Quantum annealing-based computed tomography using variational approach for a real-number image reconstruction.

Area of Science:

• Medical Imaging
• Quantum Computing
• Computational Science

Background:

• Quantum computing advancements offer potential for medical imaging, but limited qubits pose challenges.
• Computed tomography (CT) reconstruction is computationally intensive, driving the need for novel algorithms.

Purpose of the Study:

• To investigate the feasibility of quantum annealing-based computed tomography (QACT) with current qubit limitations.
• To develop and evaluate a novel variational approach for CT image reconstruction using quantum annealing.

Main Methods:

• The QACT algorithm was adapted to solve quadratic unconstrained binary optimization problems.
• A variational method was employed to approximate real numbers for image reconstruction.
• Image reconstruction was tested on sizes from 4x4 to 24x24 pixels, analyzing projection data quantity and noise impact.
• QACT results were compared against Maximum Likelihood Expectation Maximization (MLEM) and Filtered Back Projection (FBP) algorithms.

Main Results:

• Accurate CT image reconstruction was achieved using QACT with only two qubits per pixel and sufficient projection data.
• QACT demonstrated superior image quality compared to MLEM and FBP under conditions of abundant projections and low noise.
• In scenarios with limited projection data and higher noise, QACT's image quality was inferior to MLEM.

Conclusions:

• The developed QACT algorithm, utilizing a variational approach, enables real-number CT image reconstruction.
• The study confirms that two qubits per pixel are sufficient for accurate representation in QACT.
• QACT shows promise for CT reconstruction but requires further optimization for noisy and data-limited conditions.