Objective:

To quickly and effectively evaluate the therapeutic efficiency of lung cancer.

Methods:

Semi-automatic segmentation method was proposed for lung cancer in CT perfusion images and the computer-aided diagnosis (CAD) method was applied to evaluate the efficacy of vascular-targeted therapy. First

CT perfusion images of lung cancer in DICOM format were input and the modified distance regularized level set (MDRLS) algorithm was used to segment lung tumors. Then the segmented results were mapped to other CT perfusion images. By selecting the appropriate threshold

vessel-rich areas of the tumor were obtained.

The tumor volume and its vessel-rich areas were calculated respectively. Experiments were conducted with CT perfusion images in five patients before and after therapy. With the calculation of Hausdorff distance

the performance of the segmentation method was analyzed. With the comparison of the tumor volume and its vessel-rich areas before and after therapy

the therapeutic efficacy of vascular-targeted therapy was evaluated for the lung cancer.

Results:

The results showed that the proposed method achieved good segmentation performance and the therapeutic evaluation results also agreed with the pathological data.

Conclusion:

Therefore

the proposed method MDRLS may evaluate the efficiency of vascular-targeted therapy for lung cancer

which will provide the significant reference for the clinical treatment.