Digital measurement of blepharoptosis using smartphone photography and built-in markup tools: A prospective, blinded methodological comparison study with traditional ruler-based assessment

Abstract

To evaluate the accuracy, consistency, and efficiency of a novel digital measurement method for blepharoptosis grading utilizing iPhone photography and built-in markup tools, compared to the traditional ruler method. In this prospective, blinded methodological comparison study, 145 patients (185 eyes) with blepharoptosis were enrolled. Each patient underwent marginal reflex distance 1 (MRD1) measurement via both the traditional ruler method (control) and the iPhone method (experimental). The iPhone method involved capturing standardized photographs and performing measurements using the device’s native markup feature. Agreement between methods was assessed using the Intraclass Correlation Coefficient (ICC) and Bland-Altman analysis. Grading consistency was evaluated using the Kappa statistic, and operational efficiency was compared via independent t-tests. The mean MRD1 values were 2.18 ± 0.94 mm (ruler) and 2.22 ± 0.89 mm (iPhone), with no statistically significant difference (P = .188). The 2 methods demonstrated excellent agreement, with an ICC of 0.968 (95% CI: 0.956–0.977). Bland-Altman analysis showed a mean bias of −0.04 mm, with 95% limits of agreement from −0.51 mm to 0.43 mm. Clinical grading consistency was almost perfect (Kappa = 0.892). The iPhone method was significantly more efficient, requiring less time than the traditional method (64.8 ± 18.7 s vs 86.5 ± 22.3 s, P < .001). The smartphone-based digital measurement method proves to be an accurate, reliable, and efficient alternative to the traditional ruler method for quantifying MRD1 and grading blepharoptosis. It demonstrates high agreement with conventional techniques while offering advantages in standardization, traceability, and operational workflow. This approach holds significant potential for enhancing diagnostic precision in clinical practice, particularly in outpatient and primary care settings.

1. Introduction

Blepharoptosis, a common ocular condition involving both functional and aesthetic concerns, is characterized by the upper eyelid margin covering the cornea beyond the normal range, potentially leading to visual field defects, compensatory frontalis muscle contraction, and even amblyopia.^[1] A crucial step in clinical diagnosis and treatment is the precise quantification of the degree of ptosis. The marginal reflex distance (MRD1) and other objective indicators are universally used internationally for grading, with the results directly guiding the selection of surgical approaches, such as levator palpebrae superiore muscle resection or conjunctiva-Müller’s muscle resection (CFS suspension).^[2] Therefore, achieving accurate and reliable grading diagnosis critically influences treatment outcomes.

For a long time, clinical practice has primarily relied on manual measurement with a ruler, a method with significant limitations.^[3] During operation, potential sources of measurement error include the ruler’s angle misplacement, parallax effects, slight patient movement, and the examiner’s subjective judgment discrepancies. Given that clinical grading standards use millimeter-level differences as boundaries (consistent with the criteria applied throughout this study: mild ptosis: 1.5–3.5 mm, moderate ptosis: 0–1.5 mm, severe ptosis: MRD1 ≤ 0 mm),^[4] an error of merely 1–2 millimeters can lead to misclassification in the degree of ptosis. This directly affects the formulation of the surgical plan and postoperative results. This subjectivity and imprecision have long been a pressing issue in clinical practice.

In recent years, with advancements in mobile health and digital imaging technology, high-precision smartphone photography and image analysis have introduced new pathways for medical measurement. Devices like the iPhone, equipped with high-performance cameras and built-in measurement tools, enable standardized photography and digital marking for non-contact, traceable, and precise measurements. This demonstrates potential across various clinical specialties to replace traditional measurement tools, opening new avenues for enhancing diagnostic objectivity.^[5,6]

This study aims to propose a digital measurement method for blepharoptosis based on smartphone photography and its built-in markup features. In this study, the iPhone was used as the representative smartphone device for protocol implementation. Through a blinded controlled trial comparing this method against the traditional ruler method, we will systematically evaluate its accuracy, consistency, and clinical applicability. The innovation of this method lies in utilizing the native functions of widely available mobile devices to establish a diagnostic protocol that is easy to perform, yields reliable results, and is readily standardized for broader adoption. This aims to provide a new strategy for achieving more precise grading and treatment of blepharoptosis.

2. Materials and methods

2.1. General information

This study is a prospective, methodological comparison study. A total of 145 patients (185 affected eyes) who met the criteria and visited the Plastic and Aesthetic Surgery Department of our hospital between June 2024 and January 2025 were enrolled. The study protocol was reviewed and approved by the Ethics Review Committee of the Second Affiliated Hospital of University of South China. All participating patients provided written informed consent before photography and measurements. And written informed consent for the publication of all identifiable images was obtained from the participants.

2.2. Inclusion criteria

Patients with an initial clinical diagnosis of congenital or acquired blepharoptosis requiring surgical correction.

Age ≥ 18 years, possessing full civil capacity, and able to cooperate with the examination and photography.

Voluntarily participated in this study and signed the informed consent form.

All eligible patients and their affected eyes included in this study received both the traditional ruler-based and smartphone-based digital MRD1 measurements.

2.3. Exclusion criteria

Presence of eyelid edema, eyelid skin scarring, eyelid tumors, or other structural abnormalities affecting measurement.

Suffering from severe exophthalmos, blepharospasm, or other ocular diseases that might interfere with eye opening and measurement.

Previous history of eyelid or orbital surgery.

Inability to cooperate with the examination and image acquisition due to mental state or other reasons.

2.4. Sample size determination

The sample size of this study was pre-calculated based on the primary endpoint: the Intraclass Correlation Coefficient (ICC) of MRD1 measurements between the smartphone-based digital method and the traditional ruler method. The calculation was performed using PASS 15.0 software (NCSS, LLC, USA), following the standard method for ICC-based consistency study sample size estimation. The pre-specified parameters were as follows: two-sided type I error rate α = 0.05, statistical power (1-β)=90%, expected overall ICC ≥ 0.90 (clinically excellent consistency), null hypothesis ICC ≤ 0.75 (minimum threshold for clinically acceptable good consistency), and a 10% preset loss to follow-up rate. The calculation showed that a minimum of 140 valid eyes were required, and the target sample size was adjusted to 154 eyes after considering the loss to follow-up. This study ultimately enrolled 185 affected eyes from 145 patients, which exceeded the pre-specified target sample size and fully met the statistical power requirements.

2.5. Measurement process

To objectively compare the differences between the 2 measurement methods, this study adopted a prospective, within-subject paired, single-blinded design. All 185 included eyes underwent both MRD1 measurement protocols: the traditional ruler-based measurement (control method) and the smartphone-based digital measurement (experimental method, with the iPhone 15 used as the representative device in this study). The 2 measurement protocols were performed independently for each eye, with the order of measurements randomized via a computer-generated random sequence to eliminate sequence bias, and a minimum 30-minute washout interval between the 2 measurements for the same eye to avoid memory bias. All measurements were completed in a unified clinical examination room with constant indoor lighting (300 lux, no direct natural light) to avoid interference from ambient light changes. The operation time of each measurement was uniformly defined as the duration from the patient sitting in the examination chair and completing head/eye fixation to the final MRD1 value being recorded and confirmed, to ensure a standardized benchmark for efficiency comparison.

2.5.1. Traditional ruler measurement group (control group)

The operators were 2 senior plastic surgeons (both with over 5 years of specialized clinical experience and proficient in blepharoptosis grading standards) who performed all ruler-based measurements independently. The 2 physicians measured each included eye separately, and were blinded to the results of the smartphone-based digital measurements, as well as each other’s measurement results, to minimize inter-operator bias and information crossover.

2.5.1.1. Measurement steps

Positioning and Fixation: The patient sat with the head resting against the headrest of the examination chair to maintain fixation. The eyes gazed straight ahead at a distant fixed light source, ensuring the visual axis was consistent with the horizontal light.

Measurement Operation: The physician stood in front of the patient, gently pressed and fixed the patient’s forehead with the left hand to prevent compensatory eyebrow raising. The right hand held a ruler (precision 1mm), placed the ruler body perpendicular to the eyelid margin, and accurately aligned the “0” mark with the upper eyelid margin. The position of the upper eyelid margin in the natural open-eye state was read. Then, the physician gently lifted the upper eyelid with the left hand and read the position of the corneal light reflex.

Reading and Recording: MRD1 was uniformly defined as the vertical distance from the center of the pupillary corneal light reflex to the highest point of the central upper eyelid margin in the primary gaze position, which is the internationally accepted gold standard for blepharoptosis quantification. The MRD1 value was calculated as the difference between the corneal light reflex scale reading and the central upper eyelid margin scale reading. The measurement was repeated 3 times consecutively with a 10-second interval between each measurement, with the patient re-fixating on the distant light source before each repeat measurement. The average of the 3 measurements was taken as the final result for that physician.

Grading Judgment: Grading was performed based on the average measurement value according to widely accepted clinical standards, which were consistently applied to both measurement methods and all statistical analyses throughout this study: Mild ptosis (1.5~3.5 mm), Moderate ptosis (0~1.5 mm), Severe ptosis (MRD1 ≤ 0 mm).

2.5.2. iPhone measurement group (experimental group) (Fig. 1)

Figure 1.

Digital MRD1 measurement using iPhone markup tool. The photograph shows the clinical application of the iPhone measurement method for Marginal Reflex Distance 1 (MRD1). A standard ruler is positioned vertically lateral to the eye for scale calibration. Using the iPhone’s built-in markup feature, a straight line (yellow) has been digitally drawn from the corneal light reflex (reflection point) to the central upper eyelid margin. The measured distance is −1.5 mm, indicating severe blepharoptosis. Additional annotations show a palpebral fissure height of 5.5 mm and a horizontal measurement of 7.0 mm.

2.5.2.1. Operators

Two other independent plastic surgeons, who did not participate in the ruler-based measurements and were fully blinded to all clinical information and ruler measurement results of the enrolled patients, performed all smartphone-based measurements independently. Both operators had passed the pre-study standardized training and proficiency assessment, ensuring consistent operational standards.

2.5.2.2. Measurement steps

Shooting Preparation: The patient’s position, head fixation, and gaze requirements were fully consistent with the control group: the patient sat with the head firmly resting against the examination chair headrest, the forehead was gently fixed by the operator to prevent compensatory frontalis muscle contraction and eyebrow elevation, and the eyes gazed straight ahead in the primary gaze position. No distant external fixed light source was required. A standard 1 mm precision straight ruler was placed vertically and perpendicularly to the horizontal plane, at the lateral canthus level of the target eye, with the full ruler body visible in the camera frame (no need to align the “0” mark with any anatomical landmark). This ensured all variables other than the measurement tool were controlled between the 2 groups.

Image Acquisition: The operating physician used the rear main camera of the iPhone 15 (iOS 18.0 operating system) for photography, following a standardized protocol: ① The phone was fixed at a perpendicular distance of 30 cm from the patient’s corneal plane, with the built-in grid lines function enabled to align the horizontal grid line with the patient’s bilateral lateral canthus line, ensuring the camera plane was parallel to the patient’s facial plane; ② Auto-exposure and auto-focus (AE/AF) were locked on the pupil area of the target eye to avoid exposure and focus fluctuations; ③ The built-in frontal flash was activated to generate a stable corneal light reflex at the center of the pupil; ④ 2 photos were taken sequentially for each target eye: the first in the natural primary gaze state with complete eyelid opening, and the second with the upper eyelid gently lifted by the operator’s left thumb to fully expose the superior corneal limbus (for supplementary anatomical reference). Each photo was captured 3 times consecutively, and the image with clear focus, complete ruler visibility, centered corneal light reflex, and unobscured upper eyelid margin was selected as the final measurement image by a third-party independent researcher.

2.5.2.3. Image measurement and scale calibration (core standardized protocol)

The selected image was opened in the native “Photos” application of the iPhone, and the “Edit” → “Markup” function was entered for measurement, following a 3-step standardized protocol:

Step 1 Scale Calibration: The “Line” tool was used to draw a straight line along the 10 mm length of the standard ruler in the image, to calibrate the pixel-to-millimeter conversion ratio for the image;

Step 2 Anatomical Landmark Marking: The “Line” tool was used to mark 2 key anatomical points consistent with the control method: the center of the pupillary corneal light reflex, and the highest point of the central upper eyelid margin;

Step 3 MRD1 Calculation: A vertical straight line was drawn between the 2 marked points, and the length of the line was automatically calculated by the markup tool based on the pre-calibrated scale, which was recorded as the MRD1 value of the target eye.

2.5.2.4. Repeatability and grading

To assess the intra-rater reliability of the smartphone-based measurement method and eliminate operator memory bias, the same set of calibrated, de-identified photos was measured 3 times independently by the same physician on the same day, with a minimum 2-hour interval between each repeated measurement. Between assessments, the operator completed measurements for at least 20 other patients’ photos to eliminate memory of the previous marking results. During each measurement, the operator was fully blinded to all previous measurement results, patient clinical information, and traditional ruler-based measurement results to avoid subjective bias. The average of the 3 repeated measurements was taken as the final MRD1 result for the smartphone-based method. Blepharoptosis grading was performed based on this final average value, using the same standardized criteria consistently applied to both measurement methods throughout the study.

2.6. Quality control measures

2.6.1. Personnel training and standardized operations

All physicians involved in the measurements received unified standardized training before the study began. The training included: theoretical explanation (standardized definition of MRD1, blepharoptosis grading standards, principles and step-by-step operational protocols of the 2 measurement methods), practical demonstrations, and pretrial practice on 20 additional blepharoptosis patients (not included in the final study cohort). A proficiency assessment was conducted after the training: physicians were required to achieve an intra-rater reliability ICC ≥ 0.95 for 3 repeated measurements on the same eye, an inter-rater reliability ICC ≥ 0.90 between different operators, and a measurement difference of ≤ 0.5 mm for repeated measurements. Only those who passed all assessment criteria could participate in the formal data collection.

2.6.2. Image quality control

Clear criteria for image inclusion and exclusion were established. Qualified images needed to meet the following: clear focus, unambiguous eyelid contour, clearly visible corneal light reflex, and no obstruction of key anatomical landmarks (e.g., the highest point of the eyelid margin, superior corneal margin) by eyelashes or reflections. After shooting, a third-party researcher not involved in the measurements immediately reviewed the image quality. If the image did not meet the standards, it was retaken immediately to ensure all images used in subsequent analysis met the requirements.

2.6.3. Blinding design and implementation

Strict information isolation was implemented to avoid measurement bias. The operators of the ruler measurement group and the iPhone measurement group were completely independent and unaware of each other’s measurement results. During the data recording stage, all measurement results were coded by dedicated personnel unaware of the group information. Data entry adopted a “double-entry” mode, where 2 researchers independently entered the paper records into an electronic database and performed cross-checking. Any inconsistencies were corrected by tracing the original records, ensuring zero errors in data entry.

2.7. Statistical methods

All data analyses in this study were performed using SPSS statistical software (Version 26.0; IBM Corp.). The sample size calculation was completed using PASS 15.0 software (NCSS, LLC, USA), with detailed parameters and results described in Section 2.1.

Measurement data (e.g., ptosis measurement values, operation time) confirmed to follow a normal distribution by the Shapiro–Wilk test are expressed as mean ± standard deviation (x̄ ± s). Count data (e.g., gender, ptosis grade) are expressed as number (n) and percentage (%).

Accuracy Analysis aimed to evaluate the agreement between the 2 measurement methods based on the within-subject paired design. Firstly, the Intraclass Correlation Coefficient (ICC) was used to analyze the agreement between the smartphone-based digital method and the ruler-based method measurements. An ICC value ≥ 0.75 was considered indicative of good agreement, with ICC > 0.90 indicating excellent agreement. The calculation used a 2-way random-effects model to assess the absolute agreement of the 2 measurement methods. Simultaneously, a paired samples t-test was used to compare whether there was a systematic bias between the mean MRD1 values obtained by the 2 methods for the same eye. A P < .05 was considered statistically significant for the difference. To more intuitively display the range of agreement, a Bland-Altman plot was drawn, calculating the 95% Limits of Agreement (LoA).

Grading Consistency Analysis was used to evaluate the consistency between the 2 methods in clinical grading. Based on the measurement results, blepharoptosis was classified into 3 grades: mild, moderate, and severe. The Kappa test was used to assess the agreement between the grading results of the 2 methods. A Kappa value ≥ 0.75 indicated good grading consistency.

Efficiency Analysis evaluated the convenience of the 2 methods by comparing their operation times. The total time required from measurement preparation to obtaining the final result for each patient was recorded independently (unit: seconds). An independent-samples t-test was used to compare the difference in average time consumption between the 2 methods. A P < .05 was considered statistically significant for the efficiency difference.

Inter-rater and intra-rater reliability of both measurement methods were assessed using the 2-way random-effects model ICC, with ICC ≥ 0.90 indicating excellent reliability.

3. Results

3.1. General characteristics

This study ultimately included 145 eligible patients. Among them, 78 (53.8%) were male and 67 (46.2%) were female. The patients’ age ranged from 18 to 79 years, with a mean age of (47.6 ± 15.3) years. Based on clinical diagnosis types, there were 107 patients (141 eyes) with congenital blepharoptosis and 38 patients (44 eyes) with acquired blepharoptosis. Regarding affected eyes, 105 cases (72.4%) were unilateral and 40 cases (27.6%) were bilateral. (Table 1)

Table 1.

Patient demographic and clinical characteristics (n = 145).

Characteristic	Value/Number (Percentage)
Age (years)
Range	18–79
Mean ± SD	47.6 ± 15.3
Gender
Male	78 (53.8%)
Female	67 (46.2%)
Diagnosis Type
Congenital blepharoptosis	107 (73.8%)/141 eyes
Acquired blepharoptosis	38 (26.2%)/44 eyes
Eye involvement
Unilateral	105 (72.4%)
Bilateral	40 (27.6%)
Total eyes analyzed	185 eyes

3.2. Comparison of measurement values

The mean MRD1 value obtained by the traditional ruler method was (2.18 ± 0.94) mm, while the mean MRD1 value obtained by the iPhone method was (2.22 ± 0.89) mm. A paired t-test comparing the results of the 2 methods showed no statistically significant difference between the measurements (t = −1.32, P = .188). (Table 2)

Table 2.

Comparison of results between the 2 measurement methods (n = 185 eyes).

Measurement Method	MRD1 (mm, x̄ ± s)	95% Confidence Interval	t	p
Traditional Ruler Method	2.18 ± 0.94	2.04–2.32	−1.32	0.188
iPhone Measurement Method	2.22 ± 0.89	2.09–2.35

3.3. Agreement analysis

3.3.1. Intra-rater reliability analysis

The results of the 2 measurement methods showed high agreement, with an Intraclass Correlation Coefficient (ICC) of 0.968 (95% CI: 0.956–0.977, P < .001).

3.3.2. Bland-Altman analysis

The Bland-Altman analysis revealed a mean difference of −0.04 mm between the 2 methods, with 95% Limits of Agreement (LoA) ranging from −0.51 mm to 0.43 mm. The vast majority of data points (182/185, 98.4%) fell within the 95% limits of agreement. (Fig. 2)

Figure 2.

Bland-Altman plot comparing the iPhone and traditional ruler measurement methods for MRD1. The plot illustrates the agreement between the 2 methods for measuring Marginal Reflex Distance 1 (MRD1) in 185 eyes. The solid central line represents the mean difference (bias) of −0.04 mm between the methods. The upper and dashed lower lines indicate the 95% limits of agreement, ranging from −0.51 mm to 0.43 mm. The vast majority of data points (182/185, 98.4%) fall within these limits, demonstrating excellent agreement and supporting the high intraclass correlation coefficient (ICC = 0.968) reported in the results.

3.3.3. Grading consistency analysis

After clinical grading based on the measurement results, the Kappa consistency test indicated a high level of agreement between the grading results of the 2 methods (Kappa = 0.892, P < .001), with over 92% of eyes receiving an identical classification. All observed discrepancies were limited to a difference of a single adjacent grade, with no major disagreements, supporting the clinical reliability of the iPhone method for ptosis grading. (Table 3)

Table 3.

Comparison of grading gonsistency between the 2 measurement methods (n = 185 eyes).

Grading Result	Number of Eyes (n)	Percentage (%)	Notes
Complete Agreement	171	92.4%	Both methods assigned the identical grade (Mild, Moderate, or Severe).
Adjacent Grade Disagreement	14	7.6%	Methods differed by one grade (e.g., Mild vs Moderate).
Mild/Moderate Disagreement	8	4.3%	-
Moderate/Severe Disagreement	6	3.2%	-
Major Disagreement (>1 Grade)	0	0.0%	No cases where methods differed by 2 full grades (e.g., Mild vs Severe).
Statistical Agreement (Kappa)	0.892		Indicates “Almost Perfect” agreement.

3.3.4. Inter-rater and Intra-rater reliability analysis

Both measurement methods demonstrated excellent intra-rater and inter-rater reliability. For the smartphone-based digital method, the intra-rater ICC was 0.982 (95% CI: 0.975–0.987, P < .001), and the inter-rater ICC was 0.974 (95% CI: 0.965–0.981, P < .001). For the traditional ruler method, the intra-rater ICC was 0.951 (95% CI: 0.937–0.962, P < .001), and the inter-rater ICC was 0.938 (95% CI: 0.920–0.952, P < .001).

3.4. Efficiency analysis

The average time consumption for the traditional ruler method (from preparation to completion of measurement) was (86.5 ± 22.3) seconds, whereas the average time consumption for the iPhone method (from preparation to completion of photo measurement) was (64.8 ± 18.7) seconds. An independent-samples t-test showed that the time consumption of the iPhone method was significantly shorter than that of the traditional ruler method (t = 9.24, P < .001). (Table 4)

Table 4.

Comparison of efficiency between the 2 measurement methods (n = 185 eyes).

Measurement Method	Time Consumption (seconds, x̄ ± s)	95% Confidence Interval	t	P
Traditional Ruler	86.5 ± 22.3	83.3–89.7	9.24	<.001
iPhone Measurement	64.8 ± 18.7	62.1–67.5

4. Discussion

This study, through a prospective, double-blind, methodological comparison trial, systematically evaluated the diagnostic efficacy and application value of a new digital measurement paradigm based on smartphone photography and its built-in markup function for quantifying the severity of blepharptosis. The core findings indicate that the smartphone measurement method maintains a high degree of measurement agreement with the traditional ruler method while demonstrating significant superiority in operational efficiency and procedural convenience. Specifically, there was no statistically significant difference between the mean Marginal Reflex Distance 1 (MRD1) values obtained by the 2 methods (P > .05), and the Intraclass Correlation Coefficient (ICC) exceeded 0.95, indicating excellent reliability. Bland-Altman analysis further revealed that over 95% of the difference data points fell within narrow limits of agreement, strongly supporting the smartphone method as an objective and reliable alternative measurement tool with excellent repeatability and measurement precision.

The core advantage of this technique stems from its innovation over the traditional measurement paradigm, primarily reflected in the convenience of the operational workflow, the high repeatability of results, and the effective control of multiple sources of human error. The traditional ruler method, as a contact measurement technique reliant on real-time visual interpretation, is highly susceptible to interference from various factors: parallax error between the operator and the ruler, systematic bias from the difficulty in maintaining the ruler perpendicular to the eyelid margin, and instantaneous random errors introduced by patient blinking, minor movements, or varying cooperation levels. These factors collectively introduce non-negligible measurement variability, limiting the method’s reliability and objectivity.

In contrast, the smartphone method fundamentally circumvents these issues by decoupling the single, dynamic clinical procedure into 2 independent and reviewable steps: “image acquisition” and “static image analysis.” The physician first rapidly captures high-definition eye photographs under standardized conditions, a process taking only seconds and requiring no external light source. Subsequently, repeated observation, zooming, and precise marking can be performed on the static image without time pressure or environmental interference. This “in-frame analysis” mode completely eliminates instantaneous errors caused by physiological blinking, patient movement, or operator hand tremors. It focuses the measurer’s subjective judgment on the accurate identification of anatomical structures within the image, rather than on real-time reading in a dynamic setting, thereby significantly enhancing measurement repeatability and objectivity and establishing a methodological foundation for high-precision diagnostic grading.

Our findings align with the pioneering work by Choi et al,^[7] which first demonstrated the feasibility of smartphone photography for MRD1 measurement in blepharoptosis patients and identified camera and flash position variability as the primary sources of measurement deviation. Our study further extends this methodology with more robust clinical performance in 3 core aspects. First, we achieved superior measurement agreement, with an ICC of 0.968 (95% CI: 0.956–0.977), narrower 95% limits of agreement (−0.51 mm to 0.43 mm) and smaller mean bias (−0.04 mm), compared with the previously reported ICC of 0.85–0.90. This improved precision stems from our strictly standardized shooting protocol, which directly addresses the core deviation sources identified by Choi et al Second, unlike the prior study focusing mainly on shooting parameter impacts, we established a complete clinical workflow using the smartphone’s native markup tools (no additional professional software required), and confirmed almost perfect clinical grading consistency (Kappa = 0.892), verifying its value in guiding surgical decision-making. Third, we first quantitatively demonstrated the significant operational efficiency advantage of this digital method (nearly 25% time reduction, P < .001), with a strict double-blind design and full-process quality control further enhancing result robustness.

The findings of this study have important practical significance. The accuracy of blepharoptosis grading directly influences the choice of surgical procedure (e.g., levator resection vs frontalis sling) and prognosis prediction. An error of merely 1–2 mm can lead to a change in grading. The digital, traceable measurement results provided by the smartphone method offer a more solid and objective basis for clinical decision-making, helping to reduce under-treatment or over-treatment caused by measurement deviations, and ultimately holding the potential to improve surgical success rates and patient satisfaction.

Nevertheless, this study has certain limitations. First, as a single-center study, although the sample size met statistical requirements, a larger multi-center study would further validate the generalizability of this method. Second, although image quality was strictly controlled, it could still be potentially influenced by ambient lighting conditions, slight variations in shooting angle, or a weak corneal light reflex in a very small number of patients. Furthermore, the current method still requires manual marking of key points by the physician, and its measurement accuracy still somewhat depends on the operator’s experience. Future research, building upon the core principles of this method, could develop dedicated mobile applications (APPs) integrated with algorithms for automatically recognizing the eyelid margin and corneal light reflex, achieving fully automated measurement and further eliminating inter-operator variability. Concurrently, organizing prospective, multi-center, large-sample clinical validations will allow for a more comprehensive assessment of the method’s robustness across different populations, devices, and operators, promoting its adoption as part of standard clinical practice.

5. Conclusion

This study validates that utilizing the built-in functions of an smartphone for grading blepharoptosis is an accurate, efficient, and reliable method. It successfully transforms a ubiquitous consumer electronic device into a practical clinical tool, demonstrating the significant potential of digital health in enhancing the standardization and precision of ophthalmic practice. This method is particularly suitable for outpatient screening, primary care settings, and medical education, offering a novel and efficient solution for the standardized diagnosis of blepharoptosis.

Acknowledgments

The authors acknowledge all participants for their contributions to this research. And also extend our sincere gratitude to all participants involved in this study.

Author contributions

Conceptualization: Peihan Tang, Jianfei Zhang, Lan Wang, Bin Jiang.

Data curation: Peihan Tang, Jianfei Zhang, Lan Wang, Kaixi Tan, Yi Wang.

Writing – original draft: Peihan Tang, Lan Wang.

Resources: Jianfei Zhang.

Writing – review & editing: Jianfei Zhang, Feng Yang, Bin Jiang.

Formal analysis: Lan Wang, Ziqing He, Yi Wang, Feng Yang.

Investigation: Kaixi Tan, Ziqing He.

Software: Kaixi Tan, Ziqing He, Yi Wang, Bin Jiang.

Funding acquisition: Feng Yang.