Available/Emerging Quantum Advantages for Medical Images/Radiology PDF 4/15/23.
Imaging Data accounts for 90% of all Healthcare Data, with over 5.5 Billion global imaging procedures per year. Medical images
are composed of individual elements, known as pixels, each with discrete quantities of numeric representation. Machine learning
processes these pixels with central processing units (CPUs) and/or graphics processing units (GPUs) for analysis.
Quantum advantages have been observed regarding existing Human Brain/Chest MR/CT/Other Images in at least 7 articles.
These models generally consist of a conventional Convolutional Neural Network (CNN) alongside a smaller quantum
circuit/algorithm run on a simulator for "state preparation" and "measurement" tasks to improve the CNN overall classification
accuracy by up to 5%.
In a quantum environment, medical images can express using quantum bits, to prevent problems associated with classical
computing by applying a CNN framework directly to a quantum computer. Seunghyeok Oh, et al. In addition Iris Cong, et. al.
published that "The extreme complexity of many-body states often makes theoretical analysis intractable”, with conventional
methods.