Publications

BACKGROUND: No large dataset-derived standard has been established for normal or pathologic human cerebral ventricular and cranial vault volumes. Automated volumetric measurements could be used to assist in diagnosis and follow-up of hydrocephalus or craniofacial syndromes. In this work, we use deep learning algorithms to measure ventricular and cranial vault volumes in a large dataset of head computed tomography (CT) scans.

METHODS: A cross-sectional dataset comprising 13,851 CT scans was used to deploy U-Net deep learning networks to segment and quantify lateral cerebral ventricular and cranial vault volumes in relation to age and sex. The models were validated against manual segmentations. Corresponding radiologic reports were annotated using a rule-based natural language processing framework to identify normal scans, cerebral atrophy, or hydrocephalus.

RESULTS: U-Net models had high fidelity to manual segmentations for lateral ventricular and cranial vault volume measurements (Dice index, 0.878 and 0.983, respectively). The natural language processing identified 6239 (44.7%) normal radiologic reports, 1827 (13.1%) with cerebral atrophy, and 1185 (8.5%) with hydrocephalus. Age-based and sex-based reference tables with medians, 25th and 75th percentiles for scans classified as normal, atrophy, and hydrocephalus were constructed. The median lateral ventricular volume in normal scans was significantly smaller compared with hydrocephalus (15.7 vs. 82.0 mL; P < 0.001).

CONCLUSIONS: This is the first study to measure lateral ventricular and cranial vault volumes in a large dataset, made possible with artificial intelligence. We provide a robust method to establish normal values for these volumes and a tool to report these on CT scans when evaluating for hydrocephalus.

PURPOSE: To assess if a templated algorithm can improve the diagnostic performance of MRI for characterization of T2 isointense and hypointense renal masses.

METHODS: In this retrospective study, 60 renal masses with histopathologic diagnoses that were also confirmed as T2 iso- or hypointense on MRI were identified (mean ± standard deviation, range: 3.9 ± 2.5, 1.0-13.7 cm). Two semi-quantitative diagnostic algorithms were created based on MRI features of renal masses reported in the literature. Three body-MRI trained radiologists provided clinical diagnoses based on their experience and separately provided semiquantitative data for each components of the two algorithms. The algorithms were applied separately by a radiology trainee without additional interpretive input. Logistic regression was used to compare the accuracy of the three methods in distinguishing malignant versus benign lesions and in diagnosing the exact histopathology. Inter-reader agreement for each method was calculated using Fleiss' kappa statistics.

RESULTS: The accuracy of the two algorithms and clinical experience were similar (70%, 69%, and 64%, respectively, p = 0.22-0.32), with fair to moderate inter-reader agreement (Fleiss's kappa: r = 0.375, r = 0.308, r = 0.375, respectively, all p < 0.0001). The accuracy of the two algorithms and clinical experience in diagnosing specific histopathology were also no different from each other (34%, 29%, and 32%, respectively, p = 0.49-0.74), with fair to moderate inter-reader agreement (Fleiss's kappa: r = 0.20, r = 0.28, r = 0.375, respectively, all p < 0.0001).

CONCLUSION: Semi-quantitative templated algorithms based on MRI features of renal masses did not improve the ability to diagnose T2 iso- and hypointense renal masses when compared to unassisted interpretation by body MR trained subspecialists.

The acute consequences of the COVID-19 pandemic have impacted wellness strategies aimed at mitigating the pre-existing epidemic of burnout in radiology. Specifically, safety measures including social distancing requirements, effective communications, supporting remote and distributed work teams, and newly exposed employment and treatment inequities have challenged many major efforts at fostering professional fulfillment. To get our wellness efforts back on track and to achieve a new and perhaps even a better "normal" will require refocusing and reconsidering ways to foster and build a culture of wellness, implementing practices that improve work efficiencies, and supporting personal health, wellness behaviors, and resilience. Optimizing meaning in work is also critical for well-being and professional fulfillment. In addition to these earlier approaches, organizations and leaders will need to reprioritize efforts to build high-functioning cohesive and connected teams; to train, implement, and manage peer-support practices; and to support posttraumatic growth. This growth represents the positive psychological changes that can occur after highly challenging life circumstances and, when successful, allows individuals to achieve a higher level of functioning by addressing and learning from the precipitating event. Our practices can support this growth through education, emotional regulation, and disclosure, by developing a narrative that reimagines a hoped-for better future and by finding meaning through services that benefit others.

Interventional radiology continues to evolve into a more robust and clinically dynamic specialty underpinned by significant advancements in training, education, and practice. This article, prepared by members of the 2020-2021 Association of University Radiologists' task force of the Radiology Research Alliance, will review these developments, highlighting the evolution of interventional radiology pathways with attention to growing educational differences, interrelation to diagnostic radiology training, post-training practice patterns, distribution of procedures and future trends, amongst other key features important to those pursuing a career in interventional radiology as well as those in practice.

OBJECTIVE: To determine safety of shortened observation time without follow-up chest x-ray (CXR) after CT-guided transthoracic procedures (lung biopsy or fiducial placement) in patients without immediate postprocedural pneumothorax (PTX).

METHODS: Consecutive patients that underwent CT-guided procedures between January 5, 2015, and June 19, 2017, were included in this retrospective institutional review board-approved HIPAA-compliant study. Data regarding postprocedural course, complications, and clinical follow-up of the patients were obtained through a review of electronic medical records. Descriptive statistics were used.

RESULTS: There were 441 procedures for 409 patients performed; 82 procedures were excluded because of predefined criteria. In 312 of 336 asymptomatic procedures (92.9%), asymptomatic patients did not undergo CXR after procedure, with 7 of 312 of these patients (2.2%) diagnosed with delayed PTX 2 to 10 days after the procedure. In 24 of 336 procedures (7.1%), asymptomatic patients underwent CXR within 4 hours with no PTX detected, and despite that 1 of 24 of these patients (4.2%) presented with delayed PTX 7 days after procedure. When no immediate postprocedural PTX was present, rate of observation PTX and delayed PTX was 1 of 359 (0.3%) and 8 of 359 (2.2%), respectively. Average duration of monitoring for outpatients (n = 295) was 2.0 hours with median of 1.8 hours. In 23 of 359 (6.4%) procedures, the patient became symptomatic during postprocedural observation with 1 of 23 (4%) developing PTX.

CONCLUSIONS: Obtaining routine postprocedural CXRs in asymptomatic patients without immediate postprocedural PTX after CT-guided transthoracic procedures is likely not necessary given the low likelihood of PTX.

Lack of diversity in Radiology is a public health problem and may be self perpetuating as diverse candidates view the field as hostile to their entry and advancement, and consequently do not apply into the field. Solutions require understanding the obstacles, which range from enrollment in medical school to achieving leadership positions in Radiology. An understanding of the effect of demographic data on diversity in Radiology, disparate effects of Step examinations, medical school grades and induction into academic honor societies, and existing faculty disparities will allow us to better recruit, train, and retain a diverse group of physicians in our field. The downstream effect of a diverse workforce is improvement in health outcomes and disparities in medical care for our communities.

PURPOSE: Hepatic thermal ablation therapy can result in c-Met-mediated off-target stimulation of distal tumor growth. The purpose of this study was to determine if a similar effect on tumor metabolism could be detected in vivo with hyperpolarized 13C MRI.

MATERIALS AND METHODS: In this prospective study, female Fisher rats (n = 28, 120-150 g) were implanted with R3230 rat breast adenocarcinoma cells and assigned to either: sham surgery, hepatic radiofrequency ablation (RFA), or hepatic RFA + adjuvant c-Met inhibition with PHA-665752 (RFA + PHA). PHA-665752 was administered at 0.83 mg/kg at 24 h post-RFA. Tumor growth was measured daily. MRI was performed 24 h before and 72 h after treatment on 14 rats, and the conversion of 13C-pyruvate into 13C-lactate within each tumor was quantified as lactate:pyruvate ratio (LPR). Comparisons of tumor growth and LPR were performed using paired and unpaired t-tests.

RESULTS: Hepatic RFA alone resulted in increased growth of the distant tumor compared to sham treatment (0.50 ± 0.13 mm/day versus 0.11 ± 0.07 mm/day; p < 0.001), whereas RFA + PHA (0.06 ± 0.13 mm/day) resulted in no significant change from sham treatment (p = 0.28). A significant increase in LPR was seen following hepatic RFA (+0.016 ± 0.010, p = 0.02), while LPR was unchanged for sham treatment (-0.048 ± 0.051, p = 0.10) or RFA + PHA (0.003 ± 0.041, p = 0.90).

CONCLUSION: In vivo hyperpolarized 13C MRI can detect hepatic RFA-induced increase in lactate flux within a distant R3230 tumor, which correlates with increased tumor growth. Adjuvant inhibition of c-Met suppresses these off-target effects, supporting a role for the HGF/c-Met signaling axis in these tumorigenic responses.

OBJECTIVE: To review the technical feasibility of resin microsphere (SIR-Spheres®) yttrium-90 radioembolization prescribed using the medical internal radiation dose (MIRD) model.

METHODS: All radioembolization procedures for hepatic malignancies using resin microspheres with MIRD model between November 2015 and February 2019 were included in this IRB-approved study (n = 60). Student's T test was used to compare prescribed activity based on MIRD and BSA models. Adverse events were assessed immediately, 30 days, and 6 months post-treatment.

RESULTS: Sixty radioembolizations were performed in 54 patients (age 68 ± 9 years, 48-87 years, 35% female). Mean prescribed activity calculated by the MIRD model (target absorbed dose 120-200 Gy for primary and 80-200 Gy for metastatic liver cancers) was 1.7 GBq (0.3-6.4) compared with 0.6 GBq (0.12-2.1) if BSA had been used (p < 0.0001). The prescribed activity was successfully delivered in 93% (56/60) treatments. Prophylactic embolization and anti-reflux catheters were used in 20% (12/60) and 5% (3/60) treatments, respectively. No immediate post-procedural complications occurred. Abdominal pain was the most common clinical Grade 3 CTCAE in 30 days (10%) and 6 months (12%). Radiation pneumonitis occurred in 3 (5%) patients but no radiation-induced gastric ulcer or radiation-induced liver disease occurred.

CONCLUSION: MIRD dosimetry results in higher prescribed activity compared with BSA dosimetry with resin microspheres. MIRD prescribed activity with target absorbed doses up to 200 Gy can be successfully administered without prophylactic embolization in selected patients.

KEY POINTS: •MIRD dosimetry results in higher prescribed activity compared with BSA dosimetry for radioembolization. •MIRD dosimetry can be used for yttrium-90 resin microsphere radioembolization with acceptable safety profile.