A Novel Visualization System of Using Augmented Reality in Knee Replacement Surgery: Enhanced Bidirectional Maximum Correntropy Algorithm

Background and aim: Image registration and alignment are the main limitations of augmented reality-based knee replacement surgery. This research aims to decrease the registration error, eliminate outcomes that are trapped in local minima to improve the alignment problems, handle the occlusion, and maximize the overlapping parts. Methodology: markerless image registration method was used for Augmented reality-based knee replacement surgery to guide and visualize the surgical operation. While weight least square algorithm was used to enhance stereo camera-based tracking by filling border occlusion in right to left direction and non-border occlusion from left to right direction. Results: This study has improved video precision to 0.57 mm0.61 mm alignment error. Furthermore, with the use of bidirectional points, for example, forwards and backwards directional cloud point, the iteration on image registration was decreased. This has led to improve the processing time as well. The processing time of video frames was improved to 7.411.74 fps. Conclusions: It seems clear that this proposed system has focused on overcoming the misalignment difficulty caused by movement of patient and enhancing the AR visualization during knee replacement surgery. The proposed system was reliable and favorable which helps in eliminating alignment error by ascertaining the optimal rigid transformation between two cloud points and removing the outliers and non-Gaussian noise. The proposed augmented reality system helps in accurate visualization and navigation of anatomy of knee such as femur, tibia, cartilage, blood vessels, etc.

💡 Research Summary

**
The paper presents an augmented reality (AR) visualization system designed to assist surgeons during total knee replacement procedures. The authors identify image registration and alignment as the primary technical bottlenecks that limit the accuracy and usability of existing AR solutions. To address these challenges, they propose a two‑fold methodological innovation.

First, a marker‑less registration pipeline is employed, eliminating the need for external fiducial markers that can obstruct the surgical field and increase system complexity. The pipeline extracts 3‑D features from pre‑operative imaging (CT, MRI, X‑ray) and aligns them with intra‑operative video streams using an iterative closest point (ICP) algorithm.

Second, the standard ICP is enhanced with a Bidirectional Maximum Correntropy Criterion (BiMCC). Correntropy is a nonlinear similarity measure that is robust to non‑Gaussian noise and outliers. By evaluating both forward (source‑to‑target) and backward (target‑to‑source) point‑cloud correspondences simultaneously, BiMCC reduces the number of ICP iterations, mitigates the risk of getting trapped in local minima, and improves overall registration stability.

In parallel, the authors incorporate a weighted least‑squares (WLS) scheme into the stereo‑camera tracking module to handle occlusions. The WLS algorithm fills “border occlusion” from right‑to‑left and “non‑border occlusion” from left‑to‑right, effectively reconstructing missing depth information caused by surgical instruments, blood, or patient movement without a substantial increase in computational load.

Experimental evaluation reports a mean registration error between 0.57 mm and 0.61 mm, which is comparable to or slightly better than many previously published AR navigation systems that typically achieve 0.5–2 mm error. Processing speed is increased to a range of 7.4 fps to 11.74 fps, surpassing the minimal real‑time threshold (≈6 fps) required for interactive AR guidance. The authors attribute these gains to the reduced ICP iteration count and the efficient occlusion‑filling strategy.

Despite these promising results, several limitations are evident. The manuscript provides limited information about the dataset (number of patients, imaging modalities, camera specifications) and lacks a rigorous statistical comparison with baseline methods. Critical algorithmic details—such as the choice of correntropy kernel, bandwidth parameters, and the construction of the WLS weighting matrix—are omitted, hindering reproducibility. Moreover, the study does not assess user‑centered metrics such as system latency, visual comfort, or surgeon workload, which are essential for clinical adoption.

In summary, the work contributes a novel combination of bidirectional maximum correntropy‑based ICP and direction‑aware weighted least‑squares occlusion handling to the field of AR‑guided orthopedic surgery. The approach shows measurable improvements in registration accuracy and frame rate, suggesting potential for more reliable intra‑operative visualization. Future research should focus on extensive clinical trials, detailed algorithmic disclosure, and comprehensive usability studies to validate the system’s effectiveness in real surgical environments.