评价基于虚拟超声模型的腹部超声图像客观化评分方法的评价者内及评价者间一致性。

选取2019年10月北京协和医院超声医学科初入科的参加住院医师规范化培训的学员共11人。对腹部6个标准切面分别设立8个评分要素及明确的评分标准，11名学员经虚拟超声培训后采集图像。将所获图像随机打乱顺序，由2名具备10年以上超声医学工作经验的评价者对上述图像行客观化评分，每张图像评价2次，采用Cohen's Kappa系数比较其评价者内及评价者间一致性。

各超声切面评分要素的评价者内一致性及评价者间一致性普遍较好，总体Kappa值分别为0.854~1.000及0.801~1.000（P均＜0.001）；对于个别评分要素，一致性一般。11名学员经虚拟超声培训后采集图像的评分均为中等偏上。

该客观化图像评分方法对腹部超声图像评价的总分及各评分要素单项评分均非常稳定一致，为超声医师胜任力考核及临床质量控制工作提供了有效的工具方法。

To establish an objective image scoring method for abdominal ultrasound based on virtual ultrasound model, and evaluate the intra- and inter-observer consistency.

Image scoring criteria were set up for the six abdominal standard sections with eight distinct scoring elements. After training with virtual ultrasound model, 11 students collected the images, which were then randomly scrambled. Two experts with more than 10 years of clinical experience in ultrasound medicine scored the above images. Each image was evaluated twice, and the intra- and inter-observer consistency were evaluated with Cohen's kappa value.

Agreement was considered almost perfect, for the overall kappa value was between 0.854-1.000 and 0.801-1.000 (P＜0.05). For some of the scoring elements, there was a moderate degree of consistency. The scores of 11 students after virtual ultrasound training were moderate.

This new abdominal ultrasound image scoring method is highly reliable in terms of the total score and score of each element, which provides a relevant quality control tool for ultrasound operators.