E-Posters-A Novel 3D Slicer Module for Automatic Pelvic Floor Measurement and 3D Reconstruction with MRI

A Novel 3D Slicer Module for Automatic Pelvic Floor Measurement and 3D Reconstruction with MRI - Ziyun Liang, MD

Pdf Summary

This document describes a novel 3D slicer module that has been developed for automatic pelvic floor measurement and 3D reconstruction using MRI images. The module, named "Pelvic Floor," incorporates deep learning models for object detection and localization, as well as semantic segmentation, to achieve automatic measurement and reconstruction. The module also includes a coordinate correction feature that generates a new coordinate system, called the PICS reference system. A report unit is designed to automatically generate the measurement results as a file. The deep learning models were trained and validated on a set of MR images with manual annotations and successfully tested on a group of subjects. The researchers developed this module to improve pelvic floor evaluation efficiency and address the shortage of medical resources. The module is built on the 3D Slicer platform, an open-source, free medical image platform known for its advanced capabilities and wide range of functionality. The module features a user-friendly graphical user interface that allows users to choose input, measure distances, label points, generate reports, and start 3D reconstruction. The document includes various figures illustrating the pipeline, GUI design, and the results obtained from the automated measurement and 3D reconstruction. The researchers hope that this novel technique will bridge the gap between intelligent computer models and clinical usage in the field of pelvic floor evaluation.

Keywords

3D slicer module

automatic pelvic floor measurement

3D reconstruction

MRI images

deep learning models

object detection

semantic segmentation

coordinate correction

PICS reference system

report unit