Effect of conspicuity level on dental students’ visual attention to impacted maxillary canines on panoramic radiographs: an eye-tracking study

To quantitatively evaluate the visual attention patterns of dental students when examining panoramic radiographs containing impacted maxillary canines, and to determine how fixation-based metrics, visit-based metrics, and detection rates vary according to the conspicuity level of the impacted canine, defined as the degree to which the canine visually stands out from surrounding anatomical structures. Sixty-four dental students viewed nine panoramic radiographs categorized as high, moderate, or low according to the visual conspicuity of the impacted maxillary canine. Eye movements were recorded using a screen-based eye-tracking system. For polygon-defined areas of interest, fixation durations, fixation counts, visit metrics, and detection rates were calculated. Differences among conspicuity groups were analyzed using repeated-measures ANOVA, gender-based comparisons using independent-samples t-tests, and detection rates using chi-square tests. Conspicuity level significantly affected the major eye-tracking metrics. High-conspicuity canines were associated with longer viewing times, greater total and mean fixation durations, higher fixation counts, and longer visit durations than moderate- and low-conspicuity canines (all p = 0.001; η² p = 0.42–0.58). Detection rates were 84.9% for high-conspicuity, 75.5% for moderate-conspicuity, and 59.3% for low-conspicuity canines ( p = 0.001; Cramer’s V = 0.23). Male students showed higher mean fixation duration and total fixation duration than female students ( p = 0.019 and p = 0.005, respectively; Cohen’s d = 0.62 and 0.74). Dental students’ visual attention patterns while evaluating impacted maxillary canines on panoramic radiographs differed according to target conspicuity. Low-conspicuity canines received less visual engagement and showed lower visual detection rates, indicating that conspicuity is an important determinant of attention allocation during panoramic image evaluation.