Program Information
Comparison of 4DCT Binning Using Deviceless4D and Surface-Mapping Techniques
M Jermoumi1*, D Cao1 , S Zhang2 , V Mehta1 , D Shepard1 , (1) Swedish Cancer Institute, Seattle, WA, (2) Harbin Medical University Cancer Hospital, Harbin, Heilongjiang, China
Presentations
SU-H2-GePD-J(B)-2 (Sunday, July 30, 2017) 3:30 PM - 4:00 PM Room: Joint Imaging-Therapy ePoster Lounge - B
Purpose: We have compared two data-sorting techniques for 4DCT imaging. The first technique is based upon surface imaging where the breathing phase is determined using a virtual tracking point placed on the patient’s surface (SGRT-4DCT). The second technique is a solution from GE called Deviceless4D (DS-4DCT) that eliminates the need for an external surrogate by using a proprietary algorithm based upon internal anatomical features.
Methods: Ten patients with NSCLC were included in this study. Maximum Intensity Projection (MIP) imaging was used to delineate the target along with the edge detection feature of the treatment planning system (Monaco 5.00.03). The tumor positon at each breathing phase was evaluated and compared. Tumor shape features were also compared including the volume, maximum diameter, surface area, sphericity, and eccentricity. To dice coefficient was used evaluate the similarity of the segmented tumor using DS-4DCT and SGRT-4DCT.
Results: Among the ten cases, three cases failed the binning process using the surface-mapping technique due to an irregular breathing pattern. For the remaining cases, no significant difference was observed in the tumor position along all phases (0, 10, 20,30,40,50,60,70,80,90 ) along with the observed tumor motion (P>0.05). The mean volume was 8.14 ±9.6 cm3 and 8.35 ±9.7 cm3 (P=0.52) while using DS-4DCT and SGRT-4DCT, respectively. The tumor volume, maximum diameter, surface area, and sphericity using the both 4DCT binning technique showed strong correlation (r=0.99 and P<0.05). However, the eccentricity feature showed weak correlation (r=0.60 and P< 0.05). The average of dice coefficient of similarity of tumor target definition was 0.88±0.15.
Conclusion: Both DS-4DCT and SGRT-4DCT binning techniques can generate similar 4DCT data set. However, SGRT-4DCT is more dependent on a regular breathing pattern than DS-4DCT. Furthermore, a spatial tumor shape showed high degree of eccentricity while using DS-4DCT compared to SGRT-4DCT.
Contact Email: