Search

Hands-on ultrasound training is included in the curriculum of many medical specialties and is increasingly incorporated into medical school curricula. Despite published curricula for ultrasound training in these specialties, there remains a dearth of such programs for radiology residency programs. At our institution, there has been a perceived decline in ultrasound scanning comfort and skill in trainees. The purpose of this project was to assess the utility and efficacy of a hands-on simulation-based ultrasound course for radiology residents.First-year radiology residents were enrolled in a 2-week simulation-based course for the instruction of hands-on ultrasound training. With the use of a customizable commercial simulation software platform, residents completed didactic modules, virtual simulations, and phantom scans for a 2-week rotation. A dedicated simulation center with scanning models and computer-based software was provided to all residents. Self-assessments and assignments provided benchmarks of performance. All radiology residents were surveyed at the start of the academic year to assess prior experience and comfort with ultrasound scanning. First-year residents were surveyed a second time upon course completion.Presurvey and postsurvey responses suggest that participation in the 2-week ultrasound scanning course contributed to an improvement in perceived scanning knowledge and comfort for participating residents. Based on our initial experience, the scanning curriculum presented here provides a comprehensive introductory course for first-year radiology residents both for ultrasound anatomy and for scanning technique. Continued education in hands-on ultrasound skills is a crucial factor in maintaining radiology's dominance in the modality.

This study evaluates the performance of convolutional neural networks (CNNs) in risk stratifying the malignant potential of thyroid nodules alongside traditional methods such as American College of Radiology Thyroid Imaging Reporting and Data System (ACR TIRADS). The data set consisted of 651 pathology-proven thyroid nodules (500 benign, 151 malignant) from 571 patients collected at a single tertiary academic medical center. Each thyroid nodule consisted of two orthogonal views (sagittal and transverse) for a total of 1,302 grayscale images. A CNN classifier was developed to identify malignancy versus benign thyroid nodules, and a nested double cross validation scheme was applied to allow for both model parameter selection and for model accuracy evaluation. All thyroid nodules were classified according to ACR TIRADS criteria and were compared with their respective CNN-generated malignancy scores. The best performing model was the MobileNet CNN ensemble with an area under the curve of 0.86 (95% confidence interval, 0.83-0.90). Thyroid nodules within the highest and lowest CNN risk strata had malignancy rates of 81.4% and 5.9%, respectively. The rate of malignancy for ACR TIRADS ranged from 0% for TR1 nodules to 60% for TR5 nodules. Convolutional neural network malignancy scores correlated well with TIRADS levels, as malignancy scores ranged from 0.194 for TR1 nodules and 0.519 for TR5 nodules. Convolutional neural networks can be trained to generate accurate malignancy risk scores for thyroid nodules. These predictive models can aid in risk stratifying thyroid nodules alongside traditional professional guidelines such as TIRADS and can function as an adjunct tool for the radiologist when identifying those patients requiring further histopathologic workup.