Recruiting Medical Students to Interventional Radiology: Current State of Affairs Publication date: September 2019 Source: Academic Radiology, Volume 26, Issue 9 Author(s): Juri Bassuner, David Duncan, Chris Molloy, Mina S. Makary, Brycen Bodell, Dylan Assael, Riz Ahmed, Geogy Vatakencherry Objective Interventional radiology/diagnostic radiology (IR/DR) is the newest primary specialties offered to trainees, one that medical students can now apply to directly out of medical school. However, medical students are disadvantaged in that the integrated IR/DR pathway requires early decision when often radiology rotations are not part of the core clerkship curriculum. Based upon results from a survey to Integrated IR/DR Program Directors, we report strategies being used by programs to introduce and attract medical students to IR. Materials and Methods A questionnaire was written touching on various aspects of medical student engagement. The questionnaire was sent out electronically to 51 IR/DR Program Directors and answers were collated by the Society of Interventional Radiology Resident Fellow Student Section, IR Residency Training Committee. Results Eighteen responses were recorded from programs across the country. All programs encouraged applying to both DR and IR programs. All except one offered research opportunities (94%). The majority offered shadowing opportunities, had dedicated IR interest groups, and invited medical students to device workshops (78%). Planned informal opportunities for medical students to meet faculty and a dedicated department website were made available by most (67%). Little more than half invited medical students to journal clubs (59%). Formal medical student-faculty mentorship program and social media outreach initiatives like Facebook, Student Doctor Network, Twitter, LikedIn, Youtube, and podcasts rounded out the bottom two (50%). Importantly, respondents indicated that they were interested in hearing the results of the survey. Conclusion Our survey offers a snapshot of exactly what program directors are doing to address the issue of medical student recruitment.

Soft Tissue Sarcomas: Preoperative Predictive Histopathological Grading Based on Radiomics of MRI Publication date: September 2019 Source: Academic Radiology, Volume 26, Issue 9 Author(s): Yu Zhang, Yifeng Zhu, Xiaomeng Shi, Juan Tao, Jingjing Cui, Yue Dai, Minting Zheng, Shaowu Wang Rationale and Objectives The purpose of this study is to develop a radiomics model for predicting the histopathological grades of soft tissue sarcomas preoperatively through magnetic resonance imaging (MRI). Materials and Methods Thirty-five patients who were pathologically diagnosed with soft tissue sarcomas and their histological grades were recruited. All patients had undergone MRI before surgery on a 3.0T MRI scanner. Radiomics features were extracted from fat-suppressed T2-weighted imaging. We used the least absolute shrinkage and selection operator (LASSO) regression method to select features. Then three machine learning classification methods, including random forests, k-nearest neighbor, and support vector machine algorithm were trained using the 5-fold cross validation strategy to separate the soft tissue sarcomas with low- and high-histopathological grades. Results The radiomics features were significantly associated with the histopathological grades. Quantitative imaging features (n = 1049) were extracted from fat-suppressed T2-weighted imaging, and five features were selected to construct the radiomics model. The model that used support vector machine classification method achieved the best performance among the three methods, with areas under the receiver operating characteristic curves Area Under Curve (AUC) values of 0.92 ± 0.07, accuracy of 0.88. Conclusion Good accuracy and AUC could be obtained using only five radiomic features. Therefore, we proposed that three-dimensional imaging features from fat-suppressed T2-weighted imaging could be used as candidate biomarkers for preoperative prediction of histopathological grades of soft tissue sarcomas noninvasively.

Radiomics Signature: A Biomarker for the Preoperative Distant Metastatic Prediction of Stage I Nonsmall Cell Lung Cancer Publication date: September 2019 Source: Academic Radiology, Volume 26, Issue 9 Author(s): Li Fan, MengJie Fang, WenTing Tu, Di Zhang, Yun Wang, Xiuxiu Zhou, Yi Xia, ZhaoBin Li, ShiYuan Liu Objectives To evaluate the predictive value of radiomics features on the distant metastasis (DM) of stage I nonsmall cell lung cancer (NSCLC) preoperatively, by comparing with clinical characteristics and CT morphological features, and to screen the important prognostic predictors. Methods One hundred ninety-four stage I NSCLC patients were retrospectively enrolled, DM free survival (DMFS) was evaluated. The consensus clustering analysis was used to build the radiomics signatures in the primary cohort and validated in the validation cohort. The univariate survival analysis was performed in clinical characteristics, CT morphological features and radiomics signatures, respectively. Cox model was performed and C-index was calculated. Results There were 25 patients (12.9%) with DM. The median DMFS was 15 months. Three hundred thirteen radiomics features were selected, then classified into five groups, two subtypes (I and II) with each group. The RS1 showed the best prognostic ability with C-index of 0.355(95% confidence interval [CI], 0.269–0.442; p < 0.001). The histological type exhibited a good prognostic ability with C-index of 0.123 (95% CI, 0.000–0.305; p < 0.001) for DMFS. Cox model showed RS1(hazard ratio [HR] 18.025, 95% CI 2.366–137.340), pleural indentation sign (HR 2.623, 95% CI 1.070–6.426) and histological type (HR 4.461, 95% CI 1.783–11.162) were the independent prognostic factors (p < 0.05). Conclusion Radiomics provided a new modality for the distant metastatic prediction of stage I NSCLC. Patients with type II of RS1, pleural indentation sign and nonadenocarcinoma indicated the high probability of postsurgical DM.

Radiomics for Classification of Lung Cancer Histological Subtypes Based on Nonenhanced Computed Tomography Publication date: September 2019 Source: Academic Radiology, Volume 26, Issue 9 Author(s): Linning E, Lin Lu, Li Li, Hao Yang, Lawrence H. Schwartz, Binsheng Zhao Objectives To evaluate the performance of using radiomics method to classify lung cancer histological subtypes based on nonenhanced computed tomography images. Materials and Methods 278 patients with pathologically confirmed lung cancer were collected, including 181 nonsmall cell lung cancer (NSCLC) and 97 small cell lung cancers (SCLC) patients. Among the NSCLC patients, 88 patients were adenocarcinomas (AD) and 93 patients were squamous cell carcinomas (SCC). In total, 1695 quantitative radiomic features (QRF) were calculated from the primary lung cancer tumor in each patient. To build radiomic classification model based on the extracted QRFs, several machine-learning algorithms were applied sequentially. First, unsupervised hierarchical clustering was used to exclude highly correlated QRFs; second, the minimum Redundancy Maximum Relevance feature selection algorithm was employed to select informative and nonredundant QRFs; finally, the Incremental Forward Search and Support Vector Machine classification algorithms were used to combine the selected QRFs and build the model. In our work, to study the phenotypic differences among lung cancer histological subtypes, four classification models were built. They were models of SCLC vs NSCLC, SCLC vs AD, SCLC vs SCC, and AD vs SCC. The performance of the classification models was evaluated by the area under the receiver operating characteristic curve (AUC) estimated by three-fold cross-validation. Results The AUC (95% confidence interval) for the model of SCLC vs NSCLC was 0.741(0.678, 0.795). For the models of SCLC vs AD and SCLC vs SCC, the AUCs were 0.822(0.755, 0.875) and 0.665(0.583, 0.738), respectively. The AUC for the model of AD vs SCC was 0.655(0.570, 0.731). Several QRFs (“Law_15,” “LoG_Uniformity,” “GLCM_Contrast,” and “Compactness Factor”) that characterize tumor heterogeneity and shape were selected as the significant features to build the models. Conclusion Our results show that phenotypic differences exist among different lung cancer histological subtypes on nonenhanced computed tomography image.

Separating High-Z Oral Contrast From Intravascular Iodine Contrast in an Animal Model Using Dual-Layer Spectral CT Publication date: September 2019 Source: Academic Radiology, Volume 26, Issue 9 Author(s): Todd C. Soesbe, Matthew A. Lewis, Khaled Nasr, Lakshmi Ananthakrishnan, Robert E. Lenkinski Rationale and Objectives To show that water and iodine two-material decomposition images from dual-layer dual-energy spectral X-ray computed tomography (DECT) can be used to separate intravascular iodine contrast from simultaneously administered oral tantalum, tungsten, or rhenium contrast in an animal model. Materials and Methods In this Institutional Animal Care and Use Committee approved study, four female Fischer rats were given simultaneous intravenous and oral X-ray computed tomography contrast. Intravenous iodine contrast was administered via tail vein injection. Oral barium, tantalum, tungsten, or rhenium contrast was administered via gavage. The animals were imaged on a dual-layer DECT system at 120 kVp. Water and iodine two-material decomposition images (water equivalent and iodine equivalent images) were used for qualitative analysis. Computer simulations were performed using a customized DECT simulator to better understand why certain high-Z elements disappear in the iodine equivalent images and what is the theoretical range of elements with this property. Results The iodine and barium contrast appeared only in the iodine equivalent images and could not be differentiated from each other. However, the tantalum, tungsten, and rhenium contrast only appeared in the water equivalent images. This allowed iodine contrast in the bowel wall to be easily segmented from tantalum, tungsten, and rhenium contrast in the bowel lumen. Simulations confirmed that certain high-Z elements will have pixel values of ≤0 mg iodine/mL in the iodine equivalent images due to a K-edge effect associated with DECT systems. Conclusions Dual-layer DECT can separate iodine from certain high-Z elements using water equivalent and iodine equivalent images with an increased element range compared to other DECT systems. This K-edge effect could promote the development and approval of new high-Z contrast agents for DECT.

How Certain Are Your Radiology Reports And Are We Alone in Our Uncertainty? Publication date: September 2019 Source: Academic Radiology, Volume 26, Issue 9 Author(s): Elizabeth A. Krupinski

Multivariate Analysis of Radiologists’ Usage of Phrases that Convey Diagnostic Certainty Publication date: September 2019 Source: Academic Radiology, Volume 26, Issue 9 Author(s): Ronilda Lacson, Eseosa Odigie, Aijia Wang, Neena Kapoor, Atul Shinagare, Giles Boland, Ramin Khorasani Rationale and Objectives To quantify the use of Diagnostic Certainty Phrases (DCP) in radiology reports, including DCPs with good agreement (including “diagnostic of,” “unlikely” and “represents”) in connoting degree of certainty between providers based on previous studies; and to assess whether modality, presence of a trainee, radiologic subspecialty, and individual radiologists are associated with the usage of DCPs with good agreement. Materials and Methods This retrospective, IRB-approved study was conducted at an academic medical center. Radiology reports that contain DCPs were identified using information retrieval from all reports generated in 2016, excluding mammograms, obstetrical ultrasound, bone densitometry, and interventional studies. DCPs connoting good agreement were further noted. Of the reports that contained DCPs, a two-level hierarchical generalized linear model with attending as the level-two variable was performed comparing the use of DCP with good agreement while considering trainee involvement, modality, and subspecialty. Results A total of 159,151 reports out of 370,881 were found to have at least one DCP (43%). Reports of CT scans had the most number of DCP (68% of all CT reports). Breast and abdomen subspecialties were associated with use of DCP with good agreement. There was significant variation in use of DCP with good agreement between physicians that could not be explained by modality, trainee presence, and subspecialty. Conclusion Phrases to convey diagnostic certainty were commonly used in radiology reports. There is wide variation in usage of DCP with good agreement. Future interventions to reduce variation in use of DCPs may reduce ambiguity and improve quality of radiology reports.

Role of Iterative Reconstruction Algorithm for the Assessment of Myocardial Infarction with Dual Energy Computed Tomography Publication date: September 2019 Source: Academic Radiology, Volume 26, Issue 9 Author(s): Gaston A. Rodriguez-Granillo, Alejandro Deviggiano, Carlos Capunay, Macarena De Zan, Carlos Fernandez-Pereira, Patricia Carrascosa Rationale and Objectives Low monochromatic energy levels (40 keV) derived from delayed enhancement dual energy cardiac computed tomography (DE-DECT) allow the evaluation of myocardial infarcts (MI) among stable patients, although at the expense of high image noise. We explored whether the application of adaptive statistical iterative reconstruction (ASIR) to 40-keV DE-DECT (unavailable with previous software versions) might improve image quality and detection of MI in stable patients. Materials and Methods We prospectively enrolled patients with a history of previous MI, and performed delayed-enhancement cardiac magnetic resonance (DE-CMR) and DE-DECT within the same week. DE-DECT images were reconstructed with 0% and 60% ASIR. Results MI was identified in 18 (80%) patients with both DE-CMR and DE-DECT. On a per segment basis, we did not identify significant differences regarding the diagnostic performance of DE-DECT with and without ASIR [area under receiver operating characteristic curve 0.86 vs. 0.83, p = 0.10]. The application of ASIR improved the signal-to-noise ratio of DE-DECT with 0% ASIR compared to DE-DECT with 60% ASIR (6.07 ± 2.1 vs. 11.1 ± 4.5, p < 0.0001). However, qualitative assessment of MI image quality (3.35 ± 1.2, vs. 3.55 ± 1.1, p = 0.10) and diagnostic confidence (4.40 ± 0.9 vs. 4.60 ± 0.8, p = 0.10) were not significantly improved. Using DE-DECT with 60% ASIR, a threshold over 199 HU showed a sensitivity of 67% and a specificity of 92% for the detection of segments with MI. Conclusion In this study, DE-DECT allowed accurate detection of MI among stable patients compared with DE-CMR, and the application of ASIR improved signal-to-noise ratio of DE-DECT, although the diagnostic performance showed only non-significant improvements.

Utility of 13N-Ammonia PET/CT to Detect Pituitary Tissue in Patients with Pituitary Adenomas Publication date: September 2019 Source: Academic Radiology, Volume 26, Issue 9 Author(s): Zongming Wang, Zhigang Mao, Xiangsong Zhang, Dongsheng He, Xin Wang, Qiu Du, Zheng Xiao, Diming Zhu, Yonghong Zhu, Haijun Wang Rationale and Objectives It is clinically essential, but sometimes challenging, to distinguish pituitary tissue from pituitary adenomas (PAs). It is helpful to avoid damage of pituitary tissue during management. We evaluated the ability of 13N-ammonia positron emission tomography (PET)/computed tomography (CT) to locate and distinguish pituitary tissue from PAs. Materials and Methods Forty-eight patients (four with prolactinoma, 10 with Cushing's disease, 12 with acromegaly, and 22 with nonfunctional PAs) prospectively underwent magnetic resonance imaging (MRI), 13N-ammonia PET/CT, 18F-FDG PET/CT, prior to surgery. Results Pituitary position could be determined in 31 (64.5%) patients by 13N-ammonia PET/CT, and by MRI in 26 (54.2%) patients. It was detected by 13N-ammonia PET/CT and MRI in eight of eight patients (100%) with pituitary microadenoma, tumor maximum diameter (TMD) <1 cm, and in nine of 10 patients (90%) with PAs with TMD ≥1 cm, but <2 cm. In 16 patients with PAs with TMD ≥2 cm, but <3 cm, pituitary tissue position was detected by 13N-ammonia PET/CT in nine (56%), and by MRI in 8 (50.0%) patients by MRI. In 14 patients with PAs with TMD ≥3 cm, pituitary tissue position was detected by 13N-ammonia PET/CT in five (35.7%) patient, and by MRI in 1 (7.1%). In seven patients, the pituitary tissue could be detected by 13N-ammonia PET, but not by MRI, and in two patients by MRI, but not by 13N-ammonia PET. Conclusion 13N-ammonia PET/CT imaging is a sensitive means for locating and distinguishing pituitary tissue from PAs, particularly those with TMD <2 cm. It is potentially valuable in detection of pituitary tissue in patients with PAs.