Evolution and Current Clinical Applications of Subtraction Techniques in CTA

Subtraction techniques are important postprocessing tools in CT angiography (CTA) and have evolved from conventional temporal subtraction to multiple technical pathways, including energy imaging subtraction, artificial intelligence (AI)-based subtraction, and contrast enhancement boost (CE-Boost). Temporal subtraction removes bone through pixel-level subtraction between non-contrast and contrast-enhanced images; however, it is susceptible to misregistration artifacts. Energy imaging subtraction achieves bone removal by exploiting the attenuation differences between virtual monoenergetic and virtual noncontrast images and reduces radiation dose. AI-based subtraction relies on convolutional neural networks to automatically segment bone structures, yielding higher bone removal accuracy in anatomically complex regions. CE-Boost fuses an iodine map with a contrast-enhanced image, preserving the anatomical background while improving vascular contrast. This paper systematically reviews the evolution of subtraction technology, compares the principles, advantages, and disadvantages of the four types of technologies, and summarizes their value in clinical applications. It also analyzes the current challenges and future development trends. This study aimed to provide additional references for the standardized application of subtraction technology and the realization of its clinical value.