Search

Background and Purpose: Auto-contouring of complex anatomy in computed tomography (CT) scans is a highly anticipated solution to many problems in radiotherapy. In this study, artificial intelligence (AI)-based auto-contouring models were clinically validated for lymph node levels and structures of swallowing and chewing in the head and neck. Materials and Methods: CT scans of 145 head and neck radiotherapy patients were retrospectively curated. One cohort (n = 47) was used to analyze seven lymph node levels and the other (n = 98) used to analyze 17 swallowing and chewing structures. Separate nnUnet models were trained and validated using the separate cohorts. For the lymph node levels, preference and clinical acceptability of AI vs human contours were scored. For the swallowing and chewing structures, clinical acceptability was scored. Quantitative analyses of the test sets were performed for AI vs human contours for all structures using overlap and distance metrics. Results: Median Dice Similarity Coefficient ranged from 0.77 to 0.89 for lymph node levels and 0.86 to 0.96 for chewing and swallowing structures. The AI contours were superior to or equally preferred to the manual contours at rates ranging from 75% to 91%; there was not a significant difference in clinical acceptability for nodal levels I-V for manual versus AI contours. Across all AI-generated lymph node level contours, 92% were rated as usable with stylistic to no edits. Of the 340 contours in the chewing and swallowing cohort, 4% required minor edits. Conclusions: An accurate approach was developed to auto-contour lymph node levels and chewing and swallowing structures on CT images for patients with intact nodal anatomy. Only a small portion of test set auto-contours required minor edits.

Introduction: The first high-quality clinical trial to support ultrahypofractionated whole-breast irradiation (ultra-HF-WBI) for invasive early-stage breast cancer (ESBC) was published in April 2020, coinciding with the beginning of the COVID-19 pandemic. We analyzed adoption of ultra-HF-WBI for ductal carcinoma in situ (DCIS) and ESBC at our institution after primary trial publication. Methods and Materials: We evaluated radiation fractionation prescriptions for all patients with DCIS or ESBC treated with WBI from March 2020 to May 2021 at our main campus and regional campuses. Demographic and clinical characteristics were extracted from the electronic medical record. Treating physician characteristics were collected from licensure data. Hierarchical logistic regression models identified factors correlated with adoption of ultra-HF-WBI (26 Gy in 5 daily factions [UK-FAST-FORWARD] or 28.5 Gy in 5 weekly fractions [UK-FAST]). Results: Of 665 included patients, the median age was 61.5 years, and 478 patients (71.9%) had invasive, hormone-receptor-positive breast cancer. Twenty-one physicians treated the included patients. In total, 249 patients (37.4%) received ultra-HF-WBI, increasing from 4.3% (2 of 46) in March-April 2020 to a high of 45.5% (45 of 99) in July-August 2020 (P < .001). Patient factors associated with increased use of ultra-HF-WBI included older age (≥50 years old), low-grade WBI without inclusion of the low axilla, no radiation boost, and farther travel distance (P < .03). Physician variation accounted for 21.7% of variance in the outcome, with rate of use of ultra-HF-WBI by the treating physicians ranging from 0% to 75.6%. No measured physician characteristics were associated with use of ultra-HF-WBI. Conclusions: Adoption of ultra-HF-WBI at our institution increased substantially after the publication of randomized evidence supporting its use. Ultra-HF-WBI was preferentially used in patients with lower risk disease, suggesting careful selection for this new approach while long-term data are maturing. Substantial physician-level variation may reflect a lack of consensus on the evidentiary standards required to change practice.