Optimization of Computed Tomography Data Acquisition by Means of Quantum Computing

Abstract

Quantum Computing (QC) technology has made tremendous progress recently. Today, first real QC systems are operational. The Fraunhofer Institute of Integrated Circuits IIS, Division Development Center X-ray Technology is participating in the project BayQS, which aims at identifying and evaluating potential applications of quantum computing in the field of non-destructive testing by means of X-ray imaging. Computed tomography (CT) is a well-established method of 3-D imaging in medicine and industry since several decades. The quality of the resulting volumetric datasets depends essentially on a set of around 20 physical parameters controlling the set-up of X-ray source, X-ray digital detector array, movable mechanical components, sample rotation, and image processing. Hence, we started investigating the potential use of a quantum-based optimization method to optimize CT data acquisition. As a first approach, we focused on the task to minimize the number of projection images required to achieve unaltered image quality when compared to an acquisition over the full 4π solid angle. Minimizing involves a cost function. There are several ways to define this cost function. In order to transfer the task to a QC system, the problem is reformulated as a Quadratic Unconstrained Binary Optimization (QUBO). A decisive feature of the QUBO is, that is can be solved by a hybrid implementation on quantum devices, which combines a classical part to iteratively find the maximum of the cost function and a quantum circuit to evaluate the respective cost function for each combination of projections. This procedure is called quantum approximate optimization algorithm (QAOA). The experiments were performed via the qiskit script language on an IBM Quantum System One located in Ehingen, Germany. This QC system offers 27 qubits and enables first real world experiments with small test datasets.