Team determines optimal parameters for patient-specific 3D printing

Sunday, November 26 | 1:30 p.m.-1:40 p.m. | S4-SSIN01-4 | Room S401

In this talk, researchers from the University of Pittsburgh will present optimized settings for 3D printing of femurs in patients with cam-type femoroacetabular impingement (FAI).

During the talk, physician scientist in training Maxwell Lohss will explain the methods by which his team improved the conversion of segmented images into surface mesh using Mimics Medical 25.0 software. First, Lohss et al enlisted an experienced engineer to segment CT images of five femurs with cam-type FAI. STL mesh contours were then manually adjusted across the model for quality and accuracy. Next, a board-certified radiologist validated the segmentation to create a set of anatomical reference meshes.

Preset settings for low, medium, high, and optimal were used in Mimics to convert image masks to meshes. 3D deviation analyses compared each mesh to the reference, and models were evaluated for mesh irregularities including nonmanifold edges, self-intersections, highly creased edges, spikes, and holes.

For the optimal preset, for example, 40 different meshes were created by modifying the smooth factor (SF) and smoothing iteration (SI) settings from 0.3-1 and 1-20, respectively. Meshes were identified that showed an average root mean square deviation (RMSD) below 0.20 mm and a mesh that had a similar surface quality to the reference. Meshes were further evaluated using planar deviation analyses formatted around the femoral neck axis to assess error at the most clinically relevant portion of the model around the cam deformity.

The low, medium, high, and optimal settings showed an average RMSD of 1.63 mm, 0.72 mm, 0.44 mm, and 0.26 mm, respectively, Lohss reported, noting that the predefined settings in Mimics can create meshes that are error-prone and of poor quality.

However, optimizing predefined smoothing parameters in mesh conversion with SF > 0.9 and SI = 20 produced a mesh that retained the anatomical detail and surface quality of a mesh that was manually corrected and smoothed. Ultimately, the analysis identified optimal parameters for converting femur image segmentation to surface meshes with high mesh quality while maintaining accuracy, according to Lohss.

Stop by this Sunday afternoon session for more details.

Page 1 of 2
Next Page