TY - JOUR
T1 - Multi-institutional quantitative evaluation and clinical validation of Smart Probabilistic Image Contouring Engine (SPICE) autosegmentation of target structures and normal tissues on computer tomography images in the head and neck, thorax, liver, and male pelvis areas
AU - Zhu, Mingyao
AU - Bzdusek, Karl
AU - Brink, Carsten
AU - Eriksen, Jesper Grau
AU - Hansen, Olfred
AU - Jensen, Helle Anita
AU - Gay, Hiram A.
AU - Thorstad, Wade
AU - Widder, Joachim
AU - Brouwer, Charlotte L.
AU - Steenbakkers, Roel J.H.M.
AU - Vanhauten, Hubertus A.M.
AU - Cao, Jeffrey Q.
AU - McBrayne, Gail
AU - Patel, Salil H.
AU - Cannon, Donald M.
AU - Hardcastle, Nicholas
AU - Tomé, Wolfgang A.
AU - Guckenberg, Matthias
AU - Parikh, Parag J.
N1 - Funding Information:
Conflict of interest: Karl Bzdusek, PhD, is employed by Philips Healthcare. Hubertus A.M. Vanhauten, MD, participated in an advertising review for Ferring and received lecture payment from Astellas. Nicholas Hardcastle, BSc (Hons), PhD, was employed under a Philips Radiation Oncology Systems grant awarded to Professor Wolfgang Tomé at the University of Wisconsin—Madison . Wolfgang A. Tomé, PhD, FAAP, has received research grants from the NCI, Philips Radiation Oncology Systems, and Accuray; payment for patents (planned, pending, or issued) from WI Alumni Research Foundation; royalties as sole author of a monograph in mathematical physics titled Path Integrals on Group Manifolds; and serves on Scientific Advisory Board for ViewRay, Inc. Parag J. Parikh, MD, has received grants from ViewRay, Inc, and Philips Healthcare; equipment and technical support from ViewRay, Inc; grants/grants pending from Varian Medical, Philips Healthcare; and payment for lectures including service on speakers bureaus from Varian Medical.
PY - 2013/11/15
Y1 - 2013/11/15
N2 - Purpose: Clinical validation and quantitative evaluation of computed tomography (CT) image autosegmentation using Smart Probabilistic Image Contouring Engine (SPICE). Methods and Materials: CT images of 125 treated patients (32 head and neck [HN], 40 thorax, 23 liver, and 30 prostate) in 7 independent institutions were autosegmented using SPICE and computational times were recorded. The number of structures autocontoured were 25 for the HN, 7 for the thorax, 3 for the liver, and 6 for the male pelvis regions. Using the clinical contours as reference, autocontours of 22 selected structures were quantitatively evaluated using Dice Similarity Coefficient (DSC) and Mean Slice-wise Hausdorff Distance (MSHD). All 40 autocontours were evaluated by a radiation oncologist from the institution that treated the patients. Results: The mean computational times to autosegment all the structures using SPICE were 3.1 to 11.1 minutes per patient. For the HN region, the mean DSC was >0.70 for all evaluated structures, and the MSHD ranged from 3.2 to 10.0 mm. For the thorax region, the mean DSC was 0.95 for the lungs and 0.90 for the heart, and the MSHD ranged from 2.8 to 12.8 mm. For the liver region, the mean DSC was >0.92 for all structures, and the MSHD ranged from 5.2 to 15.9 mm. For the male pelvis region, the mean DSC was >0.76 for all structures, and the MSHD ranged from 4.8 to 10.5 mm. Out of the 40 autocontoured structures reviews by experts, 25 were scored useful as autocontoured or with minor edits for at least 90% of the patients and 33 were scored useful autocontoured or with minor edits for at least 80% of the patients. Conclusions: Compared with manual contouring, autosegmentation using SPICE for the HN, thorax, liver, and male pelvis regions is efficient and shows significant promise for clinical utility.
AB - Purpose: Clinical validation and quantitative evaluation of computed tomography (CT) image autosegmentation using Smart Probabilistic Image Contouring Engine (SPICE). Methods and Materials: CT images of 125 treated patients (32 head and neck [HN], 40 thorax, 23 liver, and 30 prostate) in 7 independent institutions were autosegmented using SPICE and computational times were recorded. The number of structures autocontoured were 25 for the HN, 7 for the thorax, 3 for the liver, and 6 for the male pelvis regions. Using the clinical contours as reference, autocontours of 22 selected structures were quantitatively evaluated using Dice Similarity Coefficient (DSC) and Mean Slice-wise Hausdorff Distance (MSHD). All 40 autocontours were evaluated by a radiation oncologist from the institution that treated the patients. Results: The mean computational times to autosegment all the structures using SPICE were 3.1 to 11.1 minutes per patient. For the HN region, the mean DSC was >0.70 for all evaluated structures, and the MSHD ranged from 3.2 to 10.0 mm. For the thorax region, the mean DSC was 0.95 for the lungs and 0.90 for the heart, and the MSHD ranged from 2.8 to 12.8 mm. For the liver region, the mean DSC was >0.92 for all structures, and the MSHD ranged from 5.2 to 15.9 mm. For the male pelvis region, the mean DSC was >0.76 for all structures, and the MSHD ranged from 4.8 to 10.5 mm. Out of the 40 autocontoured structures reviews by experts, 25 were scored useful as autocontoured or with minor edits for at least 90% of the patients and 33 were scored useful autocontoured or with minor edits for at least 80% of the patients. Conclusions: Compared with manual contouring, autosegmentation using SPICE for the HN, thorax, liver, and male pelvis regions is efficient and shows significant promise for clinical utility.
UR - http://www.scopus.com/inward/record.url?scp=84886080575&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84886080575&partnerID=8YFLogxK
U2 - 10.1016/j.ijrobp.2013.08.007
DO - 10.1016/j.ijrobp.2013.08.007
M3 - Article
C2 - 24138920
AN - SCOPUS:84886080575
SN - 0360-3016
VL - 87
SP - 809
EP - 816
JO - International Journal of Radiation Oncology Biology Physics
JF - International Journal of Radiation Oncology Biology Physics
IS - 4
ER -