A comparative study of clinical and radiographic outcomes in total knee arthroplasty assisted by the CT-free Smith & Nephew CORI robotic system versus the Brainlab Knee3 navigation system

Abstract

The CT-free Smith & Nephew CORI robotic system and the Brainlab Knee3 navigation system represent two distinct technological approaches for assisting total knee arthroplasty (TKA). In this retrospective comparative study, 100 consecutive patients with knee osteoarthritis who underwent TKA from December 2020 to June 2025 were allocated to either a robotic-assisted group (n = 50) utilizing the CT-free CORI system or a navigation-assisted group (n = 50) utilizing the Brainlab Knee3 system. Perioperative parameters, early and mid-term functional scores, radiographic alignment, complications, and patient satisfaction were analyzed. The results demonstrated that while both systems achieved comparable long-term functional outcomes and similar safety profiles, the CORI robotic system showed distinct advantages. It significantly reduced osteotomy time, total operative time, and intraoperative blood loss. Patients in the CORI group also experienced less early postoperative pain, better knee function, and higher range of motion at 3 days postoperatively. Radiographically, the CORI group achieved superior accuracy in restoring the hip-knee-ankle angle and controlling the coronal alignment of the femoral and tibial components. These findings suggest that although both systems are effective for TKA, the CORI robotic system offers enhanced operative efficiency, reduced early surgical trauma, faster initial recovery, and improved precision in achieving targeted mechanical alignment.

Introduction

Total Knee Arthroplasty (TKA) is the primary treatment for end-stage knee osteoarthritis. However, the long-term clinical outcomes of TKA are directly influenced by the accuracy of lower limb alignment and prosthesis positioning achieved intraoperatively [1]. A key limitation of conventional TKA lies in its reliance on the surgeon’s manual skills and visual estimation, which introduces variability in execution and may result in suboptimal implant alignment [2–4]. To mitigate variability attributable to human factors and to achieve standardization and precision in surgery, the application of computer navigation systems and robotic assistance in TKA has become increasingly prevalent and is gaining widespread adoption. Nevertheless, distinct operational differences exist between these two modalities when assisting TKA [5, 6]. On one hand, computer navigation systems establish a spatial mapping relationship between intraoperative skeletal anatomy and an optical tracking system, providing surgeons with quantitative data support beyond direct visual assessment. However, the core functions of navigation systems lie in “guidance and positioning” and “real-time feedback.” The final osteotomy is still performed by the surgeon using a hand-held oscillating saw. On the other hand, robotic systems represent not merely an extension of navigation technology but an upgrade in execution capability. This enhances the consistency of plan execution while minimizing variability at the implementation stage. Furthermore, in current preoperative planning for TKA, whether using navigation or robotic assistance, brands that rely on preoperative CT scans predominantly hold the market share. This is largely because utilizing patient-specific CT data for planning enhances procedural accuracy [7]. However, CT-dependent robotic systems introduce certain unavoidable additional costs, such as increased time consumption, higher economic burden, and risks associated with radiation exposure. Consequently, CT-free robotic systems present a distinct clinical appeal. These systems utilize intraoperative tracking combined with a probe or optical scanner to rapidly capture point cloud data of the patient’s bone, generating a personalized 3D model in real-time for surgical planning.

However, current research on robot-assisted TKA predominantly focuses on comparisons between CT-dependent and CT-free systems, different robotic brands, and varied clinical applications of the same system. Studies directly comparing computer navigation systems and conventional robotic assistance in TKA are relatively scarce. The Smith & Nephew CORI robotic system and the Brainlab Knee3 navigation system share the characteristics of being CT-free and belonging to the same corporate brand. Therefore, a comparison of the clinical efficacy between these two systems offers a degree of homogeneity. Nevertheless, comparative studies directly evaluating these two systems are rarely reported. Consequently, this study retrospectively analyses the clinical outcomes of TKA performed using the Smith & Nephew CORI robotic system and the Brainlab Knee3 navigation system in our department.

Materials and methods

Case selection

Inclusion Criteria: (1) Radiographic confirmation of end-stage knee osteoarthritis with unsatisfactory outcomes from conservative treatment. (2) Scheduled for primary unilateral TKA. (3) Prior to enrollment, all patients provided informed consent, specifically opting for and agreeing to undergo the robotic-assisted surgical procedure as part of this research. Exclusion Criteria: (1) Patients undergoing unicompartmental knee arthroplasty or revision TKA. (2) Cases where robot-assisted surgery failed or required conversion to conventional manual surgery. (3) Patients or their families who declined robot-assisted surgery.

Grouping

A total of 100 patients were enrolled in the study In the CORI group, 50 patients underwent TKA assisted by the Smith & Nephew CORI robotic system. All participants provided voluntary written informed consent prior to enrollment (Table 1).

Table 1.

Demographic characteristics of the two patient groups

Indicator	CORI group	Brainlab group	t/χ2	P
Cases	50	50
Age (year, x̅ ± s)	65.29 ± 7.85	65.79 ± 10.00	0.278	0.782
Sex (M/F)	14/36	16/34	0.019	0.663
Operated side (R/L)	13/37	15/35	0.198	0.656
BMI (kg/m2, x̅ ± s)	25.42 ± 2.20	24.98 ± 2.54	0.926	0.357
K-L grading (III/IV)	11/39	14/36	0.480	0.488
Preoperative HKA	172.51 ± 5.53	173.73 ± 7.54	0.923	0.358
Preoperative VAS score ( x̅ ± s)	4.93 ± 2.78	4.79 ± 1.78	0.300	0.765
Preoperative KSS clinical score ( x̅ ± s)	46.56 ± 8.63	47.62 ± 7.62	0.651	0.517
Preoperative KSS functional score ( x̅ ± s)	43.41 ± 6.93	45.58 ± 6.28	1.641	0.104
Preoperative WOMAC score ( x̅ ± s)	72.47 ± 7.67	74.35 ± 7.54	1.236	0.219
Preoperative knee flexion-extension ROM (°, x̅ ± s)	98.46 ± 10.74	102.52 ± 11.83	1.797	0.075

Surgical technique

All procedures were performed under general anaesthesia. A tourniquet was applied intraoperatively in all cases. Denervation of the patella was performed in all patients. The Smith & Nephew LEGION prosthesis was used in all cases.

CORI group

Following exposure, optical trackers were secured to the distal femur and the proximal tibia, respectively. Registration of various anatomical reference points was then performed sequentially as guided by the robotic system. Upon completion of registration, the surgical plan was verified and adjusted as necessary. Once satisfied with the plan, bone resection was carried out using the robotic burr attached to the system in a stepwise manner. After completing the osteotomies, trial components were inserted to assess knee stability. Upon satisfactory verification, the knee joint cavity was irrigated, and the final prosthetic components were implanted.

Brainlab group

Following exposure, optical trackers were secured to the distal femur and the proximal tibia, respectively. Registration of various anatomical reference points was then performed sequentially as guided by the navigation system. Upon completion, bone modeling and surgical planning were conducted. Once satisfied with the plan, osteotomy was performed stepwise using cutting guides. After each cut, the resection was verified using the navigation system. Upon completion of all osteotomies, trial components were inserted to assess knee stability. Following satisfactory verification, the knee joint cavity was irrigated, and the final prosthetic components were implanted.

Postoperative management

Postoperative care included anti-inflammatory therapy, anticoagulation, and multimodal analgesia. A knee X-ray was obtained on the first postoperative day. Functional knee exercises were initiated under the guidance of a physiotherapist. Patients were instructed to return for regular follow-up visits after discharge.

Observational parameters

The recorded parameters included: perioperative metrics (reference array placement time, registration time, osteotomy time, total operative time, intraoperative blood loss, and incision length); knee function at 3 days and 180 days postoperatively (KSS clinical score, KSS functional score, VAS score, WOMAC score, and range of motion - ROM); postoperative radiographic parameters (including the hip-knee-ankle angle (HKA), lateral tibial component angle (LTC), femoral trochlear component angle (FTC), lateral femoral component angle (LFC), femoral flexion component angle (FFC); as well as complications and patient satisfaction at the final follow-up.

Statistical analysis

Data analysis was performed using SPSS (v. 26.0). Normally distributed measurement data are expressed as mean ± standard deviation (x̄ ± s) and were analyzed via independent-samples t-tests for group comparisons. Count data are presented as rates and were examined for significant differences using the chi-square test or Fisher’s exact test, whichever was appropriate. Statistical significance was defined as a two-sided P < 0.05.

Results

General patient data

The group was composed of 14 male and 36 female patients, with a mean age of 65.29 ± 7.85 years (range: 57–82). Based on the Kellgren-Lawrence (K-L) grading system, the severity of osteoarthritis was categorized as grade III in 11 patients and grade IV in 39 patients. Thirteen surgeries were performed on the right knee and 37 on the left knee. In the Brainlab group, 50 patients underwent TKA assisted by the Smith & Nephew Brainlab Knee3 navigation system. This group included 16 males and 34 females, with a mean age of 65.79 ± 10.00 years (range: 55–76). According to the K-L classification, 14 and 36 patients were assigned grades III and IV, respectively. Surgeries involved 15 right and 35 left knees. Preoperative demographic and clinical baseline characteristics, such as age, gender, surgical side, BMI, and K-L grade, were comparable between the cohorts, exhibiting no statistically significant disparities (P > 0.05). The preoperative HKA angles were 172.51 ± 5.53° and 173.73 ± 7.54° (P > 0.05). Preoperative VAS score, KSS clinical and functional scores, WOMAC score, or knee ROM revealed no statistically significant intergroup differences (P > 0.05) (Table 1).

Operative time, incision length, and intraoperative blood loss in both groups

All patients successfully underwent the planned surgical procedure. The two groups exhibited comparable technical parameters, including incision length and the time required for both reference array placement and registration (P > 0.05). A notable difference was that the osteotomy time in the Brainlab group exceeded that in the CORI group by approximately 3 min, a difference that was statistically significant (P < 0.001). Consequently, the total operative time was approximately 18 min longer in the Brainlab group compared to the CORI group (P < 0.001). Furthermore, the Brainlab group exhibited greater intraoperative blood loss compared to the CORI group (P < 0.05) (Table 2).

Table 2.

Operative Time, incision Length, and intraoperative blood loss in both groups

Indicator	CORI group (n = 50)	Brainlab group (n = 50)	t	P
Reference array placement time (min, x̅ ± s)	3.52 ± 1.71	3.68 ± 158	0.486	0.628
Registration time (min, x̅ ± s)	6.68 ± 2.05	6.81 ± 1.75	0.341	0.734
Osteotomy time (min, x̅ ± s)	10.53 ± 3.29	13.73 ± 1.84	6.003	<0.001
Total operative time (min, x̅ ± s)	92.63 ± 9.93	110.45 ± 20.42	5.549	<0.001
Total incision length (cm, x̅ ± s)	13.42 ± 0.50	13.50 ± 0.74	0.633	0.528
Total blood loss (ml, x̅ ± s)	142.84 ± 41.17	167.41 ± 39.32	3.052	0.003

Comparison of VAS scores, KSS scores, ROM, and WOMAC scores between the two groups

All patients were followed up for at least 180 days. At 3 days postoperatively, patients in the CORI group demonstrated significantly lower VAS and WOMAC scores compared to the Brainlab group (P < 0.05), while exhibiting significantly higher KSS clinical and functional scores (P < 0.05). By the 180-day follow-up, all scores in both groups showed significant improvement compared to the 3-day postoperative values. By the 180-day evaluation, none of the clinical outcome scores demonstrated any significant intergroup disparity (P > 0.05) (Table 3).

Table 3.

Knee pain, function, range of motion, and comfort scores at postoperative day 3 and day 180 in both groups

Indicator	CORI group (n = 50)	Brainlab group (n = 50)	T/χ2	P
VAS score ( x̅ ± s)
3 day postoperatively	4.15 ± 0.81	5.98 ± 0.63	12.610	<0.001
180 day postoperatively	1.11 ± 0.56	1.20 ± 0.36	0.956	0.341
KSS clinical score ( x̅ ± s)
3 day postoperatively	84.53 ± 7.34	79.42 ± 6.73	3.628	<0.001
180 day postoperatively	94.97 ± 8.66	95.73 ± 7.47	0.470	0.639
KSS functional score ( x̅ ± s)
3 day postoperatively	83.42 ± 6.88	78.85 ± 5.98	3.545	0.001
180 day postoperatively	95.35 ± 9.31	93.52 ± 8.39	1.033	0.304
WOMAC score ( x̅ ± s)
3 day postoperatively	61.79 ± 7.1	69.53 ± 7.85	5.171	<0.001
180 day postoperatively	37.87 ± 6.78	35.71 ± 7.48	1.513	0.134
ROM (°, x̅ ± s)
3 day postoperatively	109.85 ± 7.53	101.52 ± 7.78	5.440	<0.001
180 day postoperatively	125.70 ± 9.37	128.84 ± 7.34	1.865	0.065

Comparison of postoperative radiographic outcomes, complication rates, and patient satisfaction between the two groups

Postoperative full-length standing radiographs of the lower limbs and anteroposterior/lateral X-rays of the operated knee were reviewed. The results indicated that the LFC angle was significantly smaller in the CORI group compared to the Brainlab group (P < 0.05). The LTC angle in the CORI group was closer to the target value of 87° (P < 0.05), and the HKA angle was closer to the target value of 180° (P < 0.05). The comparison revealed no significant difference between the groups in terms of the FFC and FTC angle measurements (P > 0.05). The complication rates were 10% in the CORI group and 14% in the Brainlab group (P > 0.05). Complications primarily included muscular calf vein thrombosis and wound drainage. All cases of muscular calf vein thrombosis were managed successfully with therapeutic anticoagulation. Wound drainage cases resolved with continued dressing changes, with no deep infections occurring. No mechanical system-related complications, pulmonary embolism, or other severe adverse events occurred in either group. The in-hospital patient satisfaction rate was 94% in both groups (P > 0.05) (Table 4).

Table 4.

Postoperative radiographic outcomes, complications, and patient satisfaction

Indicator	CORI group (n = 50)	Brainlab group (n = 50)	T/χ2	P
FFC (°, x̅ ± s)	87.80 ± 1.69	87.99 ± 1.74	0.554	0.581
FTC (°, x̅ ± s)	88.67 ± 1.45	88.98 ± 0.65	1.379	0.171
LFC (°, x̅ ± s)	7.57 ± 2.84	9.56 ± 1.44	4.419	<0.001
LTC (°, x̅ ± s)	86.87 ± 1.06	88.72 ± 1.02	8.893	<0.001
HKA (°, x̅ ± s)	178.94 ± 1.09	177.01 ± 1.01	9.184	<0.001
Complication rate [n (%)]	5(10.00)	7(14.00)	0.379	0.538
Satisfaction rate during hospitalization (%)	47/50(94.00)	47/50(94.00)	ns	1.000

Typical cases

See Figs. 1 and 2.

Fig. 1.

A Representative Case (Brainlab Navigation System). The patient was a 77-year-old female with left knee osteoarthritis (K-L III). (A) Preoperative full-length standing radiograph of both lower limbs. (B, C) Preoperative anteroposterior and lateral radiographs of the affected knee. (D) Preoperative osteotomy planning screen. (E) Postoperative anteroposterior radiograph of the knee, showing a FTC angle of 90°. (F) Postoperative lateral radiograph of the knee, measuring a LFC angle of 5.4° and a LTC angle of 87.4°. (G) Postoperative full-length standing radiograph of both lower limbs, measuring a HKA angle of 180° and a FFC angle of 90.0°. (H) Intraoperative photograph showing the operation of the Brainlab Navigation System during total knee arthroplasty

Fig. 2.

A Representative Case (CORI Robotic System). The patient was a 59-year-old female with right knee osteoarthritis (K-L IV). (A) Preoperative full-length standing radiograph of both lower limbs. (B, C) Preoperative anteroposterior and lateral radiographs of the affected knee. (D) Preoperative osteotomy planning interface. (E) Postoperative anteroposterior radiograph of the knee, showing a FTC angle of 89.7°. (F) Postoperative lateral radiograph of the knee, measuring a LFC angle of 4.5° and a LTC angle of 87.5°. (G) Postoperative full-length standing radiograph of both lower limbs, measuring a HKA angle of 179.2° and a FFC angle of 89.5°. (H) Intraoperative photograph showing the operation of the CORI Robotic System during total knee arthroplasty

Discussion

Distinctions between the two systems

The Smith & Nephew CORI system is an imageless, handheld robotic platform that integrates a high-speed burr with real-time kinetic sensors, enabling the dynamic assessment of knee kinematics during surgery. This system constructs a three-dimensional model of the knee based on intraoperative optical scan data, algorithmically generates a personalized osteotomy plan, and provides haptic feedback to constrain the resection boundaries, thereby enhancing procedural safety. Its core advantage lies in its ability to adjust soft tissue tension in real-time, achieving precise gap balancing, which is particularly beneficial for complex deformity cases [8–10]. Furthermore, its imageless design eliminates the need for preoperative CT scans, reducing radiation exposure and associated costs. However, its reliance on the surgeon to manually manipulate the burr may introduce a greater degree of operator-dependent variability. In contrast, the Brainlab Knee3 is an optical tracking-based computer navigation system [11, 12]. It monitors the lower limb mechanical axis in real-time via reference arrays attached to the bones. This system does not require preoperative imaging; instead, it acquires anatomical landmarks through intraoperative registration to calculate ideal osteotomy angles and navigate instrument positioning. Its software allows for adjustments to mechanical alignment principles, prioritising soft tissue balance and potentially reducing the need for extensive releases. However, existing reports consistently indicate that this system is associated with longer operative times. The fundamental distinction between the CORI robotic system and the Brainlab Knee3 system lies in their level of automation. The CORI system enhances osteotomy precision by actively constraining the instrument’s path. In contrast, the Brainlab Knee3 system primarily provides visual guidance to assist the surgeon’s manual execution. Regarding soft-tissue balancing, the real-time kinematic assessment capability of the CORI system surpasses the static gap measurement provided by Brainlab Knee3. Both systems are compatible with the principles of enhanced recovery after surgery. However, existing literature suggests that the CORI system demonstrates an advantage in early postoperative pain control, potentially attributable to its more precise handling of soft tissues.

Clinical efficacy of the two systems

Firstly, the most immediate finding of this study is the comprehensive superiority of the CORI group across multiple perioperative parameters, including shorter osteotomy time, shorter total operative time, and reduced intraoperative blood loss. The authors attribute this difference primarily to the distinct working principles of the two systems. The registration process and the continuous verification steps required for tracking in the Brainlab Knee3 navigation system are relatively complex. In contrast, the CORI system, as an integrated handheld robotic device, enables seamless transition from planning to execution, significantly streamlining the surgical workflow. Furthermore, by employing high-speed milling and active constraint technology, the CORI robotic system facilitates more precise bone surface preparation. This approach potentially minimizes the additional osseous and soft tissue trauma that can occur with the use of a conventional oscillating saw, which is likely the primary reason for the reduced blood loss observed in this group [13, 14].

Secondly, regarding postoperative functional recovery, this study identified a clinically significant temporal dynamic: at 3 days postoperatively, the CORI group demonstrated significantly superior outcomes in knee function scores (VAS, KSS, WOMAC) and ROM compared to the Brainlab group. However, by the 180-day follow-up, these inter-group differences were no longer statistically significant.The authors posit the following reasons for this observation [15–17]: On one hand, the CORI system provides real-time, dynamic assessment and monitoring of soft tissue balance within the knee joint during surgery. In contrast, the Brainlab Knee3 system leans towards static assessment. Consequently, superior intraoperative soft tissue balancing achieved in the CORI group likely contributed to enhanced early postoperative proprioception and patient experience. On the other hand, the assistance provided by the CORI robotic system, characterized by higher precision, a more minimally invasive technique, and reduced soft tissue trauma, offers patients a more favorable early rehabilitation experience. However, over time, the body’s robust capacity for repair and adaptation allows patients in both groups to ultimately achieve a similar plateau of good function. This finding aligns with the conclusions drawn by Kayani et al., which suggest that the advantages of robotic assistance are most pronounced in the early postoperative period.

Thirdly, regarding radiographic outcomes, the CORI group in this study demonstrated superior accuracy in the HKA angle, which reflects lower limb alignment, as well as in the LFC and LTC angles [18–20]. On one hand, regarding the HKA angle, a key metric for lower limb alignment, the CORI system achieves a higher consistency in alignment restoration. This is because it physically executes the osteotomy or milling via direct robotic arm guidance, thereby minimizing errors inherent to manual execution and translating the virtual plan into reality [21, 22]. In contrast, the osteotomy in the Brainlab group relies on the surgeon’s manual operation, which inevitably introduces a degree of variability. On the other hand, regarding sagittal prosthesis positioning, sagittal plane balance of the knee is crucial for flexion function, stability, and patellofemoral joint dynamics. Traditional navigation systems rely more heavily on the surgeon’s experience and judgment when determining sagittal plane parameters such as femoral component rotation and tibial slope. In contrast, the CORI system utilizes its dynamic ligament tension assessment tools to enable personalized sagittal plane planning guided by “gap balancing,” which is then executed with precision by the robot. This is likely the reason for its superior performance in achieving the LFC and LTC angles [23]. This further demonstrates that the advantage of robotic systems lies not only in raising the baseline level of “accuracy” but also in expanding the upper limit of “optimization,” thereby enabling the achievement of more complex, personalized balancing objectives.

Finally, the absence of a difference in complication rates and long-term satisfaction between the two groups [24–26] confirms the high safety profile of both technologies. This also indicates that, while demonstrable differences exist between robotic and navigated techniques, both modalities are capable of delivering safe and ultimately satisfactory surgical outcomes for patients, this is the result of the gradual maturity of artificial intelligence in biomechanics, surgery and other aspects [27].

This study has several limitations. First, the retrospective, non-randomized design introduces the potential for selection bias. Second, the absence of a conventional manual TKA control group limits the ability to quantify the absolute incremental benefit of each technology over the standard technique. Third, the single-center experience and relatively small sample size may affect the generalizability of the findings. Fourth, the follow-up period was limited to 180 days, lacking long-term functional and survivorship data. Furthermore, the potential influence of the surgical learning curve for each technology on perioperative outcomes was not assessed, and a formal cost-effectiveness analysis was not performed.

Conclusion

In conclusion, both the CT-free Smith & Nephew CORI robotic system and the Brainlab Knee3 navigation system demonstrated favourable clinical efficacy and can be considered reliable options in clinical practice. However, the CORI robotic system, with its more minimally invasive operative technique, superior control over limb alignment, and more precise prosthesis positioning, offers patients a better early rehabilitation experience.

Abbreviations

TKA

Total Knee Arthroplasty

BMI

Body Mass Index

CORI

Computer-Assisted Robotic Instrumentation (Smith & Nephew)

FFC

Femoral Flexion Component angle

FTC

Femoral Trochlear Component angle

LFC

Lateral Femoral Component angle

LTC

Lateral Tibial Component angle

HKA

Hip-Knee-Ankle angle

K-L

Kellgren-Lawrence grade

KSS

Knee Society Score

ROM

Range of Motion

VAS

Visual Analog Scale

WOMAC

Western Ontario and McMaster Universities Osteoarthritis Index

Author contributions

HPW and MYW participated in the study design, data collection and analysis, and wrote the main manuscript; YPL and HPW performed the surgical procedures and participated in the perioperative management; XQY , GCM, ZDT and XZS contributed to data curation and statistical analysis; HPW supervised the entire study, including conceptualization, methodology, and final review. All authors read and approved the final manuscript.

Funding

This research was funded by Sichuan Provincial Primary Health Development Research Center in 2024, North Sichuan Medical College (No. SWFZ24-Q-86) and Chengdu High-tech Medical Association “2023 annual Flurbiprofen gel paste treatment of osteoarthritis special research fund” project (2024013), Sichuan Provincial Rehabilitation Medical Association’s 2025 Annual Scientific Research Projects (No.SCKFKY20250210), 2025 Annual Science and Technology Bureau of Panzhihua City project (2025ZD-S-2), 2024 Annual Science and Technology Bureau of Panzhihua City project (2024ZD-S-87), 2023 Annual Science and Technology Bureau of Panzhihua City project ( 2023ZD-S-5) who is not involved in study design, data collection, analysis and interpretation or manuscript preparation.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Ethical approval for this investigation was granted by The Institutional Review Board of Panzhihua Central Hospital (Pankelun Trial No. [2024-008]). Written informed consent was secured from all individual participants included in the study.

Consent for publication

Written informed consent was obtained from the patient for publication of this paper.

Competing interests

The authors declare no competing interests.