Automated Image Processing Based 3D Printed Scaffolds For Critical Size Bone Fracture Treatment

Abstract

Bone tissues in critical size bone fracture cases do not re-generate naturally and require special treatment procedures, such as cast alignments, support plates and bone grafting. These procedures are risky and often have high rejection rates. Modern clinical procedures use scaffold implants that can facilitate the process of healing with a lesser risk, provide mechanical support and a porous, nutritious medium for bone tissue regeneration and recovery. They require special training, tools, and significant time to be manufactured, and are generally made at a dedicated laboratory. The whole process takes around a week to be manufactured and implanted in the fracture site. In this work a novel technique for automatic segmentation of bone fractures from CT scan images to facilitate the process of manufacturing the patient-specific scaffolds in a significantly shorter time, without the need of skilled personnel has been presented. To achieve this, the procedure of generating 3D printable models was automated using image processing and machine learning algorithms. For this, 3D CT (Computed Tomography) images were used as input (as a series of 2D slices), acquired using a micro-CT scanner for the approach. After pre-processing the acquired images (filtering), thresholding segmentation was applied to extract the bone from the scan. This step is followed by orientation optimization of the segmentation result, by taking advantage of a global optimization technique, namely Simulated Annealing, to ensure maximized visibility of the fracture in a projected view of the volume by projecting the volume on a 2D surface. Binary hole-filling techniques and bone thickness estimation is then used to create a 3D template (model) to be sent for scaffold printing to a compatible 3D printer. Experiments with both synthetic and real datasets show that the proposed method is an effective approach for creating rapid, precise, and patient-specific 3D scaffolds to treat critical-size bone fractures.