Assessment of the influence of metadata parameters on the quality of intraoral images captured via mobile phones: A comparative study using subjective expert ratings and objective ImageJ analysis

Background: Intraoral photography through mobile phones is an important tool in teledentistry, particularly for early detection of oral diseases in resource-limited settings. However, the diagnostic utility of these images hinges on their quality, which is influenced by technical metadata, environmental conditions, and operator variability. Objective: The objective is to evaluate the impact of metadata on intraoral image quality by comparing expert ratings with ImageJ metrics and identifying optimal settings for high-quality capture. Materials and Methods: This cross-sectional study analyzed 30 intraoral images captured with a Samsung Galaxy M15 5G. Ten dental experts rated each image on six quality parameters using a 3-point Likert scale. Objective measurements were obtained via ImageJ, and metadata were extracted using ExifTool. Statistical analysis examined correlations between metadata, objective metrics, and expert ratings. Results: Expert ratings showed good consistency (Cronbach’s α = 0.771) and substantial agreement (Fleiss’ κ = 0.68, 95% confidence interval: 0.62–0.74, P < 0.001). Objective analysis showed mean brightness of 113.00 ± 16.50, contrast 31.77 ± 4.43, and sharpness 2.66 ± 0.58. ISO and brightness values were significantly correlated with image brightness and red, green, and blue (RGB) intensities (P < 0.001), aligning with expert assessments. Brightness value significantly predicted red and blue channel intensities, while no metadata parameter predicted contrast, sharpness, or noise. Ideal image quality corresponded to ISO 20–25, brightness 9.0–10.0, and RGB values of R: 135–150, G: 110–125, and B: 100–115. Conclusion: ISO and brightness metadata significantly influence both objective and subjective image quality. Machine learning can leverage metadata and pixel data to predict diagnostic usability and enable real-time quality control in large-scale screenings.