Comparing Survival Outcomes of Patients With LI-RADS-M Hepatocellular Carcinomas and Intrahepatic Cholangiocarcinomas.

BACKGROUND: There is a sparsity of data evaluating outcomes of patients with Liver Imaging Reporting and Data System (LI-RADS) (LR)-M lesions. PURPOSE: To compare overall survival (OS) and progression free survival (PFS) between hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (iCCA) meeting LR-M criteria and to evaluate factors associated with prognosis. STUDY TYPE: Retrospective. SUBJECTS: Patients at risk for HCC with at least one LR-M lesion with histologic diagnosis, from 8 academic centers, yielding 120 patients with 120 LR-M lesions (84 men [mean age 62 years] and 36 women [mean age 66 years]). FIELD STRENGTH/SEQUENCE: A 1.5 and 3.0 T/3D T1 -weighted gradient echo, T2 -weighted fast spin-echo. ASSESSMENT: The imaging categorization of each lesion as LR-M was made clinically by a single radiologist at each site and patient outcome measures were collected. STATISTICAL TESTS: OS, PFS, and potential independent predictors were evaluated by Kaplan-Meier method, log-rank test, and Cox proportional hazard model. A P value of <0.05 was considered significant. RESULTS: A total of 120 patients with 120 LR-M lesions were included; on histology 65 were HCC and 55 were iCCA. There was similar median OS for patients with LR-M HCC compared to patients with iCCA (738 days vs. 769 days, P = 0.576). There were no significant differences between patients with HCC and iCCA in terms of sex (47:18 vs. 37:18, P = 0.549), age (63.0 ± 8.4 vs. 63.4 ± 7.8, P = 0.847), etiology of liver disease (P = 0.202), presence of cirrhosis (100% vs. 100%, P = 1.000), tumor size (4.73 ± 3.28 vs. 4.75 ± 2.58, P = 0.980), method of lesion histologic diagnosis (P = 0.646), and proportion of patients who underwent locoregional therapy (60.0% vs. 38.2%, P = 0.100) or surgery (134.8 ± 165.5 vs. 142.5 ± 205.6, P = 0.913). Using multivariable analysis, nonsurgical compared to surgical management (HR, 4.58), larger tumor size (HR, 1.19), and higher MELD score (HR, 1.12) were independently associated with worse OS. DATA CONCLUSION: There was similar OS in patients with LR-M HCC and LR-M iCCA, suggesting that LR-M imaging features may more closely reflect patient outcomes than histology. EVIDENCE LEVEL: 3 TECHNICAL EFFICACY: Stage 5.

Evaluation of the impact of a novel denoising algorithm on image quality in dual-energy abdominal CT of obese patients.

OBJECTIVES: Evaluate a novel algorithm for noise reduction in obese patients using dual-source dual-energy (DE) CT imaging. METHODS: Seventy-nine patients with contrast-enhanced abdominal imaging (54 women; age: 58 ± 14 years; BMI: 39 ± 5 kg/m2, range: 35-62 kg/m2) from seven DECT (SOMATOM Flash or Force) were retrospectively included (01/2019-12/2020). Image domain data were reconstructed with the standard clinical algorithm (ADMIRE/SAFIRE 2), and denoised with a comparison (ME-NLM) and a test algorithm (rank-sparse kernel regression). Contrast-to-noise ratio (CNR) was calculated. Four blinded readers evaluated the same original and denoised images (0 (worst)-100 (best)) in randomized order for perceived image noise, quality, and their comfort making a diagnosis from a table of 80 options. Comparisons between algorithms were performed using paired t-tests and mixed-effects linear modeling. RESULTS: Average CNR was 5.0 ± 1.9 (original), 31.1 ± 10.3 (comparison; p < 0.001), and 8.9 ± 2.9 (test; p < 0.001). Readers were in good to moderate agreement over perceived image noise (ICC: 0.83), image quality (ICC: 0.71), and diagnostic comfort (ICC: 0.6). Diagnostic accuracy was low across algorithms (accuracy: 66, 63, and 67% (original, comparison, test)). The noise received a mean score of 54, 84, and 66 (p < 0.05); image quality 59, 61, and 65; and the diagnostic comfort 63, 68, and 68, respectively. Quality and comfort scores were not statistically significantly different between algorithms. CONCLUSIONS: The test algorithm produces quantitatively higher image quality than current standard and existing denoising algorithms in obese patients imaged with DECT and readers show a preference for it. CLINICAL RELEVANCE STATEMENT: Accurate diagnosis on CT imaging of obese patients is challenging and denoising algorithms can increase the diagnostic comfort and quantitative image quality. This could lead to better clinical reads. KEY POINTS: • Improving image quality in DECT imaging of obese patients is important for accurate and confident clinical reads, which may be aided by novel denoising algorithms using image domain data. • Accurate diagnosis on CT imaging of obese patients is especially challenging and denoising algorithms can increase quantitative and qualitative image quality. • Image domain algorithms can generalize well and can be implemented at other institutions.

Thyroid Nodules on Ultrasound in Children and Young Adults: Comparison of Diagnostic Performance of Radiologists' Impressions, ACR TI-RADS, and a Deep Learning Algorithm.

BACKGROUND. In current clinical practice, thyroid nodules in children are generally evaluated on the basis of radiologists' overall impressions of ultrasound images. OBJECTIVE. The purpose of this article is to compare the diagnostic performance of radiologists' overall impression, the American College of Radiology Thyroid Imaging Reporting and Data System (ACR TI-RADS), and a deep learning algorithm in differentiating benign and malignant thyroid nodules on ultrasound in children and young adults. METHODS. This retrospective study included 139 patients (median age 17.5 years; 119 female patients, 20 male patients) evaluated from January 1, 2004, to September 18, 2020, who were 21 years old and younger with a thyroid nodule on ultrasound with definitive pathologic results from fine-needle aspiration and/or surgical excision to serve as the reference standard. A single nodule per patient was selected, and one transverse and one longitudinal image each of the nodules were extracted for further evaluation. Three radiologists independently characterized nodules on the basis of their overall impression (benign vs malignant) and ACR TI-RADS. A previously developed deep learning algorithm determined for each nodule a likelihood of malignancy, which was used to derive a risk level. Sensitivities and specificities for malignancy were calculated. Agreement was assessed using Cohen kappa coefficients. RESULTS. For radiologists' overall impression, sensitivity ranged from 32.1% to 75.0% (mean, 58.3%; 95% CI, 49.2-67.3%), and specificity ranged from 63.8% to 93.9% (mean, 79.9%; 95% CI, 73.8-85.7%). For ACR TI-RADS, sensitivity ranged from 82.1% to 87.5% (mean, 85.1%; 95% CI, 77.3-92.1%), and specificity ranged from 47.0% to 54.2% (mean, 50.6%; 95% CI, 41.4-59.8%). The deep learning algorithm had a sensitivity of 87.5% (95% CI, 78.3-95.5%) and specificity of 36.1% (95% CI, 25.6-46.8%). Interobserver agreement among pairwise combinations of readers, expressed as kappa, for overall impression was 0.227-0.472 and for ACR TI-RADS was 0.597-0.643. CONCLUSION. Both ACR TI-RADS and the deep learning algorithm had higher sensitivity albeit lower specificity compared with overall impressions. The deep learning algorithm had similar sensitivity but lower specificity than ACR TI-RADS. Interobserver agreement was higher for ACR TI-RADS than for overall impressions. CLINICAL IMPACT. ACR TI-RADS and the deep learning algorithm may serve as potential alternative strategies for guiding decisions to perform fine-needle aspiration of thyroid nodules in children.

Hepatic Artery Infusion Pumps: A Surgical Toolkit for Intraoperative Decision-Making and Management of Hepatic Artery Infusion-Specific Complications.

BACKGROUND: Hepatic artery infusion (HAI) is a liver-directed therapy that delivers high-dose chemotherapy to the liver through the hepatic arterial system for colorectal liver metastases and intrahepatic cholangiocarcinoma. Utilization of HAI is rapidly expanding worldwide. OBJECTIVE AND METHODS: This review describes the conduct of HAI pump implantation, with focus on common technical pitfalls and their associated solutions. Perioperative identification and management of common postoperative complications is also described. RESULTS: HAI therapy is most commonly performed with the surgical implantation of a subcutaneous pump, and placement of its catheter into the hepatic arterial system for inline flow of pump chemotherapy directly to the liver. Intraoperative challenges and abnormal hepatic perfusion can arise due to aberrant anatomy, vascular disease, technical or patient factors. However, solutions to prevent or overcome technical pitfalls are present for the majority of cases. Postoperative HAI-specific complications arise in 22% to 28% of patients in the form of pump pocket (8%-18%), catheter (10%-26%), vascular (5%-10%), or biliary (2%-8%) complications. The majority of patients can be rescued from these complications with early identification and aggressive intervention to continue to deliver safe and effective HAI therapy. CONCLUSIONS: This HAI toolkit provides the HAI team a reference to manage commonly encountered HAI-specific perioperative obstacles and complications. Overcoming these challenges is critical to ensure safe and effective pump implantation and delivery of HAI therapy, and key to successful implementation of new programs and expansion of HAI to patients who may benefit from such a highly specialized treatment strategy.

Artificial Intelligence (AI) Tools for Thyroid Nodules on Ultrasound, From the AJR Special Series on AI Applications.

Artificial intelligence (AI) methods for evaluating thyroid nodules on ultrasound have been widely described in the literature, with reported performance of AI tools matching or in some instances surpassing radiologists' performance. As these data have accumulated, products for classification and risk stratification of thyroid nodules on ultrasound have become commercially available. This article reviews FDA-approved products currently on the market, with a focus on product features, reported performance, and considerations for implementation. The products perform risk stratification primarily using a Thyroid Imaging Reporting and Data System (TIRADS), though may provide additional prediction tools independent of TIRADS. Key issues in implementation include integration with radiologist interpretation, impact on workflow and efficiency, and performance monitoring. AI applications beyond nodule classification, including report construction and incidental findings follow-up, are also described. Anticipated future directions of research and development in AI tools for thyroid nodules are highlighted.

Hepatic Arterial Infusion Pumps: What the Radiologist Needs to Know.

Hepatic arterial infusion (HAI) entails the surgical implantation of a subcutaneous pump to deliver chemotherapeutic agents directly to the liver in the setting of primary or secondary liver cancer. The purpose of HAI chemotherapy is to maximize hepatic drug concentrations while minimizing systemic toxicity, facilitating more effective treatment. HAI is used in combination with systemic chemotherapy and can be considered in several clinical scenarios, including adjuvant therapy, conversion of unresectable disease to resectable disease, and unresectable disease. Radiologists are key members of the multidisciplinary team involved in the selection and management of these patients with complex liver disease. As these devices begin to be used at more sites across the country, radiologists should become familiar with the guiding principles behind pump placement, expected imaging appearances of these devices, and potential associated complications. The authors provide an overview of HAI therapy, with a focus on the key imaging findings associated with this treatment that radiologists may encounter. ©RSNA, 2021.

Using Radiographic Domain for Evaluating Indications in Abdominal Wall Transplantation.

BACKGROUND: There is currently no description of abdominal domain changes in small bowel transplantation population or consensus of criteria regarding which patients are at high risk for immediate postoperative abdominal wall complications or would benefit from abdominal wall vascularized composite allotransplantation. METHODS: A retrospective chart review was performed on 14 adult patients receiving intestinal or multivisceral transplantation. Preoperative and postoperative computed tomography scans were reviewed, and multiple variables were collected regarding abdominal domain and volume and analyzed comparing postoperative changes and abdominal wall complications. RESULTS: Patients after intestinal or multivisceral transplantation had a mean reduction in overall intraperitoneal volume in the immediate postoperative period from 9031 cm3 to 7846 cm3 (P = 0.314). This intraperitoneal volume was further reduced to an average of 6261 cm3 upon radiographic evaluation greater than 1 year postoperatively (P = 0.024). Patients with preexisting abdominal wound (P = 0.002), radiation, or presence of ostomy (P = 0.047) were significantly associated with postoperative abdominal wall complications. No preoperative radiographic findings had a significant association with postoperative abdominal wall complications. CONCLUSIONS: Computed tomography imaging demonstrates that intestinal and multivisceral transplant patients have significant reduction in intraperitoneal volume and domain after transplantation in the acute and delayed postoperative setting. Preoperative radiographic abdominal domain was not able to predict patients with postoperative abdominal wall complications. Patients with abdominal wounds, ostomies, and preoperative radiation therapy were associated with acute postoperative abdominal complications and may be considered for need of reconstructive techniques including abdominal wall transplantation.

Implementation of a Hepatic Artery Infusion Program: Initial Patient Selection and Perioperative Outcomes of Concurrent Hepatic Artery Infusion and Systemic Chemotherapy for Colorectal Liver Metastases.

BACKGROUND: Hepatic artery infusion (HAI) combined with systemic chemotherapy is a treatment strategy for patients with unresectable liver-only or liver-dominant colorectal liver metastases (CRLM). Although HAI has previously been performed in only a few centers, this study aimed to describe patient selection and initial perioperative outcomes during implementation of a new HAI program. METHODS: The study enrolled patients with CRLM selected for HAI after multi-disciplinary review November 2018-January 2020. Demographics, prior treatment, and perioperative outcomes were assessed. Objective hepatic response was calculated according to Response Evaluation Criteria in Solid Tumors (RECIST) 1.1. RESULTS: During a 14-month period, 21 patients with CRLM underwent HAI pump placement. Of these 21 patients, 20 (95%) had unresectable disease. Most of the patients had synchronous disease (n = 18, 86%) and had received prior chemotherapy (n = 20, 95%) with extended treatment cycles (median 16; interquartile range, 8-22; range, 0-66). The median number of CRLMs was 7 (range, 2-40). Operations often were performed with combined hepatectomy (n = 4, 19%) and/or colectomy/proctectomy (n = 11, 52%). The study had no 90-day mortality. The overall surgical morbidity was 19%. The HAI-specific complications included pump pocket seroma (n = 2), hematoma (n = 1), surgical-site infection (n = 1), and extrahepatic perfusion (n = 1). HAI was initiated in 20 patients (95%). The hepatic response rates at 3 months included partial response (n = 4, 24%), stable disease (n = 9, 53%), and progression of disease (n = 4, 24%), yielding a 3-month hepatic disease control rate (DCR) of 76%. CONCLUSION: Implementation of a new HAI program is feasible, and HAI can be delivered safely to selected patients with CRLM. The initial response and DCR are promising, even for patients heavily pretreated with chemotherapy.

Reorganizing Cross-Sectional Interventional Procedures Practice During the Coronavirus Disease (COVID-19) Pandemic.

OBJECTIVE. The purpose of this article is to present strategies and guidelines that can be implemented in the performance of cross-sectional interventional procedures during the coronavirus disease (COVID-19) pandemic. CONCLUSION. Radiologists who perform cross-sectional interventional procedures can take several steps to minimize the risks to patients and radiology personnel, including screening referred patients to decide which procedures can be postponed, using appropriate personal protective equipment (PPE), minimizing the number of people involved in procedures, preserving PPE when possible, and applying proper room and equipment cleaning measures.

Association Between Magnetic Resonance Imaging-Proton Density Fat Fraction and Liver Histology Features in Patients With Nonalcoholic Fatty Liver Disease or Nonalcoholic Steatohepatitis.

BACKGROUND & AIMS: Patients with nonalcoholic fatty liver disease (NAFLD) or nonalcoholic steatohepatitis (NASH) often require histologic assessment via liver biopsy. Magnetic resonance imaging (MRI)-based methods for measuring liver triglycerides based on proton density fat fraction (PDFF) are increasingly used as a noninvasive tool to identify patients with hepatic steatosis and to assess for change in liver fat over time. We aimed to determine whether MRI-PDFF accurately reflects a variety of liver histology features in patients with NAFLD or NASH. METHODS: We performed a retrospective analysis of pooled data from 3 phase 2a trials of pharmacotherapies for NAFLD or NASH. We collected baseline clinical, laboratory, and histopathology data on all subjects who had undergone MRI analysis in 1 of the trials. We assessed the relationship between liver PDFF values and liver histologic findings using correlation and area under the receiver operating characteristic (AUROC) analyses. As an ancillary analysis, we also simulated a clinical trial selection process and calculated subject exclusion rates and differences in population characteristics caused by PDFF inclusion thresholds of 6% to 15%. RESULTS: In 370 subjects, the mean baseline PDFF was 17.4% ± 8.6%. Baseline PDFF values correlated with several histopathology parameters, including steatosis grade (r = 0.78; P < .001), NAFLD activity score (NAS, r = 0.54; P < .001), and fibrosis stage (r = -0.59; P < .001). However, PDFF did not accurately identify patients with NAS ≥ 4 (AUROC = 0.72) or fibrosis stage ≥3 (AUROC = 0.66). In a theoretical trial of these subjects, exclusion rates increased as PDFF minimum threshold level increased. There were no significant differences in cohort demographics when PDFF thresholds ranging from 6% to 15% were used, and differences in laboratory and histopathology data were small. CONCLUSIONS: In an analysis of patients with NAFLD or NASH, we determined that although The MRI-PDFF correlated with steatosis grade and NAS, and inversely with fibrosis stage, it was suboptimal in identification of patients with NAS >4 or advanced fibrosis. Although MRI-PDFF is an important imaging biomarker for continued evaluation of this patient population, liver biopsy analysis is still necessary.

Management of Thyroid Nodules Seen on US Images: Deep Learning May Match Performance of Radiologists.

BackgroundManagement of thyroid nodules may be inconsistent between different observers and time consuming for radiologists. An artificial intelligence system that uses deep learning may improve radiology workflow for management of thyroid nodules.PurposeTo develop a deep learning algorithm that uses thyroid US images to decide whether a thyroid nodule should undergo a biopsy and to compare the performance of the algorithm with the performance of radiologists who adhere to American College of Radiology (ACR) Thyroid Imaging Reporting and Data System (TI-RADS).Materials and MethodsIn this retrospective analysis, studies in patients referred for US with subsequent fine-needle aspiration or with surgical histologic analysis used as the standard were evaluated. The study period was from August 2006 to May 2010. A multitask deep convolutional neural network was trained to provide biopsy recommendations for thyroid nodules on the basis of two orthogonal US images as the input. In the training phase, the deep learning algorithm was first evaluated by using 10-fold cross-validation. Internal validation was then performed on an independent set of 99 consecutive nodules. The sensitivity and specificity of the algorithm were compared with a consensus of three ACR TI-RADS committee experts and nine other radiologists, all of whom interpreted thyroid US images in clinical practice.ResultsIncluded were 1377 thyroid nodules in 1230 patients with complete imaging data and conclusive cytologic or histologic diagnoses. For the 99 test nodules, the proposed deep learning algorithm achieved 13 of 15 (87%: 95% confidence interval [CI]: 67%, 100%) sensitivity, the same as expert consensus (P > .99) and higher than five of nine radiologists. The specificity of the deep learning algorithm was 44 of 84 (52%; 95% CI: 42%, 62%), which was similar to expert consensus (43 of 84; 51%; 95% CI: 41%, 62%; P = .91) and higher than seven of nine other radiologists. The mean sensitivity and specificity for the nine radiologists was 83% (95% CI: 64%, 98%) and 48% (95% CI: 37%, 59%), respectively.ConclusionSensitivity and specificity of a deep learning algorithm for thyroid nodule biopsy recommendations was similar to that of expert radiologists who used American College of Radiology Thyroid Imaging and Reporting Data System guidelines.© RSNA, 2019Online supplemental material is available for this article.

Using Artificial Intelligence to Revise ACR TI-RADS Risk Stratification of Thyroid Nodules: Diagnostic Accuracy and Utility.

Background Risk stratification systems for thyroid nodules are often complicated and affected by low specificity. Continual improvement of these systems is necessary to reduce the number of unnecessary thyroid biopsies. Purpose To use artificial intelligence (AI) to optimize the American College of Radiology (ACR) Thyroid Imaging Reporting and Data System (TI-RADS). Materials and Methods A total of 1425 biopsy-proven thyroid nodules from 1264 consecutive patients (1026 women; mean age, 52.9 years [range, 18-93 years]) were evaluated retrospectively. Expert readers assigned points based on five ACR TI-RADS categories (composition, echogenicity, shape, margin, echogenic foci), and a genetic AI algorithm was applied to a training set (1325 nodules). Point and pathologic data were used to create an optimized scoring system (hereafter, AI TI-RADS). Performance of the systems was compared by using a test set of the final 100 nodules with interpretations from the expert reader, eight nonexpert readers, and an expert panel. Initial performance of AI TI-RADS was calculated by using a test for differences between binomial proportions. Additional comparisons across readers were conducted by using bootstrapping; diagnostic performance was assessed by using area under the receiver operating curve. Results AI TI-RADS assigned new point values for eight ACR TI-RADS features. Six features were assigned zero points, which simplified categorization. By using expert reader data, the diagnostic performance of ACR TI-RADS and AI TI-RADS was area under the receiver operating curve of 0.91 and 0.93, respectively. For the same expert, specificity of AI TI-RADS (65%, 55 of 85) was higher (P < .001) than that of ACR TI-RADS (47%, 40 of 85). For the eight nonexpert radiologists, mean specificity for AI TI-RADS (55%) was also higher (P < .001) than that of ACR TI-RADS (48%). An interactive AI TI-RADS calculator can be viewed at http://deckard.duhs.duke.edu/∼ai-ti-rads . Conclusion An artificial intelligence-optimized Thyroid Imaging Reporting and Data System (TI-RADS) validates the American College of Radiology TI-RADS while slightly improving specificity and maintaining sensitivity. Additionally, it simplifies feature assignments, which may improve ease of use. © RSNA, 2019 Online supplemental material is available for this article.

DIfferential Subsampling With Cartesian Ordering With Respiratory Triggering Versus Conventional Liver Acquisition With Volume Acquisition: A Multiple Reader Preference Study.

OBJECTIVE: The aim of this study was to compare respiratory-triggered DIfferential Subsampling with Cartesian Ordering (rtDISCO) and breath-held Liver Acquisition with Volume Acquisition (LAVA) image quality. METHODS: In this institutional review board-approved, Health Insurance Portability and Accountability Act-compliant prospective study, 25 subjects underwent T1 imaging with rtDISCO and LAVA before and after intravenous contrast. Three readers scored individual series and side-by-side comparisons for motion and noise. Eight clinical tasks were qualitatively assessed. RESULTS: As individual series, readers rated rtDISCO images as more degraded by motion on both precontrast (mean rtDISCO score, 2.7; LAVA, 1.6; P < 0.001) and postcontrast images (rtDISCO, 2.4; LAVA, 1.8; P < 0.001). Readers preferred LAVA images based on motion on both precontrast (mean preference, -1.2; P < 0.001) and postcontrast images (mean preference, -0.7; P < 0.001) on side-by-side assessment. There was no preference between sequences for 6 of 8 clinical tasks on postcontrast images. CONCLUSIONS: Readers preferred LAVA with respect to motion but not noise; there was no preference in most of the tested clinical tasks.

Using the American College of Radiology Dose Index Registry to Evaluate Practice Patterns and Radiation Dose Estimates of Pediatric Body CT.

OBJECTIVE: Imaging registries afford opportunities to study large, heterogeneous populations. The purpose of this study was to examine the American College of Radiology CT Dose Index Registry (DIR) for dose-related demographics and metrics of common pediatric body CT examinations. MATERIALS AND METHODS: Single-phase CT examinations of the abdomen and pelvis and chest submitted to the DIR over a 5-year period (July 2011-June 2016) were evaluated (head CT frequency was also collected). CT examinations were stratified into five age groups, and examination frequency was determined across age and sex. Standard dose indexes (volume CT dose index, dose-length product, and size-specific dose estimate) were categorized by body part and age. Contributions to the DIR were also categorized by region and practice type. RESULTS: Over the study period 411,655 single-phase pediatric examinations of the abdomen and pelvis, chest, and head, constituting 5.7% of the total (adult and pediatric) examinations, were submitted to the DIR. Head CT was the most common examination across all age groups. The majority of all scan types were performed for patients in the second decade of life. Dose increased for all scan types as age increased; the dose for abdominopelvic CT was the highest in each age group. Even though the DIR was queried for single-phase examinations only, as many as 32.4% of studies contained multiple irradiation events. When these additional scans were included, the volume CT dose index for each scan type increased. Among the studies in the DIR, 99.8% came from institutions within the United States. Community practices and those that specialize in pediatrics were nearly equally represented. CONCLUSION: The DIR provides valuable information about practice patterns and dose trends for pediatric CT and may assist in establishing diagnostic reference levels in the pediatric population.

Pediatric providers and radiology examinations: knowledge and comfort levels regarding ionizing radiation and potential complications of imaging.

BACKGROUND: Pediatric providers should understand the basic risks of the diagnostic imaging tests they order and comfortably discuss those risks with parents. Appreciating providers' level of understanding is important to guide discussions and enhance relationships between radiologists and pediatric referrers. OBJECTIVE: To assess pediatric provider knowledge of diagnostic imaging modalities that use ionizing radiation and to understand provider concerns about risks of imaging. MATERIALS AND METHODS: A 6-question survey was sent via email to 390 pediatric providers (faculty, trainees and midlevel providers) from a single academic institution. A knowledge-based question asked providers to identify which radiology modalities use ionizing radiation. Subjective questions asked providers about discussions with parents, consultations with radiologists, and complications of imaging studies. RESULTS: One hundred sixty-nine pediatric providers (43.3% response rate) completed the survey. Greater than 90% of responding providers correctly identified computed tomography (CT), fluoroscopy and radiography as modalities that use ionizing radiation, and ultrasound and magnetic resonance imaging (MRI) as modalities that do not. Fewer (66.9% correct, P<0.001) knew that nuclear medicine utilizes ionizing radiation. A majority of providers (82.2%) believed that discussions with radiologists regarding ionizing radiation were helpful, but 39.6% said they rarely had time to do so. Providers were more concerned with complications of sedation and cost than they were with radiation-induced cancer, renal failure or anaphylaxis. CONCLUSION: Providers at our academic referral center have a high level of basic knowledge regarding modalities that use ionizing radiation, but they are less aware of ionizing radiation use in nuclear medicine studies. They find discussions with radiologists helpful and are concerned about complications of sedation and cost.

Increasing utilization of contrast-enhanced ultrasound during abdominal biopsies: impact of an educational training program.

RATIONAL AND OBJECTIVES: To increase utilization of contrast-enhanced ultrasound (CEUS) during ultrasound-guided targeted liver biopsies. MATERIAL AND METHODS: Two educational training interventions performed to increase use of CEUS. First, 14 radiologists (fellowship-trained in Abdominal Imaging) given didactic teaching and case presentations on the use of CEUS. Second, hands-on teaching on how to use CEUS provided to the same group. To determine the efficacy of these two interventions, radiologists completed anonymous surveys to determine the level of understanding and acceptability of using CEUS before and 6 months after CEUS training. In addition, the percentage of CEUS assisted liver biopsies was compared for the 6 months before and 6 months after the training. RESULTS: Pre-training survey completed by 11 radiologists and post-training survey completed by 9 radiologists. Before training, 11% survey responders use CEUS routinely, whereas 89% never or rarely used it. After training, 54% of respondents were new users and 100% reported they planned to use CEUS in the future. Unfamiliarity (71%) was the main reason for not using it. After training, 25% reported lack of comfort with using CEUS as the main reason for not using CEUS. During six months before training, CEUS was administered in 6% (10/172) of targeted liver biopsies. Six months after training, CEUS was used nearly twice as often (10%, 16/160, P = 0.09, 1-sided Boschloo test). The number of radiologists using CEUS increased to 57% (8/14) after training compared to 20% (3/14, P = 0.03, 1-sided Boschloo) before training. CONCLUSION: Educational training intervention increases use of CEUS during ultrasound-guided targeted liver biopsies.

Multistep Automated Data Labelling Procedure (MADLaP) for thyroid nodules on ultrasound: An artificial intelligence approach for automating image annotation.

Machine learning (ML) for diagnosis of thyroid nodules on ultrasound is an active area of research. However, ML tools require large, well-labeled datasets, the curation of which is time-consuming and labor-intensive. The purpose of our study was to develop and test a deep-learning-based tool to facilitate and automate the data annotation process for thyroid nodules; we named our tool Multistep Automated Data Labelling Procedure (MADLaP). MADLaP was designed to take multiple inputs including pathology reports, ultrasound images, and radiology reports. Using multiple step-wise 'modules' including rule-based natural language processing, deep-learning-based imaging segmentation, and optical character recognition, MADLaP automatically identified images of a specific thyroid nodule and correctly assigned a pathology label. The model was developed using a training set of 378 patients across our health system and tested on a separate set of 93 patients. Ground truths for both sets were selected by an experienced radiologist. Performance metrics including yield (how many labeled images the model produced) and accuracy (percentage correct) were measured using the test set. MADLaP achieved a yield of 63 % and an accuracy of 83 %. The yield progressively increased as the input data moved through each module, while accuracy peaked part way through. Error analysis showed that inputs from certain examination sites had lower accuracy (40 %) than the other sites (90 %, 100 %). MADLaP successfully created curated datasets of labeled ultrasound images of thyroid nodules. While accurate, the relatively suboptimal yield of MADLaP exposed some challenges when trying to automatically label radiology images from heterogeneous sources. The complex task of image curation and annotation could be automated, allowing for enrichment of larger datasets for use in machine learning development.