Comparative CT-Based Assessment of Segmentation Accuracy and Volumetric Variability in Scaphoids from Frozen and Thawed Cadaver Arms for Scaphoid Prosthesis Modeling

Purpose: Cadavers for research purposes are often stored deep frozen. To avoid repeated thawing and freezing, which could affect tissue quality, we investigated the feasibility of using a frozen and thawed cadaver wrist to model a scaphoid for a patient-specific prosthesis. We also assessed the accuracy of different segmentation methods which are necessary to smoothen the surface of the model. Methods: computed tomography (CT) scans of a frozen and thawed wrist were made. The tightly and smoothened segmented scaphoids were compared. The frozen and thawed scaphoid models were registered, and volume differences were calculated. Results: No statistical differences in volumes were detected between the segmented frozen and thawed scaphoid models. The size of the scaphoid model remained consistent between the tightly and smoothened segmented versions. Conclusion: There are no significant volume differences between frozen and thawed scaphoid bones. With regards to size, it is feasible and accurate to model a smoothened scaphoid for a patient-specific scaphoid prosthesis from a frozen cadaver arm which matches the one of the thawed wrist.