Can incorrect artificial intelligence (AI) results impact radiologists, and if so, what can we do about it? A multi-reader pilot study of lung cancer detection with chest radiography.

OBJECTIVE: To examine whether incorrect AI results impact radiologist performance, and if so, whether human factors can be optimized to reduce error. METHODS: Multi-reader design, 6 radiologists interpreted 90 identical chest radiographs (follow-up CT needed: yes/no) on four occasions (09/20-01/22). No AI result was provided for session 1. Sham AI results were provided for sessions 2-4, and AI for 12 cases were manipulated to be incorrect (8 false positives (FP), 4 false negatives (FN)) (0.87 ROC-AUC). In the Delete AI (No Box) condition, radiologists were told AI results would not be saved for the evaluation. In Keep AI (No Box) and Keep AI (Box), radiologists were told results would be saved. In Keep AI (Box), the ostensible AI program visually outlined the region of suspicion. AI results were constant between conditions. RESULTS: Relative to the No AI condition (FN = 2.7%, FP = 51.4%), FN and FPs were higher in the Keep AI (No Box) (FN = 33.0%, FP = 86.0%), Delete AI (No Box) (FN = 26.7%, FP = 80.5%), and Keep AI (Box) (FN = to 20.7%, FP = 80.5%) conditions (all ps < 0.05). FNs were higher in the Keep AI (No Box) condition (33.0%) than in the Keep AI (Box) condition (20.7%) (p = 0.04). FPs were higher in the Keep AI (No Box) (86.0%) condition than in the Delete AI (No Box) condition (80.5%) (p = 0.03). CONCLUSION: Incorrect AI causes radiologists to make incorrect follow-up decisions when they were correct without AI. This effect is mitigated when radiologists believe AI will be deleted from the patient's file or a box is provided around the region of interest. CLINICAL RELEVANCE STATEMENT: When AI is wrong, radiologists make more errors than they would have without AI. Based on human factors psychology, our manuscript provides evidence for two AI implementation strategies that reduce the deleterious effects of incorrect AI. KEY POINTS: • When AI provided incorrect results, false negative and false positive rates among the radiologists increased. • False positives decreased when AI results were deleted, versus kept, in the patient's record. • False negatives and false positives decreased when AI visually outlined the region of suspicion.

Aorta-cameral fistulas: Spectrum of CT findings and review.

Aorto-cameral fistula (ACF) is an uncommon entity, defined as an abnormal communication between the aorta and a cardiac chamber. The most common causes include ruptured sinus of Valsalva aneurysm, infective endocarditis, traumatic injury, aortic dissection, or rarely can be iatrogenic in nature. While smaller communications may initially be asymptomatic, the natural course of these connections is generally refractory heart failure as they do not spontaneously heal. Larger fistulas can be life threatening with high mortality rates, and therefore once recognized, surgery is generally considered the treatment of choice. Diagnosis, however, can be challenging, and various imaging modalities are often used for diagnosis. This review highlights common underlying etiologies, clinical manifestations, and radiologic imaging appearances of ACF to each of the cardiac chambers of this uncommon, but clinically important entity, with emphasis on CT.

Eosinophilic Myocarditis in a Patient With Sickle Cell Disease.

A 23-year-old man with sickle cell disease treated with splenectomy and allogenic stem cell transplantation presented with recurrent chest pain, elevated cardiac enzymes, and unremarkable electrocardiography. His work-up revealed eosinophilia, raising concern for eosinophilic myocarditis. Cardiac magnetic resonance imaging showed patchy late gadolinium enhancement of the left ventricular free wall, suggestive of myocarditis. He was treated with high-dose intravenous steroids followed by oral prednisone, with improvement in his symptoms and eosinophilia and a decrease in cardiac enhancement on follow-up imaging. (Level of Difficulty: Intermediate.).

Prognostication of patients with COVID-19 using artificial intelligence based on chest x-rays and clinical data: a retrospective study.

BACKGROUND: Chest x-ray is a relatively accessible, inexpensive, fast imaging modality that might be valuable in the prognostication of patients with COVID-19. We aimed to develop and evaluate an artificial intelligence system using chest x-rays and clinical data to predict disease severity and progression in patients with COVID-19. METHODS: We did a retrospective study in multiple hospitals in the University of Pennsylvania Health System in Philadelphia, PA, USA, and Brown University affiliated hospitals in Providence, RI, USA. Patients who presented to a hospital in the University of Pennsylvania Health System via the emergency department, with a diagnosis of COVID-19 confirmed by RT-PCR and with an available chest x-ray from their initial presentation or admission, were retrospectively identified and randomly divided into training, validation, and test sets (7:1:2). Using the chest x-rays as input to an EfficientNet deep neural network and clinical data, models were trained to predict the binary outcome of disease severity (ie, critical or non-critical). The deep-learning features extracted from the model and clinical data were used to build time-to-event models to predict the risk of disease progression. The models were externally tested on patients who presented to an independent multicentre institution, Brown University affiliated hospitals, and compared with severity scores provided by radiologists. FINDINGS: 1834 patients who presented via the University of Pennsylvania Health System between March 9 and July 20, 2020, were identified and assigned to the model training (n=1285), validation (n=183), or testing (n=366) sets. 475 patients who presented via the Brown University affiliated hospitals between March 1 and July 18, 2020, were identified for external testing of the models. When chest x-rays were added to clinical data for severity prediction, area under the receiver operating characteristic curve (ROC-AUC) increased from 0·821 (95% CI 0·796-0·828) to 0·846 (0·815-0·852; p<0·0001) on internal testing and 0·731 (0·712-0·738) to 0·792 (0·780-0 ·803; p<0·0001) on external testing. When deep-learning features were added to clinical data for progression prediction, the concordance index (C-index) increased from 0·769 (0·755-0·786) to 0·805 (0·800-0·820; p<0·0001) on internal testing and 0·707 (0·695-0·729) to 0·752 (0·739-0·764; p<0·0001) on external testing. The image and clinical data combined model had significantly better prognostic performance than combined severity scores and clinical data on internal testing (C-index 0·805 vs 0·781; p=0·0002) and external testing (C-index 0·752 vs 0·715; p<0·0001). INTERPRETATION: In patients with COVID-19, artificial intelligence based on chest x-rays had better prognostic performance than clinical data or radiologist-derived severity scores. Using artificial intelligence, chest x-rays can augment clinical data in predicting the risk of progression to critical illness in patients with COVID-19. FUNDING: Brown University, Amazon Web Services Diagnostic Development Initiative, Radiological Society of North America, National Cancer Institute and National Institute of Biomedical Imaging and Bioengineering of the National Institutes of Health.

Multimodal Imaging after Sudden Cardiac Arrest in an 18-Year-Old Athlete.

We report the case of a previously healthy 18-year-old male athlete who twice presented with sudden cardiac arrest. Our use of electrocardiography, echocardiography, cardiac magnetic resonance, coronary angiography, coronary computed tomographic angiography, and nuclear stress testing enabled the diagnoses of apical hypertrophic cardiomyopathy and anomalous origin of the right coronary artery. We discuss the patient's treatment and note the useful role of multiple cardiovascular imaging methods in cases of sudden cardiac arrest.

Quantitative Assessment of Right Ventricular Volumes and Ejection Fraction in Patients with Left Ventricular Systolic Dysfunction by Real Time Three-Dimensional Echocardiography versus Cardiac Magnetic Resonance Imaging.

AIMS: The aim of this study was to assess the accuracy and reproducibility of real time three-dimensional echocardiographic (RT3DE) for the determination of right ventricular (RV) volumes and function in patients with left ventricular (LV) systolic dysfunction. METHODS AND RESULTS: Dedicated RT3DE was prospectively performed to assess RV volumes and EF in patients with LV systolic function identified on routine clinical cardiac magnetic resonance (CMR) imaging. RV end-diastolic volume (RV EDV), RV end-systolic volume (RV ESV), and RV EF were obtained using an offline analysis software (TomTec) by two observers blinded to CMR results. In this population of 27 patients with LV systolic dysfunction with a mean LV EF of 36 ± 12%, RV RT3DE dataset could be assessed in 27 of 30 patients (90%). High correlation was noted between RT3DE and CMR for RV EDV, ESV, and EF (r = 0.90, 0.89, and 0.77, respectively). RV EDV was lower by RT3DE as compared to CMR (129 ± 52 vs. 142 ± 53 mL, P = 0.005) while there was no significant difference in RV ESV and RV EF (71 ± 37 vs. 77 ± 45 mL, P = 0.146; 45 ± 11 vs. 48 ± 13%, P = 0.134, respectively). The intraclass correlation coefficient ranged from 0.94 to 0.94 between measurements and from 0.84 to 0.96 between observers. CONCLUSION: Overall, RV volumes and EF assessed by RT3DE correlate well with CMR measurements in patients with LV dysfunction. RT3DE may be used as a more widely available and versatile alternative to CMR for the quantitative assessment of RV size and function in patients with LV dysfunction.

Resident experience increases diagnostic rate of thyroid fine-needle aspiration biopsies.

RATIONALE AND OBJECTIVES: The aim of this study was to determine whether the diagnostic yield of thyroid fine-needle aspirations (FNAs) changes over the course of residency training. MATERIALS AND METHODS: We identified 5418 ultrasound-guided thyroid nodule FNAs performed in our radiology department from 2004 through 2012. For each FNA, we recorded if the FNA was performed by a resident and if so the name of the resident and supervising attending radiologist. For each resident, we determined the level of training based on their graduation year from our residency program and the date of the FNA as well as prior surgical training and if they completed subsequent interventional radiology fellowship. Pathology reports were reviewed, and FNAs were classified as diagnostic or nondiagnostic (ND). Generalized mixed models were used to assess ND rate with postgraduate years, including residents with and without prior surgical training or if they subsequently completed an interventional radiology fellowship. RESULTS: Of the 5418 thyroid FNAs, 3164 (58.4%) were performed by a radiology resident under the direct supervision of an attending physician. There was a significant decrease in ND rate as postgraduate years increased (P < .05). A significant decrease in ND rate was found as postgraduate years increased for residents without prior surgical training (P = .0007) or subsequent training in interventional radiology (P = .0014); however, no significant decrease was found for residents with surgical training (P = .37) or completing an interventional radiology fellowship (P = .08). In addition, no significant difference was found for ND rate between postgraduate year 4 (PGY4) and PGY5 (P > .05). CONCLUSIONS: ND thyroid FNA rates progressively decrease with training level, suggesting that early and continued participation in procedures throughout residency improves outcomes. This is particularly true for residents without prior surgical training or subsequent interventional radiology fellowship.

Management of nodules with initially nondiagnostic results of thyroid fine-needle aspiration: can we avoid repeat biopsy?

PURPOSE: To identify demographic and ultrasonographic (US) features associated with malignancy after initially nondiagnostic results of fine-needle aspiration (FNA) to help clarify the role of repeat FNA, surgical excision, or serial US in these nodules. MATERIALS AND METHODS: This study was HIPAA compliant and institutional review board approved; informed consent was waived. Thyroid nodules (n = 5349) that underwent US-guided FNA in 2004-2012 were identified; 393 were single nodules with nondiagnostic FNA results but adequate cytologic, surgical, or US follow-up. Demographic information and diameters and volume at US at first biopsy were modeled with malignancy as outcome through medical record review. Exact logistic regression was used to model malignancy outcomes, demographic comparisons with age were made (Student t test, Satterthwaite test), and proportion confidence intervals (CIs) were estimated (Clopper-Pearson method). RESULTS: Of 393 nodules with initially nondiagnostic results, nine malignancies (2.3%) were subsequently diagnosed with repeat FNA (n = 2, 0.5%) or surgical pathologic examination (n = 7, 1.8%), 330 (84.0%) were benign, and 54 (13.7%) were stable or decreased in size at serial US (mean follow-up, 3.0 years; median, 2.5 years; range, 1.0-7.8 years). Patients with malignancies were significantly older (mean age, 62.7 years; median, 64 years; range, 47-77 years) than those without (mean age, 55.4 years; median, 57 years; range, 12-94 years; P = .0392). Odds of malignancy were 4.2 times higher for men versus women (P = .045) and increased significantly for each 1-cm increase in anteroposterior, minimum, and mean nodule diameter (1.78, 2.10, and 1.96, respectively). In 393 nodules, no malignancies were detected in cystic or spongiform nodules (both, n = 11, 2.8%; 95% CI: 1.4%, 5.0%), nodules with eggshell calcifications (n = 9, 2.3%; 95% CI: 1.1%, 4.3%), or indeterminate echogenic foci (n = 39, 9.9%; 95% CI: 7.2%, 13.3%). CONCLUSION: Very few malignancies were diagnosed with repeat FNA following nondiagnostic FNA results (two of 336, 0.6%); therefore, clinical and US follow-up may be more appropriate than repeat FNA following nondiagnostic biopsy results.

Prospective evaluation of cinefluoroscopy and chest radiography for Riata lead defects: implications for future lead screening.

INTRODUCTION: Lead insulation defects with externalization of the conductors exist in Riata defibrillator leads. Cinefluoroscopy is currently the gold standard to detect such defects. Prospective evaluation of alternative screening options such as chest radiography (CXR), which has been recommended by the FDA, is not well described. METHODS AND RESULTS: Patients with Riata leads underwent cinefluoroscopy, CXR, and device interrogation. Leads were classified as abnormal (clear cable separation), borderline, or normal by independent evaluation of cinefluoroscopy and CXR. CXR evaluation was done in two ways as follows: (1) routine CXR read by daily staff radiologists for lead screening and (2) CXR evaluation by a radiologist educated about the lead defect. One hundred two patients were evaluated at our institution. Cinefluoroscopy showed externalized conductors in 33 patients (32 %). Twenty-five of 33 patients (76 %) who had abnormal cinefluoroscopic findings had abnormal CXR findings on blinded review by the educated radiologist. All 25 patients with abnormal CXR had abnormal findings on cinefluoroscopy. Daily staff radiologists without direct education other than prompts for lead screening detected CXR abnormalities in only 8 out of 102 (8 %) cases. CONCLUSION: Cinefluoroscopy appears to be more sensitive than CXR for the detection of Riata cable extrusion. Interpretation of CXR by a radiologist with education in lead defects correlates highly with cinefluoroscopy with very high specificity. Depending on available resources for screening, CXR may be a reasonable alternative to cinefluoroscopy. Multidisciplinary collaboration across specialties (radiology and electrophysiology) can lead to improved diagnostic capability and thus the potential for enhanced quality of care.

Value of cardiovascular magnetic resonance in suspected cardiac amyloidosis.

The diagnosis of cardiac amyloid can be challenging; requiring a high clinical index of suspicion and often many diagnostic tests to confirm. We describe a case demonstrating the characteristic pattern of amyloid by cardiovascular magnetic resonance and how this imaging modality can aid in the diagnosis.

ACR Appropriateness Criteria® chronic chest pain--high probability of coronary artery disease.

Imaging is valuable in determining the presence, extent, and severity of myocardial ischemia and the severity of obstructive coronary lesions in patients with chronic chest pain in the setting of high probability of coronary artery disease. Imaging is critical for defining patients best suited for medical therapy or intervention, and findings can be used to predict long-term prognosis and the likely benefit from various therapeutic options. Chest radiography, radionuclide single photon-emission CT, radionuclide ventriculography, and conventional coronary angiography are the imaging modalities historically used in evaluating suspected chronic myocardial ischemia. Stress echocardiography, PET, cardiac MRI, and multidetector cardiac CT have all been more recently shown to be valuable in the evaluation of ischemic heart disease. Other imaging techniques may be helpful in those patients who do not present with signs classic for angina pectoris or in those patients who do not respond as expected to standard management. The ACR Appropriateness Criteria(®) are evidence-based guidelines for specific clinical conditions that are reviewed every 2 years by a multidisciplinary expert panel. The guideline development and review include an extensive analysis of current medical literature from peer-reviewed journals and the application of a well-established consensus methodology (modified Delphi) to rate the appropriateness of imaging and treatment procedures by the panel. In those instances in which evidence is lacking or not definitive, expert opinion may be used to recommend imaging or treatment.

The prevalence and clinical significance of noncardiac findings on cardiac MRI.

OBJECTIVE: This study aims to evaluate the prevalence, nature, and clinical significance of noncardiac findings (NCFs) at cardiac MRI. MATERIALS AND METHODS: We retrospectively reviewed 240 consecutive, clinically indicated cardiac MRI examinations conducted over a 21-month period. All noncardiac findings (NCFs) were recorded. Those findings that were included in the report impressions were regarded as clinically important (INCF). Electronic medical records and related imaging studies were then reviewed for all patients having INCFs to determine their actual clinical significance. A finding was significant (SNCF) if it was associated with a new diagnosis, treatment, or intervention. The prevalences of findings in the neck, chest, and abdomen were determined. RESULTS: We found 162 NCFs in 104 studies (43%), of which 94 (58%) were INCFs, and 16 (10%) were SNCFs. There was at least one INCF in 65 studies (27%)--67% of which were new--and at least one SNCF in 13 studies (5%). Compared with younger patients, patients 60 years and older were much more likely to have INCFs (43% vs 17%) and SNCFs (12% vs 1%). Overall, 29% of NCFs were in the abdomen, 70% in the chest, and 1% in the neck. The most common INCFs were pleural effusion (n = 26), air-space disease or atelectasis (n = 13), and adenopathy (n = 9). Five new cases of cancer were diagnosed, including lung (n = 2), lymphoma (n = 2), and thyroid (n = 1). CONCLUSION: NCFs are commonly encountered on cardiac MRI studies, many of which are clinically relevant. Proper recognition of NCFs is critical to the comprehensive management of patients referred for cardiac MRI.

Referrals to a dedicated cardiac MRI service: who sends what?

The purpose of this study was to retrospectively evaluate the referral patterns to the authors' cardiac magnetic resonance imaging service in its first 28 months of operation. In late November 2003, the authors' radiology practice established a cardiac magnetic resonance imaging service for a 719-bed teaching hospital and a 247-bed community hospital. Data relevant to referrals were reviewed. Between December 1, 2003, and April 1, 2006, 780 patients were imaged, 556 (71%) at the teaching hospital. Referrals came from 157 physicians in 17 different medical specialties, including adult cardiology (64%), cardiothoracic surgery (15%), pediatric cardiology (8%), internal medicine (6%), and others (7%). Overall, primary indications were function and viability evaluation (29%), aorta and valve assessment (24%), congenital heart disease (17%), arrhythmogenic right ventricular dysplasia exclusion (13%), cardiac masses (9%), pericardial disease (4%), and others (4%). Referrals for function and viability represented a much greater percentage of cases at the community hospital (53%) than at the teaching hospital (19%). The reverse was true for congenital heart disease (7% vs 21%). This study demonstrates that cardiac specialists generate the vast majority of referrals (87%) to the authors' cardiac magnetic resonance imaging service and that there are substantial differences in the referral patterns between a large teaching hospital and a smaller community hospital.

Evaluation of the cardiac surgery patient by MRI and CT imaging: the state of the art.

Over the last few years, the landscape of cardiovascular imaging has changed significantly. CT and MRI have emerged as robust imaging tools for diagnosing and characterizing a nearly exhaustive spectrum of cardiovascular diseases. With ECG-gating, advanced tomographic imaging techniques are available for imaging the beating heart and surrounding structures with unprecedented accuracy and precision. These new methods provide clinicians with substantial additional information, including--in the case of MRI--quantitative data on blood flow physiology and accurate measurements of ventricular volumes and function. Importantly, MRI has made recent dramatic strides in the evaluation of ischemic heart disease, and CT appears poised to provide an accurate alternative to catheter angiography in the anatomic assessment of obstructive coronary artery disease. The future of both of these modalities in the non-invasive evaluation of heart disease is bright.