A novel approach to double-bundle posterior cruciate ligament reconstruction: leveraging individualized robotic navigation and multimodal imaging for superior precision and feasibility

HIGHLIGHTS This work presents the first clinical application of robotic-assisted arthroscopy for anatomic double-bundle posterior cruciate ligament reconstruction, filling a notable void in the field. The system enables unprecedented surgical precision, reliably creating anterolateral and posteromedial tunnels that optimally replicate native ligament anatomy. It significantly improves procedural safety and efficiency by enhancing protection of critical neurovascular structures and obviating the need for intraoperative CT imaging. The technique secures anatomic graft placement, which is foundational for restoring natural knee kinematics and stability. Dear Editor, We read with interest the recent cadaver study by Yang et al, which compellingly demonstrated that robotic-assisted surgery could create bone tunnels with superior anatomical precision compared to conventional techniques in posterior cruciate ligament (PCL) reconstruction[1]. We would like to highlight several methodological considerations. First, the study’s comparative framework involved cadaveric specimens, which inherently limits direct extrapolation to clinical outcomes in living patients. Second, the use of intraoperative three-dimensional (3D) C-arm scanning, while precise, may introduce logistical complexities in a clinical workflow. Finally, the investigation was confined to a single-bundle reconstruction technique, which does not fully replicate the anatomical and functional goals of a biomimetic dual-bundle approach. To address these gaps, our team has developed and clinically implemented – for the first time – an Individualized Robotic Navigation system based on preoperative MRI-CT multimodal imaging[2]. This platform is specifically designed to enhance both the convenience and accuracy of anatomic double-bundle PCL reconstruction in patients. Double-bundle PCL reconstruction is theoretically more aligned with anatomical and biomechanical principles, allowing for more effective restoration of knee joint stability[3] However, its widespread clinical adoption still faces multiple challenges. First, the bony landmarks at the tibial insertion sites of the PCL are not distinct, making them difficult to clearly identify under arthroscopic view, which complicates accurate tunnel placement[4]. Second, this surgical technique entails the precise formation of two discrete tunnel bundles: the anterolateral (AL) and the posteromedial (PM). Any deviation in the angle, depth, or position of these tunnels may compromise graft isometry, tension distribution, and even lead to surgical failure or postoperative joint stiffness[4]. Moreover, due to the proximity of the PCL to critical neurovascular structures such as the popliteal artery and tibial nerve, there is a risk of drill over-penetration or excessively deep guidewire placement during tibial tunnel preparation, which could result in irreversible vascular or nerve injury[5]. In this context, our self-developed Individualized Robotic Navigation system offers a novel solution for achieving safer and more precise double-bundle reconstruction. We initially applied CT and MRI multimodal fusion technology, segmenting images layer by layer in coronal, sagittal, and axial planes (Fig. 1A). This process enabled the generation of a 3D reconstruction of the patient’s knee joint, including the femur, femoral cartilage, tibia, tibial cartilage, PCL insertion, and meniscus (Fig. 1B), allowing for precise annotation of the tibial footprints point of the PCL (Fig. 1C). The femoral tunnel was designed based on the PCL’s anatomy and its readily identifiable femoral insertion point on the medial wall of the intercondylar notch (Fig. 1D, upper). In contrast, the tibial tunnel was planned preoperatively based on the annotated tibial footprint, as this landmark is not directly visible arthroscopically (Fig. 1D, lower). After registration, the error was less than 0.5 mm. The Individualized Robotic Navigation system was then employed to drill the AL and PM bundle femoral and tibial tunnels as preoperatively planned. The system provided real-time control, confining the tunnel trajectory to within 2 mm and 1° of the planned path and limiting drill bit protrusion to 5 mm to safeguard posterior neurovascular structures (Fig. 1E). Critically, our protocol eliminates the need for intraoperative CT scanning, thereby streamlining the surgical workflow. Figure 1.: Robotically assisted posterior cruciate ligament (PCL) reconstruction. (A) Automatic segmentation of bony and ligamentous structures. (B) 3D reconstruction model of the patient’s knee joint. (C) Precise annotation of the PCL’s tibial insertion footprint. (D) Preoperative planning of the femoral and tibial tunnel positions and trajectories. (E) Drilling of the anterolateral (AL) and posteromedial (PM) bundle femoral and tibial tunnels according to the preoperative plan, executed by the universal reconstruction surgical robot system. (F) Robotic-assisted femoral tunnel drilling (AL & PM). (G) Robotic-assisted tibial tunnel drilling (AL & PM). (H) AL and PM graft bundles in situ. (I) Intraoperative assessment of graft length change across knee flexion-extension. (J) Postoperative 3D CT of femoral tunnels. (K) Tibial tunnels. Following tunnel preparation, arthroscopic evaluation confirmed excellent positioning of both the AL and PM femoral tunnels (Fig. 1F). The tibial tunnels, which were marked with two 3.0-mm Steinmann pins, were also assessed and found to be satisfactory (Fig. 1G). After pulling the autologous hamstring grafts through the tunnels, we confirmed that the AL and PM bundles were well-aligned and under good tension (Fig. 1H). Furthermore, the universal reconstruction surgical robot system measured ligament graft length changes throughout a 0–140° range of knee motion. The results demonstrated excellent length change patterns of 2.6 mm for the AL bundle and 3.2 mm for the PM bundle (Fig. 1I). Concurrently, the Lachman and posterior drawer tests were negative, indicating restored stability. Postoperative 3D CT confirmed that the femoral and tibial tunnels were in the anatomical position (Fig. 1J, K). This study has notable limitations, including a small cohort of only two patients and a lack of long-term functional outcomes. Despite its limitations, this technology has propelled a paradigm shift in orthopedic surgery by achieving dynamic, real-time, millimeter positioning accuracy. This breakthrough significantly enhances surgical precision and reproducibility, ushering in a new era of precision, personalization, and intelligent care. Ethical approval This study was approved by the Ethics Committee of Beijing Tsinghua Changgung Hospital (Protocol No. 24751-0-002) on February 28, 2025. The committee is located at No. 168 Litang Road, Changping District, Beijing, China. Consent Written informed consent was obtained from the patient for publication. The consent form is available for editorial review upon request. Sources of funding This research was supported by the Beijing Municipal Key Project (NO. Z221100007422013).