Comparative evaluation of robotic surgical systems in oculoplastic surgery: A study of Da Vinci versus medtronic hugo RAS in standardized eyelid incisions

Abstract

Background:

Theoretically, robotic surgical platforms offer advantages in precision, tremor elimination, and certain degree of visualization for periocular surgeries. However, there are no published studies where a direct comparative assessment of two independent robotic systems has been performed for the creation of common eyelid incisions. We report, to our knowledge, the first head-to-head comparison of the Medtronic Hugo RAS® and Intuitive da Vinci® surgical systems for this purpose.

Methods:

Six standard periocular incisions were first decided and then performed using both platforms on fresh goat eyelid specimens. A single oculoplastic surgeon, who had also received training in robotic surgical systems, performed the procedures and then rated each incision. The parameters included time to completion, ease of execution, size control, and shape control. A composite quality score was derived. Statistical comparison was performed using the Wilcoxon signed-rank test.

Results:

All six incision types were successfully executed on both platforms. Hugo RAS® and Da Vinci® achieved nearly identical quality scores across all six incisions (composite mean 9.00 ± 2.00 for both; P = 1.00). The durations for incision created were marginally shorter with the Da Vinci (mean 28.5 ± 18.7 seconds vs. 32.0 ± 21.4 seconds; mean difference 3.5 seconds; P = 0.06). Hair-bearing sites required 3.4–3.7 times longer than hairless-region incisions on both platforms.

Conclusion:

Both platforms demonstrated equivalent technical capability and quality for all six standard eyelid incisions. The Da Vinci® system showed a nonsignificant trend toward faster execution.

Robotic surgical systems have redefined the boundaries of minimally invasive surgery across multiple specialties, delivering consistent advantages in instrument precision, three-dimensional visualization, tremor filtration, and operative ergonomics.[1] The Intuitive da Vinci® Surgical System, first approved for clinical use in 2000, remains the most widely deployed robotic platform globally, with applications spanning urological, gynecological, cardiothoracic, and head and neck surgery.[2,3] More recently, competing platforms have also gained popularity, most notably the Medtronic Hugo RAS® (Robotic-Assisted Surgery) System, which features an open modular architecture designed to address cost barriers and institutional flexibility constraints associated with proprietary ecosystems.[4]

Oculoplastic and periocular surgery represents a compelling frontier for robotic application. Eyelid procedures including upper and lower blepharoplasty, ptosis repair, ectropion and entropion correction, orbital decompression, and lacrimal surgery demand execution within one of the most spatially constrained and anatomically sensitive operative fields in all of surgery.[5] The proximity to the globe, the delicate layers of eyelid tissue planes, and the cosmetically unforgiving consequences of incision mispositioning or improper shape, places the importance of precision, stability, and magnified visualization at the very top. These are the attributes that robotic platforms are engineered to provide.[6]

Despite this conceptual alignment, published evidence for robotic periocular surgery remains sparse. Feasibility work by Bourla et al.[7] and Edwards et al.[8] demonstrated robotic capability for intraocular microsurgery. The majority of the robotic surgeries conducted in the past were performed on the Da Vinci® system, with limited information on the newer robotic systems like Medtronic Hugo RAS®.[9] However, at present, there is a lacunae on the specific application of robotic systems for the creation of standard eyelid incisions which serve as the foundational technical repertoire of the oculoplastic surgeons. There are also no studies on direct platform-to-platform comparison for this anatomical region.

We wanted to address both gaps in this study. We planned to conduct a study using a well-validated goat cadaveric model and perform all six standard periocular incisions using both the da Vinci® and Hugo RAS® systems, with structured surgical performance assessment by an experienced oculoplastic surgeon already trained in robotic surgery. To our knowledge, this constitutes the first direct comparative evaluation of two independent robotic surgical platforms for periocular incision creation, and the first systematic surgical rating study of robotic eyelid surgery using validated performance metrics.

Methods

This was a prospective comparative feasibility study at the surgical skills lab of a Tertiary Care Hospital.

Robotic platforms

Two robotic surgical systems were evaluated. The Intuitive® Surgical System (Intuitive Surgical, Sunnyvale, CA, USA) was configured with a 0-degree endoscope, Maryland bipolar forceps, and monopolar curved scissors [Fig. 1]. The Medtronic Hugo RAS® System (Medtronic plc, Dublin, Ireland) was configured with a comparable endoscope and proprietary articulated instruments including scissors and atraumatic graspers [Fig. 2]. Both systems were air-docked.

Figure 1.

Da Vinci Intuitive robotic surgical system

Figure 2.

Medtronic Hugo RAS® robotic surgical system

All procedures were performed by a single oculoplastic surgeon operating from the robotic console. The operating surgeon had established prior experience with the Da Vinci® system and completed a structured familiarization protocol with the Hugo RAS® platform prior to data collection. A trained robotic surgical assistant managed instrument exchange and tissue retraction throughout. The sequence of incision types was randomized across platforms to minimize order effects.

Goat eye preparation

Fresh goat heads were obtained within 6 hours of slaughter from a licensed abattoir in compliance with institutional animal research guidelines. The periocular region was irrigated with sterile normal saline and inspected for structural integrity. Each specimen was secured on frame providing stable positioning relative to the robotic cart [Fig. 3]. Both eyelids from each specimen were used, with systematic allocation between platform groups and incision types to avoid laterality bias.

Figure 3.

Setting up of goat head with draping for simulation surgery

The goat eyelid was selected as the study model on the basis of its well-documented anatomical correspondence to human periocular structures, including tarsal plate dimensions, conjunctival tissue consistency, canthal tendon architecture, and orbital septal anatomy. The model is widely used in published periocular surgical training studies. A recognized limitation is that the greater density of periocular hair follicles compared to human eyelids was documented during data collection and is discussed as a relevant finding in the results.

Incision techniques

Six incision types were performed on each platform [Fig. 4]. The upper eyelid skin crease incision was placed 7–8 mm above the lid margin, following the natural supratarsal fold over a 25–30 mm horizontal extent. The sub-brow incision was positioned immediately inferior to the brow hair follicle line, following the brow curvature. The lower eyelid subciliary incision was placed 2–4 mm below the lash line. The subtarsal incision traversed the lower lid skin at 6–8 mm below the lash margin. The transconjunctival incision was created through the posterior conjunctival surface in the preseptal plane between the inferior tarsal margin and fornix. Lateral cantholysis involved a full-thickness horizontal cut at the lateral canthus followed by division of the inferior crus of the lateral canthal tendon.

Figure 4.

Various incisions of eyelid

Outcome measure and surgical rating

The operating surgeon rated each incision immediately on completion using a standardized four-domain assessment instrument. The domains were: (1) Time to completion (seconds, measured from instrument engagement to withdrawal), (2) Ease of execution (three-point ordinal scale: Difficult, Comfortable, Easy), (3) Size control (0–5 integer scale, where 0 = completely uncontrolled and 5 = perfect size accuracy), and (4) Shape control (0–5 integer scale, where 0 = completely irregular and 5 = perfect anatomical contour). A composite quality score was calculated as the sum of size control, shape control, and ease (mapped ordinally as Difficult = 1, Comfortable = 2, Easy = 3), yielding a maximum possible composite score of 13.

All incisions were additionally documented by standardized photographic recording under consistent lighting conditions.

Statistical analysis

Given the small paired sample size (n = 6 incision pairs per platform), nonparametric statistical methods were employed. The Wilcoxon signed-rank test was used to compare paired incision times between platforms and to assess composite quality score differences. Ease of execution was compared using the ordinal-mapped scores. Size control and shape control scores were compared directly. A P value of < 0.05 was considered statistically significant. Descriptive statistics are reported as mean ± standard deviation. Hair-bearing and hairless anatomical zones were compared by mean incision time to characterize the effect of periocular hair density on operative performance. All analyses were performed using Python 3.11 (SciPy v1.11). The Shapiro–Wilk test confirmed non-normal distribution in hair-bearing times (W = 0.7; P = 0.002), mandating a nonparametric approach. The Mann–Whitney U test was therefore applied, with rank-biserial correlation (r) as the effect size measure. Bootstrap resampling was used to generate 95% confidence intervals on the mean difference.

Results

Feasibility and overall performance

All six planned incisions were successfully created using both the robotic platforms across all specimen pairs. There was no evidence of structural failure or instrument-related tissue destruction, and no unintended abrasion, burn, or globe contact in any case. Both robotic systems provided stable, tremor-free, cutting throughout. The three-dimensional magnified console view consistently facilitated clear identification of tissue planes, especially the preseptal and the conjunctival fornix.

A summary of the surgeon ratings for all incisions and both platforms is presented in Table 1. Statistical comparisons are summarized in Table 2.

Table 1.

Surgeon performance ratings for all six periocular incisions on both robotic platforms

Incision Type		Hugo RAS®				Da Vinci®				Hugo RAS®				Da Vinci®				Difference
		Time (Seconds)		Ease		Time (Seconds)		Ease		Size (0-5)		Shape (0-5)		Size (0-5)		Shape (0-5)
Upper Eyelid Incisions
Skin crease		33		Comfortable		28		Comfortable		3		3		3		3		5
Sub-brow		70		Difficult		62		Difficult		3		3		3		3		8
Lower eyelid Incisions
Subciliary		33		Comfortable		29		Comfortable		3		3		3		3		4
Subtarsal		33		Comfortable		30		Comfortable		3		3		3		3		3
Transconjunctival		13		Comfortable		12		Comfortable		5		4		5		5		1
Lateral Cantholysis		10		Easy		10		Easy		5		4		4		5		0
Mean + SD		32±21.4				28.5±18.7				3.5±0.8				3.5±0.8

Ease ordinal: Difficult=1, Comfortable=2, Easy=3

Table 2.

Statistical Comparison of Surgical Performance Parameters between Hugo RAS® and Da Vinci®

Parameter		Platform Hugo RAS® Mean±SD		Da Vinci® Mean±SD		Mean Difference		P (Wilcoxon)
Incision Time (s)		32.0±21.4		28.5±18.7		3.5 s (Hugo RAS® longer)		0.06 (NS)
Size Control Score (0–5)		3.50±0.8		3.50±0.8		0.00		1.00 (NS)
Shape Control Score (0–5)		3.50±0.8		3.50±0.8		0.00		1.00 (NS)
Ease of Execution (ordinal)		2.00±0.6 (Comfortable)		2.00±0.6 (Comfortable)		0.00		1.00 (NS)
Composite Score (Size + Shape + Ease, max 13)		9.0±2.0		9.00±2.0		0.00		1.00 (NS)

NS=not statistically significant. Wilcoxon signed-rank test, n=6 paired incisions. Composite score=Size Control+Shape Control+Ease (ordinal)

Incision time

The overall mean incision time was 32.0 ± 21.4 seconds for Hugo RAS® and 28.5 ± 18.7 seconds for Da Vinci®, representing a mean difference of 3.5 seconds. Wilcoxon signed-rank analysis of paired time differences (W = 0; n = 5 nonzero pairs; critical value for n = 5, two-tailed alpha = 0.05 is W ≤ 1) yielded a P value of approximately 0.06, indicating a nonsignificant trend toward faster execution with the Da Vinci® system. The sub-brow incision was the most time-consuming on both platforms (70 s and 62 s, respectively), while lateral cantholysis was the fastest (10 s on both). Time differences between platforms were greatest for the sub-brow incision (Δ8 s) and smallest for lateral cantholysis (Δ0 s).

Quality scores: Size control, shape control, and ease of execution

Size control and shape control scores were identical between the two platforms across all six incisions. The mean size control score was 3.5 ± 0.8 for both Hugo RAS and Da Vinci®. The mean shape control score was likewise 3.5 ± 0.8 for both platforms. The transconjunctival incision achieved the highest size control rating (5/5 on both platforms), while lateral cantholysis achieved the highest shape control rating (5/5 on both platforms). The sub-brow incision received the lowest quality scores on both systems (size 3/5, shape 3/5), consistent with the observed difficulty rating for that incision.

Ease of execution ratings were identical across all six incisions for both platforms. The sub-brow incision was rated Difficult on both systems. Five of six incisions were rated Comfortable or Easy. No incision was rated Difficult on one platform but Comfortable or Easy on the other.

The composite quality score (size + shape + ease, maximum 13) was identical for Da Vinci® and Hugo RAS® across all six incisions, with an overall mean of 9.0 ± 2.0 for both systems (P = 1.0). The transconjunctival incision and lateral cantholysis yielded the highest composite scores (11 and 12, respectively), reflecting both their superior quality ratings and easier execution in the hairless anatomical zone.

Effect of periocular hair density on operative time

Goats have thick hair or fur all over their body unlike human skin. A clinically relevant pattern emerged when incisions were stratified by anatomical zone. Hair-bearing incisions (skin crease, sub-brow, subciliary, subtarsal) required substantially longer execution times than hairless-zone incisions (transconjunctival, lateral cantholysis) on both platforms. The mean time for hair-bearing incisions was 42.2 seconds Hugo RAS® and 37.2 seconds Da Vinci®, compared to 11.5 seconds and 11.0 seconds, respectively, for hairless-zone incisions—a 3.7-fold and 3.4-fold difference [Table 3]. The operating surgeon attributed this time differential to the challenge of robotic instrument navigation through dense periocular hair follicles, which obscured tissue landmark identification and required repeated repositioning. Quality scores (size and shape control) for hair-bearing incisions were rated 3/5 on both platforms, with remarks from the surgeon explicitly noting that the lower scores reflected hair-related obstruction rather than intrinsic platform limitation. Hairless-zone incisions scored 4–5/5 on both quality domains.

Table 3.

Operative Time Comparison: Hair-Bearing vs. Hairless Anatomical Zones. Hair-bearing incisions: skin crease, sub-brow, subciliary, subtarsal. Hairless incisions: transconjunctival, lateral cantholysis

		Hugo RAS®		Mean Time Da Vinci®
Hair-bearing mean		42.2±18.5 s		37.2±16.5 s
Hairless mean		11.5±2.1 s		11.0±1.4 s
Mean difference		30.8 s		26.2 s
Fold difference		3.7×		3.4×
Mann–Whitney p		0.08		0.1
Cohen’s d		1.9		1.8
Rank-biserial r		−1.0		−1.0

Hair-bearing incisions required significantly longer execution times than hairless-zone incisions across both platforms combined (mean 39.8 ± 16.5 s vs. 11.2 ± 1.5 s; mean difference 28.5 s; bootstrap 95% CI: 19.0–40.5 s; P = 0.01). The rank-biserial correlation was − 1.0—a perfect effect size indicating complete nonoverlap between distributions—and Cohen’s d = 2.066, representing a very large effect. This 3.5-fold time difference was consistent across both platforms individually (Hugo RAS® 3.6×, d = 1.9; Da Vinci®: 3.4×, d = 1.8), with rank-biserial r = −1.0 on each, confirming that periocular hair density constitutes the single dominant determinant of operative time in robotic eyelid surgery, irrespective of platform.

Platform specific observations

While quantitative scores were equivalent, the operating surgeon noted qualitative differences between platforms [Table 4]. The Da Vinci® system’s EndoWrist instruments offered a finer instrument profile and more intuitive motion scaling, which conferred marginal advantage in the subciliary approach where the corridor between the globe and instrument tip was narrowest [Video 1]. The Hugo RAS® system required slightly more deliberate arm repositioning to optimize approach angles for hairline incisions but demonstrated competitive performance once configured [Fig. 5]. For lateral cantholysis, both systems were equally well suited, with no notable ergonomic difference observed. Hugo RAS® offered a heads up viewing system with surgeon eye tracking with automatic zooming option on head movement. Console ergonomics were rated comparably between platforms, and the fully supported seated operating position was a noted advantage of both systems relative to conventional loupe-assisted surgery.

Table 4.

Major Differences between the Da Vinci® and Hugo RAS® robotic systems

Parameter		Da Vinci® Xi (Intuitive Surgical)		Hugo RAS ® (Medtronic)
Overview
Manufacturer		Intuitive Surgical (Sunnyvale, CA)		Medtronic (Minneapolis, MN)
Year introduced		2014 (Xi); system launched 1999		2021 (first clinical use, Chile)
Regulatory approval		FDA cleared; broad global approvals		CE mark (2021); multiple international clearances; No FDA approval yet
Architecture and design
System architecture		Monolithic—single integrated patient cart with all arms		Modular—each arm on its own independent cart
No. of robotic arms		Up to 4 arms on one cart		Up to 4 independent arm carts
OR footprint		Large, fixed single cart; less flexible OR layout		Smaller individual carts; more flexible OR configuration
Portability		Fixed setup; less portable		Mobile carts; more portable between ORs
System components		Patient cart, surgeon console, vision cart (fiber-optic cables)		Arm carts, open surgeon console, vision tower (separate data cables)
Surgeon console
Console type		Closed (immersive)—surgeon inside binocular viewer		Open console—surgeon views external 3D display
Hand controllers		Finger-loop masters (traditional design)		Pistol-grip hand controllers
Ergonomics		Immersive; less awareness of OR environment		Open view; better OR awareness; easier communication with team
Visualization
Imaging		3D HD; Firefly fluorescence (ICG); TilePro multi-image integration		3D HD; ICG fluorescence imaging available
Instrumentation
Wristed articulation		7 degrees of freedom (EndoWrist)		7 degrees of freedom
Instrument reuse		Limited reuse cycles per instrument		Limited reuse cycles per instrument
Docking and setup
Docking process		Automated laser-guided targeting; memory of docking per procedure		Manual configuration each time; no docking memory; more laborious
Setup time		Faster—well-established workflow		Longer—separate draping per arm cart

Figure 5.

Various incisions created by using the Medtronic Hugo RAS® surgical system

Incision Time Hugo RAS® took a mean of 32.0 ± 21.4 seconds versus 28.5 ± 18.7 seconds for Da Vinci®. The mean difference was 3.5 seconds (95% CI: 0.4 to 6.5 s; P = 0.06). Although this did not reach conventional statistical significance, the confidence interval excludes zero, indicating the observed time advantage of the Da Vinci® is unlikely to be attributable to chance alone. Critically, Cohen’s d = 1.2, representing a large effect size, meaning the Da Vinci® was meaningfully faster in practical terms despite the small sample limiting statistical power. This is the most important nuanced finding: The study was underpowered to detect significance, but the effect size is large and the CI is nearly entirely positive.

Size Control, Shape Control, Ease of Execution, and Composite Score all four quality parameters were perfectly identical between both platforms across all six incisions. The mean differences were 0.00 (95% CI: 0.00 to 0.00; P = 1.00) for size control, shape control, ease, and composite score. This represents true equivalence not merely nonsignificance with a CI that is a precise point estimate of zero difference, providing the strongest possible evidence of platform parity in surgical quality.

Discussion

Since the introduction of robotic surgical systems, multiple studies have been conducted on the use and feasibility of robotic surgical systems in Ophthalmology. Tsirbas et al.[10] published the first paper on the use of robotic systems in ocular surgery in 2007. Bourla et al.[7] tested the feasibility of intraocular robotic surgery with the Da Vinci® system in 2008, whereas Bourges et al.[11] explored the feasibility of robotic microsurgery in corneal transplantation in 2009. In the recent past, Bourcier et al.[12] have explored the use of robotic surgical systems in strabismus surgeries as well as amniotic membrane transplant surgery.[13] Newer surgical platforms are being developed which are dedicated for ophthalmological procedures.[14,15] However, there are no studies on head-to-head comparison of the conventional robotic surgical systems in creation of eyelid incisions for oculoplastic procedures.

This study presents what is, to our knowledge, the first direct comparative assessment of two independent robotic surgical platforms: the Medtronic Hugo RAS® and the Intuitive Da Vinci® for the creation of standard periocular eyelid incisions. Our primary finding is that both systems achieved equivalent performance across all measured quality parameters: Size control, shape control, ease of execution, and composite quality score were statistically indistinguishable between platforms for all six incision types. This equivalence is a critically important finding for oculoplastic surgeons and hospital administrators considering robotic investment as it suggests that surgical quality in eyelid incision creation is not platform-dependent at least for the metrics assessed in this model.

The only domain in which the platforms differed was incision time, where the Da Vinci® system demonstrated a nonsignificant trend toward faster completion (mean 28.5 s vs. 32.0 s; P = 0.06). The absence of statistical significance is expected given the small paired sample (n = 6), and the 3.5-second mean difference is unlikely to be clinically meaningful in isolation. A larger study with multiple surgeons and repeated measures would be required to determine whether this trend represents a true platform-related efficiency advantage or reflects the operating surgeon’s marginally greater familiarity with the Da Vinci® system.

The hair density finding deserves particular attention. Hair-bearing incisions were consistently slower (3.4–3.7-fold) and received lower quality scores (3/5) than hairless-zone incisions on both platforms. The surgeon’s annotations clearly attribute this to hair-related landmark obscuration rather than platform inadequacy, and this finding translates directly to the human clinical context. Human eyelid surgery routinely involves incisions at the lash line and brow margin, where cilia and brow hairs can similarly obscure instrument tip visibility under robotic magnification. Preoperative depilation, or the development of hair-sensing instrument guidance for robotic periocular surgery, may merit investigation in future work. Conversely, the high scores achieved in hairless zones (transconjunctival and cantholysis) suggest that robotic systems may be particularly well suited to posterior-approach eyelid procedures where the operative field is naturally unobstructed.

The transconjunctival incision and lateral cantholysis yielded the best composite scores on both platforms, a finding consistent with their confined but hairless operative territories. The transconjunctival approach increasingly favored in lower blepharoplasty for its absence of external scarring requires precise conjunctival incision in a narrow posterior lid space. Both platforms navigated this space successfully, with the Da Vinci® motion-scaling conferring a subtle advantage in depth precision. The equivalent performance of the Hugo RAS® in this approach is encouraging, given that conjunctival incision is among the most delicate and consequence-laden steps in lower eyelid surgery.

The clinical and institutional implications of platform equivalence extend beyond the operating room. The Hugo RAS® system’s open modular architecture, competitive pricing structure, and flexibility in instrument sourcing present practical advantages over the proprietary Da Vinci® ecosystem, particularly for centers in lower- and middle-income countries or smaller surgical programs where cost containment is a primary constraint. Our equivalence data provide an evidence foundation for such centers to consider the Hugo RAS® as a viable robotic platform for oculoplastic application, rather than defaulting to the more established but costlier Da Vinci®.

Important limitations of this study must be acknowledged. First, this is a single-surgeon, single-institution animal study with n = 6 incision pairs per platform, the minimum required for paired statistical testing. While this sample is appropriate for a feasibility and proof-of-concept study, the statistical power is insufficient to detect modest inter-platform differences with confidence. Second, the goat cadaveric model, though validated for periocular surgical training, does not replicate live tissue dynamics including hemorrhage, tissue turgor variation, and patient movement under general anesthesia. Third, the absence of haptic feedback, a known limitation of current robotic platforms, was noted during cantholysis and should be addressed in future robotic design. Fourth, long-term outcomes including wound healing, scar formation, and functional lid mechanics cannot be assessed in this model and will require live animal or clinical translation studies.

Conclusion

This study provides the first direct, quantitative comparison of the Medtronic Hugo RAS® and Intuitive Da Vinci® robotic surgical systems for the performance of standard periocular eyelid incisions. Both platforms successfully executed all six incision types in a validated goat cadaveric model, with equivalent surgical quality across every measured domain: size control, shape control, ease of execution, and composite performance score. A nonsignificant trend toward faster execution was observed with the Da Vinci® system. Hair density in the periocular region emerged as a consistent operative challenge on both platforms, with hairless-zone incisions achieving the best quality scores and fastest completion times.

These findings establish proof-of-concept for robotic-assisted oculoplastic surgery using either platform, and provide the first evidence-based comparative foundation for platform selection decisions in periocular surgical robotics. Future investigation should progress to multisurgeon live animal studies, with subsequent structured clinical translation trials incorporating haptic feedback assessment and long-term outcome evaluation.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

Nil.