89. Impiego del software "MIM Maestro" per la segmentazione automatica del distretto pelvi in radioterapia. (December 2018)

• Record Type: Journal Article
• Authors: Casati, M.; Calusi, S.; Marrazzo, L.; Piffer, S.; Wandael, Y.; Grela, K.; Di Dio, A.; Simontacchi, G.; Pallotta, S.
• Is Part Of: Physica medica. Volume 56(2018)Supplement 2
• Page Start: 116
• Page End: 117
• Publication Date: 2018-12
• Subjects: Medical physics -- Periodicals; Biophysics -- Periodicals; Biophysics -- Periodicals; Imagerie médicale -- Périodiques; Radiothérapie -- Périodiques; Rayons X -- Sécurité -- Mesures -- Périodiques; Physique -- Périodiques; Médecine -- Périodiques; 610.153
• DOI: 10.1016/j.ejmp.2018.04.099
• Languages: English
• ISSNs: 1120-1797

Abstract

Abstract : Purpose: The feasibility of adaptive radiotherapy is based on the possibility of rapid contouring and re-planning. This work aims to evaluate if an automatic workflow for pelvis structures segmentation developed with MIM Maestro (MIM Software, Inc., Cleveland, OH) could help to reduce contouring time and inter-operator variability. Methods: Four physicians contoured manually the pelvis structures (lymph-nodal CTV, rectum, bladder and femoral heads) of 3 patients (M1) registering the delineation time. The same procedure was repeated after one month (M2). Contours were automatically generated (A) with an atlas-based auto-segmentation workflow developed in MIM Maestro and then corrected by each physician (AC). The editing time was registered. The workflow combined a proprietary 38 subject atlas "High risk prostate" for rectum with two home-made atlases for other structures: one (73 subjects) for bladder and femoral heads, the other (14 subjects) for lymph-nodes. Dice similarity coefficient (DSC) was used to compare contours. Intra-operator variability was assessed by comparison of M1 and M2 contours (test A). Inter-operator variability (test B) was evaluated by comparing the 4 sets of original manual (M1) contours with a "gold standard" (GS). This was generated by applying STAPLE algorithm to the manual (M1) contours. To evaluate any possible reduction of inter-operator variability due to the introduction of auto-contouring, also AC contours were compared (test C). Finally AC editing time was compared to M1 contouring time (test D). Results: Results of test A, B and C are summarized in the box-whisker plot. As presumable, inter-operator variability exceeds intra-operator variability and the introduction of auto-contouring reduces inter-operator variability. The results of test D are summarized in the table. An average (4 physicians and 3 patients) time saving of 10% for CTV and 17% for other structures was obtained. Conclusions: This work demonstrates the potentiality of auto-contouring to reduce time and inter-operator variability.