针对目前关于人眼调节机制理论的争论,为更加准确地研究人眼调节时MRI图像的形态变化特点,应用一种新的数字化图像处理方法提高图像识别效果。利用高分辨的MRI采集同一受试者在调节时的人眼图像,通过最大互信息方法对图像对位频闪,实现图像的匹配定位,便于眼科医师对人眼组织的变化进行定性观察与分析。用最大互信息的配准方法编制的软件处理后的图像可以更加直观准确地便于眼科医师的识别判断,在此基础上可以方便地获得定性变化的信息。结果提示最大互信息的配准方法应用于人眼调节时的MRI图像可以提高临床医师对于图像变化的识别质量,并为图像变化的定性和定量研究提供更好的基础。
A novel digital method was proposed for processing of magnetic resonance images (MRI) pictures of accommodated human eyes to get accurate differentiated information for study of mechanism of accommodation of human eyes. High-resolution MRI pictures of accommodated human eyes were collected and processed by mutual information method to get more recognized information to assist ophthalmologists for quantitative observation and analysis. The MRI pictures processed by the mutual information method were easier to be recognized by the ophthalmologists. Results show that the mutual information method could be accepted as an efficient method to improve the recognized accuracy of MRI pictures of accommodated human eyes and provide better basis for qualitative and quantitative studies.