Lung Image Volume Registration Using the Affine Transformation and Demons Algorithm
A method to register the expiration and inspirationbreath-hold HRCT lung image volumes was presented. Weconsidered that the deformation between the expiration andinspiration of lung can be decomposed into a global affinetransformation and a local deformation. Before registration, wesegmented the lung parenchyma from thoracic HRCT slices.Then, we used a prior anatomy knowledge based method to findthe global affine transformation parameters, and re-sampled aside of lung to align the corresponding one. The Demonsalgorithm was employed to realize local deformation. In theexperiment results, the images after registration exhibit a goodalignment about the lung shape and the inner structures. Thequantitative criterions indicate that the average volume overlapbefore registration is 0.7982, but after the global transformation,it becomes 0.8936 and 0.9544 after Demons. The greatdescending of RMSE shows that the average intensities ofcorresponding voxels become more and more similar. Ourmethod about lung image volume registration can help theanalysis of lung respiration function.
Yingyue Zhou Hongbin Zang
School of Information Engineering Southwest University of Science and Technology MianYang, Sichuan, School of Manufacturing Science and Engineering Southwest University of Science and Technology MianY
国际会议
成都
英文
1-4
2010-06-18（万方平台首次上网日期，不代表论文的发表时间）