Efficacy of the CALM® Algorithm in Reducing Motion-Induced Artifacts in CBCT Imaging: A Fractal Dimension Analysis of Trabecular Bone.

Abstract

Objective: The primary goal of this investigation was to ascertain the efficacy of the CALM^® motion artifact reduction algorithm in diminishing motion-induced blurriness in Cone Beam Computed Tomography [CBCT] images. The assessment was conducted through Fractal Dimension [FD] analysis of the trabecular bone.

Methods and materials: A desiccated human mandible was subjected to Planmeca ProMax 3D^® scanning under eight distinct protocols, marked by variations in motion presence [at 5, 10, and 15 degrees] and the deployment of CALM^®. In every scan, five distinct regions of interest [ROIs] were designated for FD analysis, meticulously avoiding tooth roots or cortical bone. The FD was computed employing the box-counting method with Image-J 1.53 software.

Results: Our findings reveal that a 5-degree motion does not significantly disrupt FD analysis, while a 10-degree motion and beyond exhibit statistical differences and volatility among the sites and groups. A decreased FD value, signifying a less intricate or "rough" bone structure, correlated with amplified motion blurriness. The utilization of CALM^® software seemed to counteract this effect in some instances, reconciling FD values to those akin to the control groups. Nonetheless, CALM^®'s efficacy differed across sites and motion degrees. Interestingly, at one site, CALM^® application in the absence of motion resulted in FD values considerably higher than all other groups.

Conclusion: The study indicates that motion, particularly at 10 degrees or more, can considerably impact the FD analysis of trabecular bone in CBCT images. In some situations, the CALM^® motion artifact reduction algorithm can alleviate this impact, though its effectiveness fluctuates depending on the site and degree of motion. This underscores the necessity of factoring in motion and the employment of artifact reduction algorithms during the interpretation of FD analysis outcomes in CBCT imaging. More research is necessary to refine the application of such algorithms and to comprehend their influence on different sites under varying motion degrees.