Assessment of a semiautomated pelvic floor measurement model for evaluating pelvic organ prolapse on MRI

Abstract

The objective of this study was to assess the performance of a semiautomated pelvic floor measurement algorithmic model on dynamic magnetic resonance imaging (MRI) images compared with manual pelvic floor measurements for pelvic organ prolapse (POP) evaluation. We examined 15 MRIs along the midsagittal view. Five reference points used for pelvic floor measurements were identified both manually and using our semiautomated measurement model. The two processes were compared in terms of accuracy and precision. The semiautomated pelvic floor measurement model provided highly consistent and accurate locations for all reference points on MRI. Results also showed that the model can identify the reference points faster than the manual-point identification process. The semiautomated pelvic floor measurement model can be used to facilitate and improve the process of pelvic floor measurements on MRI. This will enable high throughput analysis of MRI data to improve the correlation analysis with clinical outcomes and potentially improve POP assessment.