AI software algorithm may provide accurate measurements of spinopelvic parameters
Use of an artificial intelligence software algorithm may provide measurements of spinopelvic parameters on lateral lumbar radiographs that are of comparable accuracy to surgeons, according to published results.
Samuel K. Cho, MD, chief of spine surgery at Mount Sinai West, and colleagues sequentially collected images from 816 patients receiving lateral lumbar radiographs and used the images to develop a convolutional neural network segmentation algorithm. used 80% of the radiographs to train and validate the convolutional neural network, which combined with a computer vision algorithm to create a pipeline for the fully automated measurement of spinopelvic parameters from lateral lumbar radiographs. Researchers used the remaining 20% of radiographs to test the pipeline, of which 40 radiographs were selected and manually measured by three surgeons for comparison.
Results showed the convolutional neural network had an area under the receiver operating curve of 0.956. found no significant differences algorithm measurements of L1-S1 Cobb angle, pelvic incidence, pelvic tilt and sacral slope. The algorithm also had similar mean absolute difference to spine surgeons for L1-S1 Cobb angle, pelvic tilt, pelvic incidence and sacral slope gold standard measurement, according to results.
Incorporating AI software that can automatically measure radiographs can make the workflow of a busy office more efficient so clinicians can spend more time with patients. Cho added AI software can also help surgeons in the operating room by automatically performing measurements on intraoperative radiographs to assess positioning of implants and alignment of bone.
“We can now reliably use technologies like AI to perform straightforward, menial tasks like measuring radiographs efficiently,” Cho, who is also professor in the department of orthopedic surgery and neurosurgery at the Icahn School of Medicine at Mount Sinai, told Healio Orthopedics. “As orthopedic surgeons, achieving proper alignment is crucial to patient outcome following surgery. AI can help with this and hopefully improve surgical outcome.”
Collapse
Read more about