.

• The paper proposes a hybrid Quantum-Classical Convolutional Neural Network (QCCNN) integrating a variational quantum circuit (VQC) with a classical ResNet-18 base for feature extraction and improved dementia detection using MRI scans.
• The developed QCCNN achieved a testing accuracy of 97.5% on an MRI dataset, outperforming a classical CNN which reached 91.5% accuracy, demonstrating enhanced precision and recall.
• Integrating quantum principles via the VQC projects data into a higher-dimensional space, demonstrating a tangible step toward realizing quantum advantage in machine learning for biomedical applications like dementia diagnosis.

Overview of Quantum-Classical Neural Network for Dementia Detection

The paper "Implementing a Hybrid Quantum-Classical Neural Network by Utilizing a Variational Quantum Circuit for Detection of Dementia" presents a novel approach by integrating quantum computing with classical neural networks to enhance the accuracy of dementia detection via MRI scans. The core contribution of this study is the development of a hybrid Quantum-Classical Convolutional Neural Network (QCCNN) that leverages variational quantum circuits (VQC) to improve the classification accuracy over traditional classical neural networks.

Hybrid Neural Network Architecture

The architecture proposed in the paper incorporates a fully-connected classical layer intertwined with a quantum layer driven by a VQC. A ResNet-18 model serves as the base for feature extraction, outputting 512 features. These features are further reduced using a VQC that applies unique quantum operations, leading to enhanced computational efficiency and improved accuracy by leveraging quantum superposition and entanglement.

Dataset and Performance

The study utilized an MRI dataset focused on classifying brains into demented and non-demented categories. The hybrid QCCNN is tested against classical CNN on measures of accuracy, precision, recall, and F1-score. Results indicate that the QCCNN achieved a testing accuracy of 97.5%, substantially higher than the 91.5% accuracy of the classical counterpart. Specifically notable is the QCCNN's enhanced precision and recall rates, demonstrating strengthened robustness in true positive identification and reduced false negatives.

Quantum Mechanics Integration

In detail, the variational quantum circuit exploits quantum entanglement and superposition through gates like the Hadamard, CNOT, and Rotation-Y gates. These operations are foundational to projecting the dataset from a classical feature space into a higher-dimensional quantum space, thus aiding superior classification performance. The paper underscores the significance of this innovative integration, showing it as a tangible step toward realizing quantum advantage in machine learning applications.

Implications and Future Directions

The implications of this research are significant both practically and theoretically. By integrating quantum mechanisms, the hybrid neural network reduces the complexity required for high-accuracy results, thereby demonstrating potential transformative effects in biomedical applications, such as early and accurate dementia diagnosis. On a theoretical level, this study contributes to the body of research on quantum machine learning, illustrating additional pathways for applying VQCs in other binary classification tasks.

Moving forward, anticipated future research could explore scaling these hybrid networks further, integrating additional quantum-inspired layers, and tuning hyperparameters to capitalize on the limited qubits of NISQ-era systems. With the progressive development of quantum hardware, more comprehensive real-world implementations and transfer learning opportunities could amplify the impact of such hybrid models.

In conclusion, the paper presents a compelling case for the utility of quantum-classical models in medical imaging and beyond. Future advancements in quantum computing resources will likely unlock even broader applications and further elevate the performance and efficiency of AI technologies.