Robotic-assisted vs. traditional medial patellofemoral ligament reconstruction: a comparative study of surgical precision and clinical outcomes

Abstract

Purpose

To evaluate the feasibility and clinical effectiveness of reconstructing the Medial Patellofemoral Ligament (MPFL) using robotic-assisted reconstruction.

Methods

This retrospective cohort study encompassed 46 patients who underwent medial patellofemoral ligament reconstruction at Yichang People’s Hospital between January 2022 and January 2024. Patients were categorized into a conventional surgery group (control group, n = 24) and a robot-assisted surgery group (experimental group, n = 22) based on whether robotic assistance was used during the procedure. The primary endpoints included the error margin between the femoral tunnel entry point, a predefined reference landmark, and the quantity of intraoperative fluoroscopic exposures. Secondary endpoints consisted of knee function scores at postoperative and final follow-up evaluations, patellar stability, operative duration, length of hospital stay, and intraoperative blood loss.

Results

The experimental group demonstrated a significantly higher degree of precision in femoral tunnel placement than the control group, as indicated by a smaller mean distance from the tunnel entry point to the reference landmark (P < 0.05). Additionally, the experimental group markedly reduced intraoperative fluoroscopic exposures relative to the control group (P < 0.05). No significant differences were observed between the two groups regarding postoperative or final follow-up patellar stability or knee function scores (P > 0.05). Furthermore, the experimental group incurred significantly shorter hospital stays and experienced less intraoperative blood loss than the control group (P < 0.05). At the same time, the operative time did not present any significant differences between groups (P > 0.05).

Conclusion

Robot-assisted MPFL reconstruction significantly enhances femoral tunnel positioning accuracy and surgical efficiency compared to conventional methods. This approach offers a promising surgical option for improving precision and efficiency in the management of recurrent patellar dislocation, with potential implications for future research.

Introduction

Recurrent patellar dislocation (RPD) is a common knee disorder in which the patella dislocates laterally when the knee is near full extension due to laxity and structural damage of the medial knee ligaments; the patella relocates to its original position when the knee is flexed. It is often accompanied by persistent dull pain, periarticular swelling, and a sense of knee instability[1, 2]. Epidemiological studies show that children and young women are significantly more likely to experience patellar dislocation than men. Conservative treatment is commonly adopted for first-time patellar dislocations, and only a tiny percentage of patients can return to their usual level of activity. Given the high recurrence rate of up to 50%,surgical intervention has become a vital and widely accepted therapeutic option for RPD patients [3–5]. MPFL repair is currently one of the most fundamental and significant effective techniques, and its effectiveness after surgery is well known. Research has demonstrated that the degree of graft tension in MPFL repair is essential for both surgical success and positive outcomes, and that either excessive and insufficient tension might impact postoperative results. Identifying the appropriate femoral tunnel entry point during conventional surgery is crucial. Multiple fluoroscopic adjustments are necessary to refine the entry point until it falls within a 5 to 7 mm range of the Schöttle point [6–8]. Several traditional methods have been developed to address this challenge, including specialized 3D navigation templates, palpation-based techniques, and imaging-guided localization. Among these, robotic-assisted technology has emerged as a promising alternative, offering enhanced precision and reproducibility.

Orthopedic surgical robots are rapidly developing, demonstrating excellent outcomes due to their distinct advantages over traditional surgery in precision, accuracy, and flexibility [9–11]. The robot possesses significant potential within the field of sports medicine. It is poised to assist physicians in achieving accurate positioning, making necessary adjustments, and ensuring proper graft alignment during surgical procedures. By monitoring and dynamically tracking the graft status in real time, the surgeon can effectively address the critical issue of inaccurate positioning. This capability enables the precise orientation of the femoral tunnel in real-time, thereby preventing needles from inadvertently entering the intercondylar region of the femur, which may result in medical complications or injury. As an emerging technology, robot-assisted surgery is still evolving. However, with increasing clinical experience, its advantages are becoming more apparent.The benefits of these techniques become clearer.

This study collected data from 46 patients who underwent medial patellofemoral ligament reconstruction at the hospital between January 2022 and January 2024. The patients were classified into two groups based on whether robotic assistance was used during the surgical procedure: the conventional surgery group (control group, n = 24) and the robot-assisted surgery group (experimental group, n = 22). Subsequently, the Schöttle point localization method and robotic assistance were employed to accurately identify the MPFL. Following acquiring a standard lateral knee image through intraoperative fluoroscopy utilising C-arm X-ray technology, the surgical procedures were conducted using the conventional Schöttle point localization technique and the robot-assisted localization method.

The objectives are: ① To evaluate whether tethered robotic-assisted medial patellofemoral ligament (MPFL) reconstruction improves the accuracy of femoral tunnel placement and clinical outcomes in comparison to traditional techniques. ②. Provide clinical evidence supporting the application of orthopedic robotic systems in MPFL reconstruction.

Information and methods

Patients and methods

This investigation constituted a retrospective cohort study. According to the established inclusion and exclusion criteria, data were meticulously collected on 46 patients who underwent MPFL reconstruction at our hospital between January 2022 and January 2024. The patients were categorized into two groups: the control group, which included 24 cases, and the experimental group, which comprised 22 cases. All surgeries were performed by the same surgical team, which had extensive experience in ligament reconstruction. Prior to initiating the robotic-assisted procedures, the team completed formal preoperative training in robotic surgery. During the procedures, standard lateral views of the knee were acquired using intraoperative fluoroscopy with a C-arm X-ray machine, and the surgical interventions were executed employing the traditional Schöttle point localization method. Additionally, the conventional Schöttle point localization method and the Tianjie robot-assisted localization method were utilized to complete the operations. The surgical equipment employed included the TiRobot® orthopaedic surgical robot (Beijing Tinavi Medical Technologies Co., Ltd., Beijing, China) and a mobile C-arm X-ray machine (Siemens, Germany).

Inclusion criteria

1. Recurrent dislocation diagnosis and age exceeding 14 years;

2. Conservative treatments failed to assist whenever a patellar dislocation occurred twice or more;

3. Persistent patellar instability symptoms, and significant medial support band lateral movement or laxity for assessment;

4. Fear test (+);

5. Patients who consented to operations and exhibited appreciation of the surgical strategy;

6. Willingness to participate in follow-up for more than 6 months.

Exclusion criteria

1. In addition to MPFL reconstruction, other bony orthopedic procedures were required in certain patients. (These indications included: Tibial tubercle–trochlear groove distance(TT-TG) > 20 mm, Q angle > 20°, anterior tilt angle of the femur ≥ 30°, high patellar Caton’s index > 1.2, and deformity of a sliding groove with a Dejour > B above);

2. Patients with severe cardiopulmonary comorbidities;

3. Patients who demonstrated a lack of cooperation or engagement in treatment and follow-up processes;

Ethical approval and consent to participate

This research was approved by the Medical Ethics Committee of Yichang Central People’s Hospital (Ethics No. 2023-182-01). All procedures were conducted in accordance with the Declaration of Helsinki (amended in 2000). Given the retrospective nature of the study, written informed consent was obtained from all participants and their families prior to surgery.

Data availability

Considering the sensitive nature of the study data and the imperative of maintaining patient privacy, the datasets produced and/or analyzed throughout this study are not accessible to the public. Participants did not provide consent for the disclosure of their personal information, which includes, but is not limited to, signed informed consent documents, imaging data, and follow-up outcomes. Nevertheless, data may be made available upon reasonable request from the corresponding author, provided that all ethical and privacy requirements are strictly observed.

Surgical approach

Control group: After successful anesthesia, the patient was placed in the supine position, and the surgical area was routinely disinfected and draped. An oblique incision was made on the anteromedial aspect of the knee to harvest the semitendinosus tendon, which was excised for autograft preparation. A 3-cm oblique incision was then made over the medial epicondylar region of the femur to expose the medial femoral soft tissue attachments. The saddle area between the medial epicondyle and the adductor tubercle was palpated and identified, and a Kirschner wire was inserted at the center of this area. Lateral fluoroscopic imaging using a mobile C-arm X-ray unit was performed to confirm the appropriate entry point. The position was adjusted until satisfactory alignment was achieved, and a 5.0-mm femoral tunnel was created along the guidewire. Along the medial edge of the patella, two threaded anchors were inserted. The sutures from the anchors were used to secure the prepared autograft tendon. The free end of the tendon was then passed through the femoral tunnel. The knee was flexed and extended to assess graft tracking. At 30° of knee flexion, an interference screw was inserted into the femoral tunnel to fix the graft. After confirming that the reconstructed MPFL was under appropriate tension, the surgical incisions were closed in layers. The procedure was completed with a sterile dressing and compression bandage.

Experimental group: After successful anesthesia, the patient was was placed in the supine position, and the operation area was routinely disinfected and toweled. An oblique incision was made in the anterior medial aspect of the knee, and the semitendinosus tendon was completely excised to prepare the autograft tendon. A mobile C-arm X-ray machine obtained a standard lateral view of the knee. Robotic navigation system preparation: Kirschner pins were inserted into the anterior side of the lower and middle femur to fix the tracer, connected to the Tenguet robotic navigation device, imported the knee X-ray data, determined the Schotte point on the image with the assistance of the Tenguet robot, and carried out the path planning, and accurately inserted the guiding pins in the direction of the robotic arm navigation, and the results of fluoroscopy by the C-arm X-ray machine were satisfactory (Fig. 1). 2 anchors with wires were selected along the medial patellar margin and inserted. Anchor nails were placed along the medial edge of the patella, the tail line of the anchor nails was knotted to fix the prepared autograft tendon, and the tail part of the tendon was passed through the femoral tunnel, the knee was flexed and extended, and in the position of 30° of flexion of the knee, an interference screw was screwed in along the femoral tunnel to fix the tendon and the reconstructed MPFL was confirmed to be in good tension. Then the surgical incision was closed one by one. The surgery was completed with a sterile dressing and a pressure bandage.

Fig. 1.

Surgery in a typical case. Note: Robot-assisted femoral tunnel localization was employed to facilitate MPFL reconstruction for a 14-year-old female patient diagnosed with recurrent patellar dislocation, following three previous dislocations. A: Preoperative orthopedic X-ray of the left knee; B: Preoperative CT scan; C: Orthopedic robotic navigation system utilized during the procedure; D: Intraoperative fluoroscopy from the C-arm X-ray machine to obtain the patient’s standard lateral radiographs; E~H: The guide pin was employed to determine the femoral tunnel needle entrance site in the orthopedic radiography; I: The guide pin was accurately positioned by the direction of the robotic arm along the planned trajectory, with fluoroscopy findings from the C-arm X-ray machine being noted satisfactory

Postoperative care

(1) 1 to 3 days post-surgery involves fixing the knee joint in a straight position using a joint brace. Patients are advised to gradually commence a series of exercises, including partially weight-bearing ambulation with double crutches, progressively increasing the intensity of their activities. Weight-bearing exercises may be initiated following the operation if the patient demonstrates tolerance. (2) Typically, patients can achieve full weight-bearing ambulation within 2 to 3 weeks after the surgery. A knee immobilization band must be utilized to maintain knee extension during ambulation for 6 weeks. (3) Patients are encouraged to achieve a minimum flexion of 90° in their knees within 3 weeks following the surgical procedure. (4) Regular exercise may commence approximately 4 to 6 months after surgery. (5) The specific timeline for recovery may vary by individual differences among patients.

Outcomes

The general perioperative data of patients from both groups were meticulously recorded. The knee pain visual analog score (VSA), the knee function Lysholm score, and the knee function Kujala score were documented and compared between the two patient groups during the preoperative period and at three months and six months postoperatively. Measurements of patellar stability indexes, including Patellar congruence angle (CA), patellar tilt angle (PTA), and lateral Patellar Displacement (LPD), were conducted and compared between the groups using preoperative and postoperative (1–3 days) knee CT and X-ray imaging data. A postoperative review was performed on the knee CT reconstruction data, focusing on measuring and comparing the distance between the femoral tunnel entry point and the MPFL anatomical point of surgery in the two patient groups. This assessment aimed to evaluate the positioning accuracy and create a femoral tunnel entry point map for visualization and analysis.

Statistical analysis

The SPSS 25.0 statistical software was employed for the analysis of data. Continuous variables that adhered to a normal distribution were represented as mean ± standard deviation (Mean ± SD). Before conducting intergroup comparisons, the assumption of homogeneity of variance was assessed. When this assumption was satisfied, independent sample t-tests were utilized; otherwise, Welch’s t-test (Satterthwaite approximation) was implemented. For continuous variables that did not conform to a normal distribution, data were presented in terms of median and interquartile range (IQR), with comparisons between groups being carried out using the Mann–Whitney U test. Categorical variables were conveyed as frequencies and percentages, and comparisons were performed using the chi-square (χ²) test. All statistical tests were two-tailed, with a p-value of < 0.05 considered statistically significant.

Sample size calculation

The research team estimated the sample size by comparing the means of two independent samples utilizing a two-sample approach t-test. According to previously published data [12], the mean distance from the entry point of the femoral tunnel to the reference landmark was 4.7 mm in the experimental group, with a standard deviation of 1.2 mm, and 6.1 mm in the control group, which also exhibited a standard deviation of 1.2 mm. The sample size was determined utilizing the Med Sci Sample Size Tools (MSST), with a significance level (α) established at 0.05 and a statistical power (1 − β) of 0.9 (β = 0.1). An allocation ratio of 1:1 between the experimental and control groups was assumed. It should be noted that the sample size calculation did not incorporate potential loss to follow-up. The formula employed for this estimation is presented below:

n=(Z_α/2 ​+ Z_β) × (σ₁² + σ₂²) ​/ (µ₁ - µ₂)².

Where, Z_α/2 = 1.96, Z_β = 1.28, σ₁ = σ₂ = 1.2, µ₁​ = 4.7, µ₂ = 6.1.

Substituting these values into the equation produces:

n=(1.96 + 1.28)² × (1.2² + 1.2²) / (4.7–6.1)² = 30.24 / 1.96 ≈ 15.43.

Consequently, the minimum necessary sample size for each group was estimated to be approximately 16 patients. The achieved sample sizes in the control group (n = 24) and experimental group (n = 22) exceeded the minimum requirement, ensuring adequate statistical power and reliability of the results.

Results

Demographic and clinical characteristics, including age, sex, body mass index (BMI), and severity of patellar instability, were systematically collected and compared between the two groups to evaluate potential confounding factors. No statistically significant differences were observed in these variables, which indicates a relatively balanced comparison (P > 0.05). This finding further suggests that the two groups of patients were comparable (Table 1).

Table 1.

Basic characteristics

Project	Control group (n = 24)	Experimental group (n = 22)	Test statistic	P-value
Age (x ± s, years)	16.0 (16.0, 20.8)	17.0 (14.0, 19.0)	Z = -0.022	0.982
Gender (male/female, n)
Male	5	7	χ2 = 0.718	0.397
Female	19	15
BMI (x ± s, kg/m2)	23.5 ± 4.1	25.0 ± 3.0	t = -1.492	0.143
Frequency (x ± s, times)	2.0 (2.0, 3.0)	2.0 (2.0, 3.0)	Z = -0.012	0.990
Times (x ± s, years)	2.5 (0.4, 7.0)	5.5 (0.9, 13.0)	Z = -1.112	0.266

Patellar stability comparison

No significant differences were observed between the control and experimental groups in preoperative measurements of the CA, PTA, or LPD (P > 0.05). Both groups demonstrated substantial improvements in all three parameters following the surgical procedure. The experimental group exhibited a significantly lower postoperative LPD than the control group, suggesting improved patellar alignment (P < 0.05). However, postoperative CA and PTA did not show significant differences between the groups (P > 0.05). (Table 2).

Table 2.

Preoperative and postoperative changes in patellofemoral joint stability

Project	Control group (n = 24)	Experimental group (n = 22)	Effect Size	95% CI	P-value
CA (x ± s, °)
Preoperative	25.6 ± 2.7	24.4 ± 3.0	0.42	[–0.43, 2.67]	0.151
Postoperative	1.5 ± 1.5	1.7 ± 1.4	–0.14	[–0.84, 0.48]	0.598
PTA (x ± s, °)
Postoperative	29.2 ± 5.8	28.0 ± 6.6	0.20	[–2.55, 4.95]	0.512
Postoperative	10.6 ± 2.9	10.3 ± 4.3	0.08	[–1.28, 1.82]	0.743
LPD (x ± s, mm)
Postoperative	16.9 ± 4.8	15.5 ± 4.0	0.32	[–1.35, 4.23]	0.292
Postoperative	4.0 ± 1.0	3.4 ± 0.8	0.66	[0.17, 1.02]	0.007

Note: CA: Congruence Angle, the angle between the bisector of the sulcus angle and the line from the sulcus apex to the inferior pole of the patella. PTA: Patellar Tilt Angle, the angle between the line connecting the highest points of the medial and lateral femoral condyles and the maximum transverse diameter extension of the patella. LPD: Lateral Patellar Displacement, the perpendicular distance from the medial edge of the patella to a line connecting the highest points of the medial and lateral femoral condyles, drawn through the medial condyle apex

Scores of knee joint function

Knee function scores for both groups were evaluated preoperatively and during the follow-up assessments conducted three and six months postoperatively, indicating continuous enhancement over time (Table 3).

Table 3.

Preoperative and postoperative subjective knee function scores in both groups

Project	Control group (n = 24)	Experimental group (n = 22)
VAS score
Preoperative	3.0 (2.0, 3.5)	3.0 (2.0, 3.0)
3 months	2.0 (1.3, 2.0)	2.0 (1.0, 2.0)
6 months	0 (0, 1.0)	0.5 (0, 1.0)
Kujala score
Preoperative	54.2 ± 1.7	54.1 ± 1.5
3 months	65.8 ± 2.4	66.1 ± 2.1
6 months	70.6 ± 2.0	71.2 ± 1.9
Lysholm score
Preoperative	62.2 ± 3.5	60.4 ± 2.7
3 months	69.5 ± 1.7	70.1 ± 1.8
6 months	73.3 ± 1.8	72.9 ± 2.3

Note: VAS indicates knee pain measured by the Visual Analog Scale (minimum unit 0.5); the Kujala score assesses patellofemoral joint function; the Lysholm score evaluates overall knee function

No significant differences were observed between the control (n = 24) and experimental (n = 22) groups in any patient-reported outcome measures (all P > 0.05). Preoperative Visual Analog Scale (VAS) scores were 3.0 in both groups and improved to 0 in the control group and 0.5 in the experimental group at the final follow-up (P > 0.05). Kujala scores increased from 54.2 ± 1.7 to 70.6 ± 2.0 in the control group and from 54.1 ± 1.5 to 71.2 ± 1.9 in the experimental group (P > 0.05). Similarly, Lysholm scores rose from 62.2 ± 3.5 to 73.3 ± 1.8 in the control group and from 60.4 ± 2.7 to 72.9 ± 2.3 in the experimental group, respectively (P > 0.05). (Table 4).

Table 4.

Comparison of knee function scores before surgery and at final follow-up visit

Project	Control group (n = 24)	Experimental group (n = 22)	Effect Size	95% CI	P-value
VAS score
Preoperative	3.0 (2.0, 3.5)	3.0 (2.0, 3.0)	-	-	0.640
Final follow-up	0 (0, 1.0)	0.5 (0, 1.0)	-	-	0.980
Kujala score
Preoperative	54.2 ± 1.7	54.1 ± 1.5	0.06	[–0.84, 1.04]	0.348
Final follow-up	70.6 ± 2.0	71.2 ± 1.9	–0.30	[–2.26, 1.14]	0.923
Lysholm score
Preoperative	62.2 ± 3.5	60.4 ± 2.7	0.56	[–0.10, 3.90]	0.059
Final follow-up	73.3 ± 1.8	72.9 ± 2.3	0.19	[–1.26, 2.06]	0.191

Note: - Not applicable

Comparison of femoral tunnel positioning accuracy

The average distance from the femoral tunnel positioning point to the reference point was notably reduced in the experimental group (3.4 ± 1.0 mm) compared to the control group (5.6 ± 1.5 mm) (P < 0.05), indicating an enhancement in the precision of tunnel placement facilitated by robot-assisted navigation. (Table 5).

Table 5.

Comparison of femoral tunnel positioning accuracy between two groups (mm, x̄ ± s)

Project	Control group (n = 24)	Experimental group (n = 22)	Effect Size	95% CI	P-value
Distance (mm)	5.6 ± 1.5	3.4 ± 1.0	1.71	[1.45,2.95]	0.001

Note: Distance: Mean distance from the femoral tunnel entry point to the reference point

Comparison of perioperative outcomes

The two groups exhibited no significant differences in surgical duration or length of hospital stay (P > 0.05). However, intraoperative blood loss was significantly reduced and the number of fluoroscopic exposures was markedly diminished in the experimental group compared to the control group (P < 0.05). (Table 6).

Table 6.

General perioperative data comparison

Project	Control group (n = 24)	Experimental group (n = 22)	Effect Size	95% CI	P-value
Surgical time (±s, min)	74.9 ± 6.9	77.8 ± 6.5	0.43	[–6.93, 1.13]	0.156
Hospital stay (±s, days)	6.3 ± 1.5	6.1 ± 1.4	0.14	[–0.89, 1.29]	0.795
Blood loss (±s, ml)	63.8 ± 10.7	27.27 ± 9.4	3.70	[29.6, 43.3]	0.001
Fluoroscopy (±s, time)	3.0 (3.0, 4.0)	1.0 (1.0, 2.0)	-	-	0.001

Note: Fluoroscopy: Number of fluoroscopic exposures. -: Not applicable

Discussion

Physiological significance of the medial patellofemoral ligament

MPFL is a critical structure for maintaining medial stability of the knee joint. At approximately 30° of flexion, it provides about 50–60% of the medial restraining force, essential in limiting excessive lateral displacement [13–15]. Biomechanical studies show that during knee flexion of 0° to 110°, the MPFL changes length by only 1.1 mm, indicating its isometric nature [16, 17]. This characteristic ensures balanced patellar kinematics and uniform intra-articular stress distribution. The main goal of anatomical MPFL reconstruction is to restore its native anatomical configuration and biomechanical function accurately. Accurate femoral tunnel placement ensures graft isometry and optimal postoperative function. Misplacement of the femoral tunnel, anteriorly or posteriorly, alters graft tension and joint stress, increasing the risk of instability, cartilage damage, and poor outcomes. Optimizing femoral tunnel placement to restore the anatomical integrity and biomechanical function of MPFL remains a critical area for ongoing research. Prior studies show MPFL is a fan-shaped ligament arising from the femoral adductor tubercle and extending anteriorly. Based on its fiber orientation, the MPFL can be subdivided into three distinct portions: The superior fibers lie deep to the vastus medialis obliquus, coursing anterosuperior to blend with the fascia of the vastus intermedius. The middle fibers run anteriorly, merging with the superficial tendon of the vastus medialis obliquus and terminating at the superomedial border of the patella. The inferior fibers extend anteroinferior, inserting onto the medial patella. The patella’s border is at its proximal two-thirds. Near the adductor tubercle, the MPFL fibers start slender and narrow; they widen and thicken towards the medial border of the patella, reaching their greatest thickness. Congenital deficiencies in the medial femoral condyle may increase the risk of ligamentous tearing. In this study, we identified the femoral attachment point of the MPFL as the midpoint between the medial femoral epicondyle and the adductor tubercle. This landmark was the reference for assessing femoral tunnel placement, allowing objective quantification of the tunnel’s deviation from the native site. Anatomical studies guide optimal tunnel positioning based on native MPFL attachment [18–21]. The Schöttle point is the gold standard for femoral tunnel placement in clinical practice. Still, it does not align precisely with the true anatomical center of the MPFL attachment on the medial femoral condyle. Its localization accuracy is affected by systematic errors from different fluoroscopic projection angles. The conventional method for femoral tunnel placement uses C-arm fluoroscopy and lacks real-time feedback. After the guidewire puncture, the surgeon must adjust the wire’s insertion angle and position based on images from a mobile C-arm to achieve the desired tunnel orientation. This increases operative time and radiation exposure risk. The guide pin’s insertion depth depends on the surgeon’s experience and tactile feedback, meaning slight angular deviations in the lateral view from C-arm fluoroscopy can cause significant errors in femoral tunnel placement. A 2.5° deviation can lead to over 5 mm positioning error in the femoral tunnel [22].

Selection of reference landmarks

Robot-assisted surgery uses intraoperative imaging to plan and monitor guide pin trajectories in real time. This approach significantly diminishes the frequency of fluoroscopic exposures and the necessity for repeated adjustments, thereby facilitating a more precise determination of the optimal femoral tunnel position. For postoperative assessment, we utilized thin-slice (0.6 mm) CT scans to reconstruct three-dimensional images, which enable accurate measurement of tunnel positioning deviations. This approach affords a more authentic depiction of the medial anatomical structures of the knee, thereby improving the accuracy and objectivity of the evaluation. Utilizing computer-assisted measurement tools allows for precise analysis of the femoral tunnel’s actual postoperative position and facilitates quantifying its spatial relationship to pertinent anatomical landmarks. In contrast, traditional postoperative radiographs are constrained by the limitations of two-dimensional imaging, which frequently obstructs the clear identification of the guide pin’s entry and exit points. Furthermore, image overlap can significantly undermine measurement precision. Postoperative follow-up imaging is also constrained by practical limitations related to facility availability and time restrictions. The inability to replicate intraoperative fluoroscopic angles and the necessity for high patient compliance complicates the procurement of standardized lateral knee views, adversely impacting the accuracy of postoperative evaluations. While conventional computed tomography (CT) imaging theoretically permits planar distance measurements, it solely represents a singular axial slice. The selection of a representative slice differs among individuals and introduces a significant degree of subjectivity, potentially influencing the consistency and comparability of the measurements. Furthermore, CT scans of the knee frequently inadequately portray the medial joint anatomy, thereby restricting their utility in postoperative assessments. In contrast, three-dimensional computed tomography (3D-CT) reconstruction provides a more intuitive and spatially accurate approach to postoperative evaluation, enhancing the precision and repeatability. Nevertheless, the inherent limitations of imaging technology must be considered, including the potential measurement errors that systematic image smoothing processes may introduce. Despite these challenges, three-dimensional reconstruction presents distinct advantages in postoperative analysis, offering a more reliable radiographic foundation for clinical evaluation.

Advantages of Robotic-Assisted surgery

This study demonstrates that robot-assisted surgery significantly improves the accuracy of femoral tunnel placement in MPFL reconstruction compared to conventional techniques. The reduced deviation between the tunnel center and the anatomical insertion point (P < 0.05) underlines the superior spatial precision achievable with robotic assistance. These results suggest that satisfactory short-term clinical outcomes can be achieved with either technique. However, robotic systems’ superior precision and radiation-sparing benefits may offer additional long-term advantages, particularly in complex cases or among less experienced surgeons. Notably, the operative time in the robotic group was slightly longer than in the conventional group, although the difference was not statistically significant (P > 0.05). This is likely due to the learning curve of adopting new robotic systems. In the early phase of the study, the surgeon’s limited experience with the platform, along with the assistant’s unfamiliarity, occasional fixation issues, and the need for intraoperative instruction, collectively contributed to prolonged surgical times. Nevertheless, as proficiency increased and team coordination improved, operative time in the robotic group decreased significantly, underscoring the presence of a learning curve and the potential for workflow optimization with continued use. Moreover, the robotic system’s capacity to reduce intraoperative radiation exposure was a consistent advantage throughout the study. The statistically significant reduction in fluoroscopy usage in the robotic group (P < 0.05) highlights the system’s potential to mitigate occupational radiation risks, which is particularly important for orthopedic teams regularly exposed to fluoroscopy in daily practice. In conclusion, although the short-term clinical outcomes were similar between the two groups, robot-assisted MPFL reconstruction demonstrated clear intraoperative advantages in precision, safety, and radiation reduction. With increased familiarity and refinement of surgical protocols, robotic systems are poised to play an increasingly valuable role in enhancing the quality and safety of ligament reconstruction procedures.

Postoperative assessment of tunnel distribution was conducted using 3D-CT images of the knee. The results revealed that, compared to the conventional group, the entry points of the femoral tunnels in the robotic group were more centrally clustered, indicating superior tunnel localization accuracy with robotic technology’s assistance. (Fig. 2). Biomechanical studies suggest that a deviation of the femoral tunnel exceeding 5 mm from the anatomical insertion may result in a length change of over 12 mm in MPFL, subsequently increasing medial cartilage pressure. In this study, the robotic group demonstrated superior tunnel accuracy, which may help reduce potential complications from graft malposition.

Fig. 2.

Distribution of femoral tunnel localization points on the medial femoral condyle. Note: (A) Preoperative plain CT scan of the femur; (B) Postoperative plain CT scan of the femur; (C) Threedimensional reconstruction showing the relationship between the femoral tunnel entry point and the anatomical attachment of the medial patellofemoral ligament(MPFL) (AT: adductor tubercle; ME: center of the medial epicondyle); (D) Schematic distribution of femoral tunnel entry points (mm), with red dots indicating the roboticassisted group, blue dots indicating the conventional group, and the circle center marking the anatomical MPFL attachment site

Postoperative complications and cause analysis

All patients in both groups demonstrated primary (Grade A) wound healing during their hospitalization, with no evidence of wound infection present. Furthermore, no adverse events were reported during the follow—up period, including patellar radiolucency or graft complications or failure. In the conventional group, the mean deviation of femoral tunnel placement was 5.6 ± 1.5 mm, with four patients surpassing the acceptable threshold (7.2 mm, 7.3 mm, 9.5 mm, and 7.6 mm). Despite this, no graft failure or patellar instability cases were observed at the final follow-up, and all patients maintained good patellar stability. However, several limitations and potential sources of variability should be acknowledged. First, slight rotation or inclination during 3D-CT may have impacted the precision of measurements. Second, the compression effects stemming from the insertion of interference screws may have displaced the graft center, modifying its biomechanical characteristics behavior. Furthermore, image reconstruction does not accurately replicate anatomical reality; the smoothing of images during processing can obscure bony landmarks, thereby diminishing the precision of tunnel localization. Moreover, the relatively young age of most patients may have facilitated soft tissue compensation, potentially concealing the effects of tunnel malposition. Previous studies have indicated that complications arising from non-anatomic reconstruction often manifest two to five years postoperatively. However, our follow-up period was relatively brief, possibly leading to underestimating the long-term outcome risks. Ultimately, minor intraoperative variations in the knee flexion angle during the drilling process may have influenced the final tunnel positioning, thereby affecting the consistency of reconstruction outcomes.

5、Limitations

This study examined the application of robotic-assisted surgery in MPFL reconstruction and produced promising short-term clinical outcomes. Nevertheless, several limitations require attention in forthcoming research endeavors. Initially, only short-term outcomes were evaluated due to the limited sample size, which resulted in the absence of long-term follow-up data. Considering that the recurrence rate and functional recovery following MPFL reconstruction generally manifest after one year, the six-month follow-up duration is comparatively brief. Furthermore, there is a critical need for further optimization of the surgical technique and for large-scale studies featuring extended follow-up to validate the clinical efficacy and long-term stability of robotic-assisted MPFL reconstruction. Secondly, the current follow-up duration is inadequate for a comprehensive assessment of long-term outcomes. While short-term improvements in knee function have been noted, additional data are necessary to evaluate graft healing, ligament laxity, and the potential for recurrent dislocation. Robotic technology significantly diminishes human error, reduces radiation exposure, and enhances surgical precision and safety. Another limitation of this study is the absence of a formal matching strategy to control for potential confounding variables. Although no significant differences were observed in baseline characteristics between the two groups, the lack of techniques such as propensity score matching may still allow for residual confounding. Nonetheless, challenges such as equipment calibration, image quality, and the accuracy of real-time monitoring remain critical areas for improvement. Intraoperative adjustments and clinical supervision are likewise essential. Furthermore, robotic systems may increase surgical costs, and discovering methods to integrate this technology while preserving cost-effectiveness poses challenges in clinical practice. This study primarily concentrated on the precision of tunnel placement and short-term functional recovery after robotic-assisted MPFL reconstruction. It did not encompass an analysis of surgical cost-effectiveness, the learning curve, or patient satisfaction. Owing to ethical considerations, direct intraoperative exposure of medial bony landmarks for accurate measurement was not practicable, which may have led to specific measurement errors. Notwithstanding these limitations, our findings suggest that robotic-assisted MPFL reconstruction offers significant advantages in femoral tunnel localization and results in favorable short-term outcomes. Nevertheless, its efficacy in restoring patellar stability and enhancing long-term knee function necessitates further validation.

Significantly, robotic technology enhances surgical precision and offers potential benefits in minimizing the learning curve. Traditional MPFL reconstruction necessitates high anatomical accuracy and surgical expertise. In contrast, implementing robotic systems, through preoperative planning and real-time navigation, can facilitate the learning process and diminish adaptation time for surgeons. Future research ought to investigate this technology’s utility among surgeons with varying experience levels and its prospective role in standardized training programs. Moreover, with ongoing advancements in robotic systems, their potential for personalized surgical planning warrants consideration. Robotic systems possess the potential to provide tailored surgical strategies that enhance postoperative recovery through the integration of imaging data and patient-specific anatomical features. Future studies must assess the long-term clinical value of robotic-assisted MPFL reconstruction.

Conclusion

Robot-assisted MPFL reconstruction significantly enhances femoral tunnel placement accuracy and streamlines the surgical workflow compared to conventional techniques. These results improve outcomes for recurrent patellar dislocation and provide valuable guidance for future research.

Author contributions

Hongzhi Fang: As the first author, drafted the manuscript, conducted data analysis, and assisted in measurements. Shang Zhenghui: As the corresponding author and the primary surgeon, performed the surgeries and supervised and guided the manuscript preparation. Tianli Du: Provided high-quality illustrations for the manuscript and assisted in organizing and measuring the data.

Funding

This work was not funded by any source.

Data availability

No datasets were generated for this study; however, the datasets used and/or analyzed during the current study are not publicly available due to the sensitive nature of clinical data and patient privacy considerations. De-identified data may be made available from the corresponding author upon reasonable request and subject to compliance with institutional privacy regulations.

Declarations

Ethical approval and consent to participate

This study was approved by the Medical Ethics Committee of Yichang Central People’s Hospital (Ethics No. 2023-182-01). All procedures were conducted by the Helsinki Declaration (amended in 2000). Given the study’s retrospective nature, written informed consent was obtained from all participants and their families before surgery.

Consent for publication

Not applicable.

Ethics committee approval

Medical Ethics Committee of Yichang Central People’s Hospital (Ethics No. 2023- 182- 01).

Human ethics and consent to participate declarations

All participants and their families provided informed consent and signed the relevant consent forms. The study was approved by the Medical Ethics Committee of Yichang Central People’s Hospital.

Competing interests

The authors declare no competing interests.