Faculty Bibliography

The left ventricular outflow tract (LVOT) velocity time integral (VTI) is an easily measured echocardiographic stroke volume index analog. Low values predict adverse outcomes in left ventricular failure. We postulate the left ventricular VTI may be a signal of right ventricular dysfunction in acute pulmonary embolism, and therefore a predictor of poor outcomes. We retrospectively reviewed echocardiograms on all Pulmonary Embolism Response Team activations at our institution at the time of pulmonary embolism diagnosis. Low LVOT VTI was defined as ⩽ 15 cm. We examined two composite outcomes: (1) in-hospital death or cardiac arrest; and (2) shock or need for primary reperfusion therapies. Sixty-one of 188 patients (32%) had a LVOT VTI of ⩽ 15 cm. Low VTI was associated with in-hospital death or cardiac arrest (odds ratio (OR) 6, 95% CI 2, 17.9; p = 0.0014) and shock or need for reperfusion (OR 23.3, 95% CI 6.6, 82.1; p < 0.0001). In a multivariable model, LVOT VTI ⩽ 15 remained significant for death or cardiac arrest (OR 3.48, 95% CI 1.02, 11.9; p = 0.047) and for shock or need for reperfusion (OR 8.12, 95% CI 1.62, 40.66; p = 0.011). Among intermediate-high-risk patients, low VTI was the only variable associated with the composite outcome of death, cardiac arrest, shock, or need for reperfusion (OR 14, 95% CI 1.7, 118.4; p = 0.015). LVOT VTI is associated with adverse short-term outcomes in acute pulmonary embolism. The VTI may help risk stratify patients with intermediate-high-risk pulmonary embolism.

Markov random field (MRF) has been widely used to incorporate a priori knowledge as penalty or regularizer to preserve edge sharpness while smoothing the region enclosed by the edge for pieces-wise smooth image reconstruction. In our earlier study, we proposed a type of MRF reconstruction method for low-dose CT (LdCT) scans using tissue-specific textures extracted from the same patient's previous full-dose CT (FdCT) scans as prior knowledge. It showed advantages in clinical applications. This paper aims to remove the constraint of using previous data of the same patient. We investigated the feasibility of extracting the tissue-specific MRF textures from an FdCT database to reconstruct a LdCT image of another patient. This feasibility study was carried out by experiments designed as follows. We constructed a tissue-specific MRF-texture database from 3990 FdCT scan slices of 133 patients who were scheduled for lung nodule biopsy. Each patient had one FdCT scan (120 kVp/100 mAs) and one LdCT scan (120 kVp/20 mAs) prior to biopsy procedure. When reconstructing the LdCT image of one patient among the 133 patients, we ranked the closeness of the MRF-textures from the other 132 patients saved in the database and used them as the a prior knowledge. Then, we evaluated the reconstructed image quality using Haralick texture measures. For any patient within our database, we found more than eighteen patients' FdCT MRF texures can be used without noticeably changing the Haralick texture measures on the lung nodules (to be biopsied). These experimental outcomes indicate it is promising that a sizable FdCT texture database could be used to enhance Bayesian reconstructions of any incoming LdCT scans.

Objectives: To differentiate pulmonary tumor type by volumetric iodine quantification and texture analysis on dual-energy CT images.
Material(s) and Method(s): Radiology information system search for all contrast-enhanced DECT chest examinations from 1/1/2015-4/30/ 2018 was performed, filtering for those with pathology within 120 days, yielding 80 cases of pathologically-proven pulmonary lesions. 73 lesions meeting inclusion criteria were manually volumetrically segmented via open-source software using the low-kV DECT dataset. 3D-iodine quantification was achieved by mapping between high/low energy HU on a representative 2D-image, and applying to surrounding slices, with absolute iodine normalized to mid-descending aorta. Full Width Tenth Maximum was applied to each normalized iodine histogram, providing a single comprehensive measure of relative iodine concentration. Volumetric iodine values and first order texture features were assessed using Hoteling's T-squared multivariate analysis.
Result(s): 72 individuals (37 women, 35 men) with mean age 64 years were included. 44 primary, 25 metastatic, and 3 benign lesions were assessed; 22 with history of chemotherapy. Mean time between histopathologic sampling and imaging was 26 days. Mean, median and minimum volumetric iodine concentration were significant (P< 0.05) in distinguishing primary versus metastatic lesions, with P<0.01 for these measures between de novo primary and metastatic lesions. Metastatic lesions demonstrated higher mean iodine (1.2 mg/mL) in comparison to primary lesions (0.83 mg/mL). Mean and median 3D-iodine values significantly (P<0.05) differed between primary lung adenocarcinoma and squamous lesions (mean 0.95 mg/mL and 0.49 mg/mL, respectively). Skewness significantly differed between de novo versus treated primary (P=0.01), and metastatic versus primary lesions (P=0.03), and entropy between de novo primary and metastatic lesions (P<0.001), and treated versus non-treated metastases (P= 0.03).
Conclusion(s): Volumetric iodine quantification significantly differs between de novo primary versus metastatic, and primary lung adenocarcinoma versus squamous lesions. Texture features may also have a role in distinguishing tumor type and treatment response. Clinical Relevance Application: Potential role of DECT in distinguishing tumor type and treatment response

Background: Subsolid nodules account for 20% of lesions detected on CT lung cancer screening, and many are incidentally detected on routine CT exams. Persistent subsolid nodules may correlate with adenocarcinoma spectrum, though commonly are due to other etiologies, emphasizing the importance of accurate nodule characterization and management recommendations. This review will used a case-based approach to showcase how lesion description and management may differ from that of solid nodules, including application of Fleischner guidelines and discussion of management options. Educational Goals/Teaching Points: * Highlight technical pitfalls in evaluation of subsolid nodules * Highlight interpretive pitfalls in evaluation of subsolid nodules * Review differential considerations including adenocarcinoma spectrum: AAH, AIS, MIA, Invasive adenocarcinoma * Describe differential considerations beyond the adenocarcinoma spectrum, such as: * Neoplastic: lymphoma, metastatic disease, treatment response * Infectious/Inflammatory: organizing pneumonia, eosinophilic pneumonia, alveolar sarcoid, various infection, drug reaction * Adapt management guidelines to diverse clinical scenarios: age, multiplicity, baseline versus follow-up, incidental versus screening * Explain treatment options for subsolid adenocarcinoma lesions including emerging technology such as cryoablation
Conclusion(s): Subsolid nodules are commonly encountered clinically, with accurate interpretation based on avoiding technical pitfalls and providing relevant descriptors to differentiate actionable findings. Management guidelines and further steps must take into account nodule features, clinical risk factors, and patient specific considerations

Background: Anterior mediastinal lesions may be radiographically challenging to diagnose. Helpful imaging findings for the radiologist and pathologist, and relevant reporting information for the clinician will be reviewed. * Educational Goals/Teaching Points: Review multimodality imaging of thymic lesions with histopathologic correlation * Thymic hyperplasia * Thymic cyst * Thymolipoma * Staging of thymic carcinoma * Highlight distinguishing features between common anterior mediastinal lesions * Lymphoma * Germ Cell tumor * Thyroid * Present atypical anterior mediastinal lesions * Ectopic Parathyroid * Lymphangioma * Lipoma and Lipomatosis * Sarcoma * Nodal mesothelioma
Conclusion(s): Pathologists and radiologists often work in tandem to arrive at a unifying diagnosis, and understanding the radiopathologic correlation of anterior mediastinal lesions will serve to improve diagnosis and reporting

Diseases that are predominantly peribronchovascular in distribution on computed tomography by definition involve the bronchi, adjacent vasculature, and associated lymphatics involving the central or axial lung interstitium. An understanding of diseases that can present with focal peribronchovascular findings is useful for establishing diagnoses and guiding patient management. This review will cover clinical and imaging features that may assist in differentiating amongst the various causes of primarily peribronchovascular disease.

PURPOSE/OBJECTIVE:To determine the utility of an automated radiology-pathology feedback tool. METHODS:We previously developed a tool that automatically provides radiologists with pathology results related to imaging examinations they interpreted. The tool also allows radiologists to mark the results as concordant or discordant. Five abdominal radiologists prospectively scored their own discordant results related to their previously interpreted abdominal ultrasound, CT, and MR interpretations between August 2017 and June 2018. Radiologists recorded whether they would have followed up on the case if there was no automated alert, reason for the discordance, whether the result required further action, prompted imaging rereview, influenced future interpretations, enhanced teaching files, or inspired a research idea. RESULTS:There were 234 total discordances (range 30-66 per radiologist), and 70.5% (165 of 234) of discordances would not have been manually followed up in the absence of the automated tool. Reasons for discordances included missed findings (10.7%; 25 of 234), misinterpreted findings (29.1%; 68 of 234), possible biopsy sampling error (13.3%; 31 of 234), and limitations of imaging techniques (32.1%; 75/234). In addition, 4.7% (11 of 234) required further radiologist action, including report addenda or discussion with referrer or pathologist, and 93.2% (218 of 234) prompted radiologists to rereview the images. Radiologists reported that they learned from 88% (206 of 234) of discordances, 38.6% (90 of 233) of discordances probably or definitely influenced future interpretations, 55.6% (130 of 234) of discordances prompted the radiologist to add the case to his or her teaching files, and 13.7% (32 of 233) inspired a research idea. CONCLUSION/CONCLUSIONS:Automated pathology feedback provides a valuable opportunity for radiologists across experience levels to learn, increase their skill, and improve patient care.

PURPOSE/OBJECTIVE:To develop natural language processing (NLP) to identify incidental lung nodules (ILNs) in radiology reports for assessment of management recommendations. METHOD AND MATERIALS/METHODS:We searched the electronic health records for patients who underwent chest CT during 2014 and 2017, before and after implementation of a department-wide dictation macro of the Fleischner Society recommendations. We randomly selected 950 unstructured chest CT reports and reviewed manually for ILNs. An NLP tool was trained and validated against the manually reviewed set, for the task of automated detection of ILNs with exclusion of previously known or definitively benign nodules. For ILNs found in the training and validation sets, we assessed whether reported management recommendations agreed with Fleischner Society guidelines. The guideline concordance of management recommendations was compared between 2014 and 2017. RESULTS:The NLP tool identified ILNs with sensitivity and specificity of 91.1% and 82.2%, respectively, in the validation set. Positive and negative predictive values were 59.7% and 97.0%. In reports of ILNs in the training and validation sets before versus after introduction of a Fleischner reporting macro, there was no difference in the proportion of reports with ILNs (108 of 500 [21.6%] versus 101 of 450 [22.4%]; P = .8), or in the proportion of reports with ILNs containing follow-up recommendations (75 of 108 [69.4%] versus 80 of 101 [79.2%]; P = .2]. Rates of recommendation guideline concordance were not significantly different before and after implementation of the standardized macro (52 of 75 [69.3%] versus 60 of 80 [75.0%]; P = .43). CONCLUSION/CONCLUSIONS:NLP reliably automates identification of ILNs in unstructured reports, pertinent to quality improvement efforts for ILN management.

Pyrite FeS₂ has long been considered a potential earth-abundant low-cost photovoltaic material for thin-film solar cells but has been plagued by low power conversion efficiencies and open-circuit voltages. Recent efforts have identified a lack of understanding and control of doping, as well as uncontrolled surface conduction, as key roadblocks to the development of pyrite photovoltaics. In particular, while n-type bulk behavior in unintentionally doped single crystals and thin films is speculated to arise from sulfur vacancies (V_S), proof remains elusive. Here, we provide strong evidence, from extensive electronic transport measurements on high-quality crystals, that V_S are deep donors in bulk pyrite. Otherwise identical crystals grown via chemical vapor transport under varied S vapor pressures are thoroughly characterized structurally and chemically, and shown to exhibit systematically different electronic transport. Decreased S vapor pressure during growth leads to reduced bulk resistivity, increased bulk Hall electron density, reduced transport activation energy, onset of positive temperature coefficient of resistivity, and approach to an insulator-metal transition, all as would be expected from increased V_S donor density. Impurity analyses show that these trends are uncorrelated with metal impurity concentration and that extracted donor densities significantly exceed total impurity concentrations, directly evidencing a native defect. Well-controlled, wide-range n-doping of pyrite is thus achieved via the control of V_S concentration, with substantial implications for photovoltaic and other applications. The location of the V_S state within the gap, the influence of specific impurities, unusual aspects to the insulator-metal transition, and the influence of doping on surface conduction are also discussed.