Comprehensive Summary
This study, conducted by Chang et al., examines the current state of AI in spinal imaging through a narrative review of articles published in the last 3 years. Various studies have developed and tested techniques that span the whole course of spinal surgery. First, through Deep Learning Reconstruction, image acquisition could be accelerated. Based on the obtained spine images, segmentation of spine anatomy can be automated. This information can be inputted into AI models to plan surgical interventions. After the surgery, imaging data could be used to predict the surgical outcomes through machine learning algorithms. There have been various improvements in AI in spinal surgery. Nevertheless, the effectiveness of AI techniques still needs to be validated with clinical outcomes.
Outcomes and Implications
Image recognition and analysis is a strong field for AI applications. With spinal surgery requiring extensive use of imaging, the application of AI would enhance patient care and decrease the demand for further resources. Currently, the advancements are promising, with Magnetic Resonance and Computed Tomography spinal imagining being well established. Nevertheless, emerging uses such as image analysis and surgery planning technology requires continuous improvements and testing.