Effects of Indirect Pupillary Light Reflex Stimulation With Colored Light on Corneal Specular Microscopy Performance in Dogs Authors: Minju Kim, Hyunwoo Suk, Seonmi Kang, Kangmoon Seo Published: 2026 ABSTRACT ABSTRACTObjectiveTo compare the effects of indirect pupillary light reflex (PLR) stimulation using differently colored lights on corneal specular microscopy performance in dogs and to determine the influence of pupil diameter.Animals Studied20 client‐owned dogs (38 eyes) with healthy corneas.ProceduresIn this prospective study, one eye underwent specular microscopy while the contralateral eye was exposed to white, red, or blue light in a randomized order. The duration of the technique was measured from tracking onset to imaging completion. A 30‐s time limit was imposed to prioritize animal welfare, and trials not completed within this duration were recorded as unsuccessful. Pupil diameter was measured immediately before imaging completion or at 30 s in unsuccessful trials.ResultsBlue light was associated with the highest success rate within 30 s (97.4%), followed by white (93.4%) and red (88.2%) lights. Linear mixed‐model analysis showed that blue light was also associated with the shortest procedural duration and greatest degree of miosis, whereas red light was associated with longer procedural durations and weaker pupillary constriction. After adjusting for light condition, pupil diameter was not independently associated with procedural duration. Because pupil diameter and light condition were closely related, their respective contributions to imaging efficiency could not be fully distinguished.ConclusionsIndirect PLR stimulation with blue light enhanced the efficiency of corneal specular microscopy in conscious dogs. The relative contributions of miosis and other wavelength‐dependent physiological or optical effects remain uncertain.Keywords: blue light, corneal specular microscopy, dogs, indirect pupillary light reflex, miosis, tapetal reflection Introduction The corneal endothelium plays a vital role in maintaining corneal hydration and transparency. Endothelial cell density (ECD) and morphological parameters are important structural indicators of corneal health. In dogs, endothelial dysfunction leads to corneal edema and increased corneal thickness, which can result in significant vision loss [1, 2]. Canine corneal endothelial cells have a regular hexagonal arrangement, and cell density naturally decreases with age: it typically ranges from approximately 2500 cells/mm2 in young dogs to 2100 cells/mm2 in geriatric individuals [3]. Specular microscopy is the clinical standard for the noninvasive evaluation of the corneal endothelium, as it facilitates high‐resolution imaging and rapid quantitative analysis. However, successful examination depends on technical factors, such as autoalignment and autofocusing, and biological variables, including animal cooperation and tear film stability [4, 5]. In veterinary settings, factors such as microsaccades, interference from the third eyelid, and pupil recognition errors frequently cause unsuccessful examinations [4, 5, 6]. Furthermore, performing specular microscopy under general anesthesia is often impractical because the procedure requires stable ocular positioning [6]. Therefore, shorter procedural durations may improve the efficiency and reliability of specular microscopic imaging in conscious dogs. Unlike humans, most dogs possess a tapetum lucidum in the fundus that reflects incident light to enhance nocturnal vision [7]. During specular microscopy, this tapetal reflection can interfere with the autofocus and image recognition algorithms of the device and reduce efficiency [8]. Quantitative evidence regarding the effect of tapetal reflection on procedural success is limited, and standardized strategies to mitigate this potential interference have not been established. Consequently, approaches that modulate fundus reflection, either optically or physiologically, may improve clinical imaging efficiency. Pupil size is the primary factor for regulating the amount of light entering and emanating from the eye. The pupillary light reflex (PLR) is the physiological mechanism by which the iris constricts in response to light stimulation [9]. In dogs, approximately 75% of optic nerve fibers decussate at the optic chiasm, resulting in a pronounced consensual PLR [10]. By utilizing this response, miosis can be induced in the test eye by applying light stimulation to the contralateral stimulated eye without direct illumination. This strategy was hypothesized to reduce the effective area of tapetal reflection while minimizing glare and scattered light artifacts in the test eye. The spectral sensitivity of the pupillary response varies with wavelength. Blue light (approximately 480 nm) effectively stimulates melanopsin‐containing intrinsically photosensitive retinal ganglion cells (ipRGCs) to produce sustained and robust miosis, whereas red light (approximately 630 nm) has a limited effect on this pathway [11, 12]. Therefore, this study aimed to compare the effects of indirect pupillary light reflex stimulation using different wavelengths on corneal specular microscopy performance in conscious dogs. We further evaluated whether pupil diameter was independently associated with procedural duration after accounting for light condition. Materials and Methods Study Population This prospective study enrolled client‐owned dogs with clinically healthy corneas to evaluate the efficiency of corneal specular microscopy during indirect PLR stimulation using different light wavelengths. The participants were recruited from staff‐owned pets at Seoul National University Veterinary Medical Teaching Hospital. Written informed consent was obtained from all owners before participation. Signalment data, including breed, age, and sex, were recorded for each participant. Screening and Exclusion Criteria All dogs underwent a comprehensive ophthalmic examination, including the Schirmer tear test‐1, neuro‐ophthalmic assessment of the menace response and dazzle reflex, intraocular pressure measurement, slit‐lamp biomicroscopy, fluorescein staining, and indirect ophthalmoscopy. Eyes were excluded if they showed corneal abnormalities such as moderate‐to‐severe edema, scarring, or pigmentation that could interfere with endothelial imaging. Dogs were included only if their fundus appeared normal on indirect ophthalmoscopy. Eyes showing retinal degenerative changes such as retinal vessel attenuation were excluded. Dogs with mature cataracts were also excluded to ensure visibility of the tapetum lucidum and allow evaluation of the effect of tapetal reflection. In addition, tapetal status was recorded to allow an exploratory evaluation of the potential influence of fundus reflection on procedural duration. The examinations were immediately discontinued if the dogs showed excessive stress or aggression. Specular Microscopy Protocol Using Differently Colored Lights All procedures were performed in a standardized environment with constant ambient lighting. The stimulus intensities were measured at a distance of 7 cm using a handheld digital lux meter (MT‐912, URCERI, China) and averaged 3025 lx across all conditions: white, 3042 lx; red, 3029 lx; and blue, 3005 lx. Colored light stimuli were delivered using a precision illuminator (BPI‐50, RetinoGraphics Inc., Lawrence, Kansas, USA) equipped with narrowband light‐emitting diodes (blue, 480 nm; red, 630 nm). White light was provided by a 3.5‐V Finnoff transilluminator (Welch Allyn, Skaneateles Falls, New York, USA; 2950 K). Specular microscopy was performed using an EM‐4000 specular microscope (Tomey Corp., Nagoya, Japan). Light stimulation was applied only to the contralateral eye to induce consensual miosis in the test eye. Direct stimulation of the test eye was impractical during imaging because the eye was partially obscured by the device, and supplementary direct illumination could interfere with the reflected slit beam required for specular microscopy [4]. This approach was intended to reduce tapetal reflection, while preventing direct light‐induced artifacts in the imaged eye. Each participant underwent two trial cycles under each light condition (white, red, and blue). The order of the light stimuli was randomized for each dog using a computer‐generated randomization table. To minimize residual miosis or fatigue, a 1‐min rest interval was allowed between eye changes and light conditions, and a 5‐min rest interval was allowed between trial cycles. The procedure was repeated when the device failed to achieve focus, when performance was compromised by ocular movement or eyelid interference, or when fewer than 150 cells were analyzed per image [13]. Outcome Measures The procedural duration was defined as the interval from the onset of tracking to the appearance of the imaging completion signal on the display. To prioritize animal welfare and minimize prolonged ocular surface exposure during forced eye opening, a maximum procedure duration of 30 s was predefined, and trials not completed within this interval were classified as unsuccessful [14, 15]. Pupil diameter was quantified by analyzing video recordings of the specular microscopy display. The last video frame immediately preceding the appearance of the imaging completion signal was selected and analyzed using ImageJ software (Bethesda, Maryland, USA). The normalized pupil diameter was calculated by fitting an ellipse to the pupil margin and expressing the major axis relative to the frame height. All pupil diameter measurements were performed once by a single observer using the predefined video frame selection criteria. Intraobserver repeatability was not formally assessed. This normalized value was converted to millimeters using a screen‐referenced calibration factor. In unsuccessful cases, the pupil diameter was measured at the 30‐s time point. Hippus was predefined as a repetitive, spontaneous fluctuation in pupil size under continuous light stimulation after initial pupillary constriction [16, 17]. It was considered present when at least two observable cycles of narrowing and re‐enlargement were identified without obvious eyelid interference or marked changes in gaze direction. Persistent miosis without repeated oscillation was not considered hippus. Because no universally accepted threshold for hippus assessment exists under these examination conditions, hippus was recorded as an observational finding rather than a definitive diagnostic or mechanistic outcome [16]. Among the parameters provided by the instrument, the corneal ECD, coefficient of variation in cell area (CV), and percentage of hexagonal cells (HEX) were selected as the primary morphological indices for analysis [5, 13, 18]. Statistical Analysis All statistical analyses were performed using R software (version 4.5.2) with the lme4, lmerTest, and emmeans packages. Linear mixed models were used to analyze the procedural duration and pupil diameter, with dog included as a random effect to account for repeated measurements within individuals. Trials in which image acquisition was not completed within 30 s were classified as unsuccessful and excluded from the linear mixed‐model analysis of procedural duration, rather than being assigned a duration of 30 s, because the primary analysis was based on completion within the predefined examination period. Observations from dogs without a tapetum lucidum were excluded from the primary analyses of pupil diameter and procedural duration because the principal hypothesis concerned the effects of light stimulation on imaging in eyes with a reflective tapetum lucidum. Before fitting the final model for the procedural duration, the effects of test eye (left vs. right) and trial order (first vs. second) were assessed in a preliminary linear mixed model. Neither variable was significantly associated with the procedural duration in the preliminary model, and both were excluded from the final model for parsimony. The final procedural duration model included pupil diameter and light condition as fixed effects, whereas the pupil diameter model included light condition as a fixed effect. The overall effect of light condition was first evaluated in each model followed by Tukey‐adjusted post hoc pairwise comparisons using estimated marginal means. Binary outcomes, specifically the occurrence of hippus and procedural success, were evaluated using generalized linear mixed models (GLMMs) with a binomial distribution, with dog included as a random effect. Because hippus was absent under blue light, its occurrence was analyzed only under white‐ and red‐light conditions; light condition and trial order were included as fixed effects. A separate binomial mixed model assessed whether hippus was associated with procedural success within the 30‐s limit, treating hippus status, light condition, and trial order as fixed effects. The effects of light conditions on the corneal endothelial morphological indices (ECD, CV, and HEX) were evaluated using repeated‐measures analysis of variance, with dog included as an error term. Only complete repeated‐measures cases with available data under all three light conditions were included. Statistical significance was set at p < 0.05. Results Demographics Twenty client‐owned dogs (38 eyes) were included. Breeds included Poodles (n = 5), Bichon Frises (n = 3), Maltese (n = 3), and one each of Beagle, Pomeranian, Dachshund, Chihuahua, and Shih Tzu, along with mixed breeds (n = 4). The mean age of the dogs was 8.8 ± 3.7 years (range, 1–13 years). Of the 40 eyes initially examined, two were excluded: one eye of a Maltese dog had a pre‐existing corneal opacity that precluded successful specular microscopy, and one eye of a Shih Tzu dog had conjunctival and corneal pigmentation secondary to medial canthal trichiasis. Accordingly, the data from 38 eyes of 20 dogs were included in the final dataset. One eye with severe congenital iris atrophy was retained despite the absence of a visible pupillary light reflex because the cornea was sufficiently clear for inclusion in the study and did not meet the predefined exclusion criteria. In addition, two dogs (one Pomeranian and one Dachshund) lacked the tapetum lucidum upon fundus examination. Because the study hypothesis concerned imaging performance in eyes with a reflective tapetum lucidum, these dogs were excluded from the primary mixed‐model analyses and evaluated descriptively. Data Pre‐Processing and Model Simplification For the analysis of procedural duration, observations from dogs without a tapetum lucidum and trials that did not result in successful imaging within the 30‐s protocol limit were excluded, yielding a final dataset of 188 valid observations. To evaluate whether this protocol adequately minimized order‐related effects, the influence of the test eye (left vs. right) and trial order (first vs. second) on the procedural duration was assessed. In a linear mixed model with dog as a random effect, neither test eye (p = 0.192) nor trial order (p = 0.252) was significantly associated with procedural duration. Accordingly, both variables were excluded from the subsequent models for parsimony, and the final models focused on light conditions and pupil diameter. Changes in Pupil Diameter According to Light Condition Linear mixed‐model analysis showed a significant overall effect of light conditions on pupil diameter (F = 69.755, p < 0.0001; Table 1, Figure 1). Using blue light as the reference, the pupil diameter increased by a mean of 0.80 mm under white light and 1.46 mm under red light (both p < 0.0001). These findings indicated that blue light was associated with the greatest degree of miosis among the tested conditions. Post hoc comparisons confirmed significant differences among all three conditions, with the pupil diameter increasing in the order of blue, white, and red. TABLE 1Effect of light condition on pupil diameter.Variable/ComparisonEstimateSEdfTest statistic p Light condition, overall——2, 168.11F = 69.755< 0.0001White vs. Blue0.8050.121168.09t = 6.671< 0.0001Red vs. Blue1.4550.124168.37t = 11.757< 0.0001Red vs. White0.6510.124168.11t = 5.241< 0.0001Abbreviations: df, degrees of freedom; SE, standard error. Effect of light condition on pupil diameter. Abbreviations: df, degrees of freedom; SE, standard error. Light color significantly affected pupil diameter of the test eye. Boxplots showing pupil diameter under blue, white, and red light conditions in dogs (n = 18). Pupil diameter increased in the order of blue, white, and red. Boxplots represent the median, interquartile range, and range. Results were analyzed using a linear mixed model with Tukey post hoc tests. Imaging Success Rate and Procedure Duration The imaging success rate within the 30‐s clinical limit varied according to light condition. Blue‐light stimulation yielded the highest success rate at 97.4% (74/76 trials), followed by white light at 93.4% (71/76 trials) and red light at 88.2% (67/76 trials). Most unsuccessful trials occurred under red light. These unsuccessful trials were observed in eyes with features such as pupillary hippus, absent or severely impaired PLR, or persistently large pupil diameter despite prolonged stimulation. In several cases, successful imaging was achieved only after the predefined 30‐s limit, suggesting that this limit served as a pragmatic clinical cutoff rather than an absolute threshold (Table S1). When the valid procedural duration was modeled as the dependent variable, with pupil diameter and light condition as fixed effects, the light condition showed a significant overall effect (F = 6.508, p = 0.0019; Table 2, Figure 2). Compared with blue light, white and red light were associated with procedural durations that were 2.59 and 2.08 s longer, respectively. In contrast, the pupil diameter was not independently associated with procedural duration after adjusting for light conditions (p = 0.328). Post hoc analysis showed that the procedural duration associated with blue‐light stimulation was significantly shorter than that associated with white‐ and red‐light stimulation. No significant difference was observed between white‐ and red‐light conditions. Collectively, these results indicated that blue light provided the most efficient imaging conditions, yielding both the highest success rate and shortest procedural duration. The examination procedure under blue‐, red‐, and white‐light stimulation is demonstrated in Videos 1, 2, 3. TABLE 2Factors affecting duration of corneal specular microscopy.Variable/ComparisonEstimateSEdfTest statistic p Pupil diameter, mm0.3130.31867.54 t = 0.9860.3275Light condition, overall——2, 175.17 F = 6.5080.0019White vs. Blue2.5850.729182.19 t = 3.5440.0014Red vs. Blue2.0820.840174.68 t = 2.4780.0375White vs. Red0.5040.732178.77 t = 0.6880.7707Abbreviations: df, degrees of freedom; SE, standard error. Factors affecting duration of corneal specular microscopy. Abbreviations: df, degrees of freedom; SE, standard error. Blue light reduced procedural duration of corneal specular microscopy in dogs. Boxplots showing procedural duration under blue, white, and red light conditions in dogs (n = 18). Procedural duration was shortest under blue light. Boxplots represent the median, interquartile range, and range. Results were analyzed using a linear mixed model with Tukey post hoc tests. Corneal specular microscopy by blue‐light stimulation, followed by successful image acquisition and endothelial cell analysis. Video content can be viewed at https://onlinelibrary.wiley.com/doi/10.1111/vop.70259. Corneal specular microscopy by red‐light stimulation, followed by successful image acquisition and endothelial cell analysis. Video content can be viewed at https://onlinelibrary.wiley.com/doi/10.1111/vop.70259. Corneal specular microscopy by white‐light stimulation with unsuccessful endothelial imaging. Although the examination proceeded to the Analyze step, a usable endothelial image was not obtained. Video content can be viewed at https://onlinelibrary.wiley.com/doi/10.1111/vop.70259. Association Between Pupil Diameter and Procedural Duration Visual inspection of the scatterplot suggested that procedural duration tended to increase with increasing pupil diameter within each light condition (Figure 3). Pupil diameter was not independently associated with procedural duration after adjustment for light color. Scatterplots with linear regression lines show the relationship between pupil diameter and procedural duration under blue, white, and red light conditions. Shaded areas indicate 95% confidence intervals. Data were analyzed using a linear mixed model. Exploratory Analysis of Hippus Occurrence No hippus events were observed under blue light (0/76 trials). Because this caused complete separation, inferential analysis was restricted to white‐ and red‐light conditions. In a mixed‐effects logistic regression model, red light was associated with higher odds of hippus occurrence than white light (estimate = 2.59, SE = 0.88, p = 0.003), whereas trial order was not significant (p = 0.159). In a separate model adjusted for light condition and trial order, hippus was significantly associated with lower imaging success within 30 s (estimate = −3.48, SE = 1.43, p = 0.015), whereas light condition and trial order were not significant. Comparison of Corneal Endothelial Parameters Across Light Conditions To determine whether light conditions affected corneal endothelial morphological indices, repeated‐measures analysis of variance (ANOVA) was performed for complete cases with data available under all three light conditions. Corneal ECD did not differ significantly among light conditions (F = 0.605, p = 0.552). Similarly, neither CV nor HEX differed significantly among conditions (CV: F = 0.160, p = 0.853; HEX: F = 0.120, p = 0.887). These results indicated that the endothelial measurements remained comparable across the three light conditions (Figure 4). Representative corneal specular microscopy images obtained under different light conditions. Representative images from the same eye of Dog 9 during trial 1 are shown from top to bottom: (A) blue‐, (B) white‐, and (C) red‐light stimulation of the contralateral eye. Corresponding trial data, including pupil diameter and procedural duration, are provided in Table S1. Discussion Despite controlling for illuminance across all conditions, significant differences in pupil diameter and specular microscopy efficiency were observed among the spectrally distinct light conditions. Small fluctuations in the distance between the light source and the eye may have increased data variability, but the overall pattern of our findings was consistent with previous reports suggesting that pupillary responses in clinical settings are more closely related to melanopic‐weighted stimulation than to photopic illuminance alone [19]. From a physiological perspective, the blue light used in this study (approximately 480 nm) likely induced stronger miosis because its wavelength overlaps with the peak spectral sensitivity of melanopsin‐containing ipRGCs, whereas red light (approximately 630 nm) has relatively limited overlap with this region [20]. These spectral differences provide a plausible physiological basis for the greater miosis observed with blue‐light stimulation [20]. Furthermore, activation of the melanopsin/ipRGC pathway is known to contribute to sustained pupillary constriction [11, 20]. A more sustained pupillary response may therefore have contributed to imaging efficiency under blue‐light stimulation; however, this mechanism was not directly evaluated in the present study and should be regarded as hypothetical [21]. Additional analyses of hippus further suggested that light conditions may affect pupillary stability and imaging performance. These findings indicate that wavelength‐dependent differences in indirect PLR behavior could influence imaging efficiency, particularly in eyes with reflective ocular backgrounds. Notably, red‐light stimulation produced weaker and more variable pupillary responses. During indirect PLR evaluation, hippus was observed more frequently under red light, even in dogs with otherwise normal funduscopic findings. In an exploratory mixed‐effects analysis, hippus was not observed under blue light, was more likely under red than white light, and was associated with reduced imaging success after adjustment for light conditions and trial order. Although the present study was primarily designed to assess miosis and imaging efficiency rather than dynamic PLR characteristics, these observations suggest that response stability, in addition to constriction magnitude, may be relevant when interpreting wavelength‐dependent pupillary responses in clinical settings [19, 20, 21]. Methodologically, pupil diameter was selected as the principal analytical metric, which was consistent with previous studies [22]. Area‐based quantification was considered less suitable in this clinical setting because maintaining a perfectly standardized imaging geometry was not feasible in conscious dogs. In two‐dimensional video‐based analysis, the pupil area can be disproportionately affected by eye movement and image angle. In addition, the diameters derived from ellipse fitting may be subject to systematic measurement errors [23], which should be considered when interpreting absolute values. In the present study, pupil diameter was measured from the last video frame immediately preceding the appearance of the imaging completion signal or, in unsuccessful cases, at the 30‐s time point. This approach was intended to provide a standardized measure of the pupillary state most relevant to the imaging procedure. Hippus observed during red‐light stimulation may also have introduced additional variability into pupil diameter assessment, because repeated cycles of constriction and dilation can make the identification of a single representative pupil size more challenging [11]. Blue light was associated with the greatest miosis and shortest procedural duration. Although this supports a potential role for miosis in imaging efficiency, the respective contributions of pupil constriction and light wavelength could not be fully distinguished because pupil diameter and light condition were closely related. Moreover, the lack of an independent association between pupil diameter and procedural duration in the multivariable model does not exclude a clinically relevant contribution of miosis. Additional wavelength‐dependent physiological or optical factors may also have contributed [19, 20], although these mechanisms were not directly evaluated in the present study. One possible explanation is that blue‐light stimulation may have improved imaging through wavelength‐dependent effects on the consensual pupillary response, which could indirectly facilitate autofocus by reducing the exposed tapetal area in the test eye. However, pupil diameter alone did not account for the between‐condition differences after light condition was included in the model. Only two dogs without a tapetum lucidum were included, and this interpretation should therefore be considered hypothesis‐generating rather than definitive. Nevertheless, the observed pattern raises the possibility of an interaction between tapetal reflection and wavelength‐dependent pupillary responses, although this hypothesis could not be directly tested in the present study. Importantly, the use of different light conditions did not significantly alter the measured corneal endothelial morphological indices, including ECD, CV, and HEX [5, 13, 18]. These results indicate that endothelial measurements remained comparable across the three light conditions despite improved imaging performance with blue light. Clinically, these findings suggest that imaging efficiency can be optimized without systematically altering the principal morphological parameters evaluated in this study [5, 13, 18]. However, ECD, CV, and HEX reflect endothelial morphology rather than image quality. Because no standardized assessment of endothelial image quality was performed, it remains unclear whether blue‐light stimulation improved only acquisition efficiency or also the quality of the acquired endothelial images. This study has several limitations. First, measurements were performed in conscious dogs without sedation or eyelid speculum placement to preserve clinical applicability and minimize invasiveness [5]. Consequently, movement of the dogs, blinking, and variable eyelid exposure may have introduced measurement variability. Second, light intensity was standardized using photopic illuminance in lux rather than spectral irradiance or a melanopic‐weighted metric. Although the nominal wavelengths of the stimuli were specified, equal photopic illuminance did not necessarily indicate equivalent spectral irradiance or melanopic stimulation across light conditions. Therefore, the retinally effective stimulus intensity could not be quantitatively compared across wavelengths. Third, pupil diameter was measured immediately before imaging completion, reflecting the pupillary state most relevant to image acquisition rather than maximal physiological constriction. Furthermore, age‐related lenticular changes (e.g., nuclear sclerosis) may have attenuated tapetal reflection and thereby reduced the dependence of successful imaging on full miosis. These factors, together with measurement variability and limited statistical power, may have contributed to the lack of an independent association between pupil diameter and procedural duration. Intraobserver repeatability and measurement error were not formally evaluated for the video‐based pupil diameter measurements, which may have introduced additional measurement variability. Fourth, retinal function was not evaluated using electroretinography, and subject inclusion relied solely on funduscopic appearance; therefore, subclinical retinal dysfunction cannot be ruled out. Fifth, trials exceeding the predefined 30‐s limit were excluded from the analysis of procedural duration. Because unsuccessful trials occurred more frequently under red‐light stimulation, their exclusion may have affected the comparison of procedural duration among light conditions. Therefore, procedural duration should be interpreted together with the imaging success rate. Review of these unsuccessful trials suggested that their failure may have been multifactorial, with possible contributors including severe iris abnormalities and pupillary hippus that may have destabilized the indirect PLR. Hippus was assessed using observational criteria rather than quantitative pupillography and should therefore be regarded as a descriptive variable rather than a definitive mechanistic outcome. Moreover, successful imaging after 30 s in some eyes suggests that this limit was a pragmatic stopping point rather than an absolute threshold (Table S1). Future studies could consider time‐to‐event approaches that account for right‐censored observations. Sixth, only dogs with clinically healthy corneas were included. Therefore, the present findings may not be directly generalizable to dogs with corneal disease, particularly corneal edema or opacity, because reduced corneal transparency and altered optical properties may independently impair autofocus and image acquisition. Finally, only two dogs without a tapetum lucidum were included, precluding a meaningful comparison according to tapetal status. Therefore, the present study could not determine whether the beneficial effect of blue‐light stimulation was dependent on tapetal reflection, and the proposed tapetum‐related mechanism should be regarded as hypothesis‐generating. Safety considerations are also important when applying blue light in a clinical setting [24]. Short‐wavelength light carries a greater theoretical risk of photochemical retinal injury than long‐wavelength light [25, 26]. Therefore, blue light stimulation should be used judiciously. Future studies should establish safe and standardized clinical protocols linked to evaluations of stimulus intensity, exposure duration, and cumulative retinal light exposure in dogs [24, 25, 26]. Taken together, these findings suggest that indirect induction of the pupillary light reflex using blue light in the contralateral stimulated eye may improve the efficiency of corneal specular microscopy in the test eye of conscious dogs. Among the tested conditions, blue‐light stimulation was associated with the highest imaging success rate and the greatest degree of miosis. However, because pupil diameter and light condition were closely related, the respective contributions of miosis and wavelength‐dependent effects could not be fully distinguished. Additional physiological or optical mechanisms may have contributed, but these possibilities remain hypothetical and require further investigation [19, 20]. Blue‐light stimulation was associated with improved procedural efficiency without significant differences in the principal corneal endothelial morphological parameters [5, 13, 18]. This approach may therefore be useful for optimizing specular microscopy in dogs. Whether similar benefits occur in other tapetum‐bearing species remains to be determined. Author Contributions Seonmi Kang: methodology, writing – review and editing, supervision, validation. Minju Kim: investigation, visualization, writing – original draft, software, data curation, methodology, formal analysis. Kangmoon Seo: conceptualization, funding acquisition, writing – review and editing, supervision, validation. Hyunwoo Suk: data curation, methodology, investigation. Disclosure Artificial Intelligence Statement: The authors have not used AI to generate any part of the manuscript. Ethics Statement The study protocol was approved by the Institutional Animal Care and Use Committee of Seoul National University. All procedures were non‐invasive, and written informed consent was obtained from all owners prior to the inclusion of their dogs in the study. The study protocol was approved by the Institutional Animal Care and Use Committee of Seoul National University (SNU‐IACUC; No. SNU‐260121‐7). Conflicts of Interest The authors declare no conflicts of interest. Supporting information Table S1: Raw data on animal characteristics (breed, age, and sex), eye and tapetum status, and trial‐level measurements (procedural duration, pupil diameter, and hippus) from two repeated trials (T1 and T2) under blue‐, white‐, and red‐light stimulation. Acknowledgments This study was partially supported by the Research Institute for Veterinary Science, Seoul National University. Data Availability Statement The data that support the findings of this study are available in Table S1 of this article. References Associated Data Supplementary Materials Table S1: Raw data on animal characteristics (breed, age, and sex), eye and tapetum status, and trial‐level measurements (procedural duration, pupil diameter, and hippus) from two repeated trials (T1 and T2) under blue‐, white‐, and red‐light stimulation. Data Availability Statement The data that support the findings of this study are available in Table S1 of this article.