April 8, 2021
This article, Deep learning approach for automatic landmark detection and alignment analysis in whole-spine lateral radiographs, co-authored by aetherAI and Chang Gung Memorial Hospital was accepted and published by a prestigious journal, Scientific Report as recognized its deep learning approach breakthrough.
Human spinal balance assessment relies considerably on sagittal radiographic parameter measurement. Deep learning could be applied for automatic landmark detection and alignment analysis, with mild to moderate standard errors and favourable correlations with manual measurement. In this study, based on 2210 annotated images of various spinal disease aetiologies, it developed deep learning models capable of automatically locating 45 anatomic landmarks and subsequently generating 18 radiographic parameters on a whole-spine lateral radiograph.
(A schematic diagram of the pipeline for parameter estimation is illustrated.)
In the assessment of model performance, the localization accuracy and learning speed were the highest for landmarks in the cervical area, followed by those in the lumbosacral, thoracic, and femoral areas. The proposed automatic alignment analysis system was able to localise spinal anatomic landmarks with high accuracy and to generate various radiographic parameters with favourable correlations with manual measurements.