Kundel H L, Nodine C F
Radiology. 1975 Sep;116(3):527-32. doi: 10.1148/116.3.527.
Ten radiologists were shown a series of 10 normal and 10 abnormal chest films under two viewing conditions: a 0.2-second flash and unlimited viewing time. The results were compared in terms of verbal content, diagnostic accuracy, and level of confidence. The overall accuracy was surprisingly high (70% true positives) considering that no search was possible. Performance improved as expected with free search (97% true positives). These data support the hypothesis that visual search begins with a global response that establishes content, detects gross deviations from normal, and organizes subsequent foveal checking fixations to conduct a detailed examination of ambiguities. The total search strategy then consists of an ordered sequence of interspersed global and checking fixations.
十名放射科医生在两种观察条件下观看了一系列10张正常胸部X光片和10张异常胸部X光片：0.2秒的闪光和无限制的观察时间。从言语内容、诊断准确性和置信度方面对结果进行了比较。考虑到无法进行搜索，总体准确率出奇地高（真阳性率为70%）。自由搜索时，表现如预期那样得到了改善（真阳性率为97%）。这些数据支持了这样一种假设，即视觉搜索始于一种全局反应，这种反应确立内容、检测与正常情况的明显偏差，并组织随后的中央凹检查注视以对模糊之处进行详细检查。然后，总的搜索策略由一系列穿插的全局注视和检查注视组成。
