1：Experimental Design and Method Selection:

The study compared two main baseline detrending approaches: increasing the high-pass filter cutoff from 1 to 10 Hz and estimating and removing the trend using polynomials of increasing order (1 to 10) applied to different parts of the ERG signal (prestimulus interval, prestimulus and postsignal interval, or whole signal). These were compared to the ISCEV recommended band-pass filter (0.3–300 Hz). Statistical methods included coefficient of repeatability (COR%), unpaired Student's t-test, and area under the receiver operating characteristic curve (AUC).

2：3–300 Hz). Statistical methods included coefficient of repeatability (COR%), unpaired Student's t-test, and area under the receiver operating characteristic curve (AUC). Sample Selection and Data Sources:

2. Sample Selection and Data Sources: Participants included 20 glaucoma patients and 18 age-matched healthy controls recruited from outpatient clinics. Exclusion criteria were visual acuity worse than 6/12, diabetes, recent ophthalmic surgery, and other eye conditions. Glaucoma patients had confirmed glaucomatous disc appearance and visual field defects.

3：List of Experimental Equipment and Materials:

Equipment used included the RETeval device for ERG recording, DTL-like electrode, gold-cup skin electrodes, tropicamide 0.5% for pupil dilation, and an International Light photometer for calibration. Software included Matlab for offline processing.

4：5% for pupil dilation, and an International Light photometer for calibration. Software included Matlab for offline processing. Experimental Procedures and Operational Workflow:

4. Experimental Procedures and Operational Workflow: Pupils were dilated, electrodes were placed, participants were light-adapted, and photopic ERGs were recorded using the RETeval device with 50 red flashes on a blue background. Signals were acquired at 2 kHz sampling frequency and processed offline. Detrending methods were applied, outlier traces were removed using robust principal component analysis, and ERG parameters (a-wave, b-wave, PhNR) were extracted from averaged traces.

5：Data Analysis Methods:

Data were analyzed using COR% for repeatability, unpaired t-tests for group comparisons, and AUC for diagnostic performance. Polynomial fitting and high-pass filtering were implemented in Matlab.