A new quantum-enhanced approach to AI-driven medical imaging system

Medical Imaging Systems (MIS) play a crucial role in modern medicine by providing accurate diagnostic and treatment capabilities. These systems use various physical processes to create images inside the human body for healthcare professionals to identify and address medical conditions. There is a growing interest in integrating artificial intelligence (AI) in medicine from various sources recently. Presently, with improved algorithms and more significant availability of training data, AI can help or even replace some of the tasks that were being performed by medical professionals. Typically, most MIS performance enhancements are achieved by leveraging transistor-based technologies. However, such implementations showcase certain disadvantages: for instance, slow processing speeds, high power consumption, large physical footprints, and restricted switching frequencies, especially in the GHz range. This could limit the effective performance and efficiency of MIS. Quantum computing, in turn, today appears as an alternative, at least for fully digital circuits in MIS; QCA provides advantages related to higher intrinsic switching speeds (up to terahertz) compared with transistor-based technologies, along with an improved throughput owing to its inherent compatibility with pipelining. QCA also has minimum power consumption and a smaller area of circuitry, which makes it amply suitable for establishing frameworks in circuit design for AI applications. The performance requirement in AI is real-time with minimum energy consumption and minimum cost. The ALU, in this regard, forms the basis for processing and computation units within processor systems. The method presented in this work benefits from the merits of QCA for an ALU design featuring low complexity, high performance, minimum power consumption, maximum speed, and reduced area. This approach has been able to successfully integrate the design of adders and multiplexers with that of basic gates to reduce latency and energy consumption with the aim of improving AI in MIS. The development and simulation of the proposed designs are carefully carried out using QCADesigner 2.0.03 software. A comparison of the different structures proposed shows significant improvements in complexity vs. cell count vs. power consumption compared to earlier designs, hence promising quantum computing for the MIS capability development.