Diagnosis of acute heart failure in CT pulmonary angiography: feasibility and accuracy.

OBJECTIVES: To evaluate the feasibility and accuracy of diagnosing acute heart failure (HF) with CT pulmonary angiography (CTPA) in emergency department patients. METHODS: In this retrospective single-center study, we evaluated 150 emergency department patients (mean age 65 ± 17 years) undergoing CTPA with a fixed scan (100 kVp) and contrast media protocol (60 mL, 4 mL/s) who had no pulmonary embolism (PE). Patients were subdivided into training cohort (n = 100) and test cohort (n = 50). Three independent, blinded readers measured the attenuation in the right ventricle (RV) and left ventricle (LV) on axial images. The ratio (HUratio) and difference (HUdiff) between RV and LV attenuation were calculated. Diagnosis of acute HF was made on the basis of clinical, laboratory, and echocardiography data. Optimal thresholds, sensitivity, and specificity were calculated using the area under the curve (AUC) from receiver operating characteristics analysis. RESULTS: Fifty-nine of the 150 patients (40%) were diagnosed with acute HF. Attenuation measurements showed an almost perfect interobserver agreement (intraclass correlation coefficient: 0.986, 95%CI: 0.980-0.991). NT-pro BNP exhibited moderate correlations with HUratio (r = 0.50, p < 0.001) and HUdiff (r = 0.50, p < 0.001). In the training cohort, HUratio (AUC: 0.89, 95%CI: 0.82-0.95) and HUdiff (AUC: 0.88, 95%CI: 0.81-0.95) showed a very good performance to diagnose HF. Optimal cutoff values were 1.42 for HUratio (sensitivity 93%; specificity 75%) and 113 for HUdiff (sensitivity 93%; specificity 73%). Applying these thresholds to the test cohort yielded a sensitivity of 89% and 89% and a specificity of 69% and 63% for HUratio and HUdiff, respectively. CONCLUSION: In emergency department patients undergoing CTPA and showing no PE, both HUratio and HUdiff have a high sensitivity for diagnosing acute HF. KEY POINTS: • Heart failure is a common differential diagnosis in patients undergoing CT pulmonary angiography. • In emergency department patients undergoing CT pulmonary angiography and showing no pulmonary embolism, attenuation differences of the left and right ventricle have a high sensitivity for diagnosing acute heart failure.

Dual-Energy Computed Tomography in Stroke Imaging: Technical and Clinical Considerations of Virtual Noncontrast Images for Detection of the Hyperdense Artery Sign.

OBJECTIVE: The technical feasibility of virtual noncontrast (VNC) images from dual-energy computed tomography (DECT) for the detection of the hyperdense artery sign (HAS) in ischemic stroke patients was investigated. METHODS: True noncontrast (TNC) scans of 60 patients either with or without HAS (n = 30 each) were investigated. Clot presence and characteristics were assessed on VNC images from DECT angiography and compared with TNC images. Clot characterization included the level of confidence for diagnosing HAS, a qualitative clot burden score, and quantitative attenuation (Hounsfield unit [HU]) measurements. RESULTS: Sensitivity, specificity, and accuracy of VNC for diagnosing HAS were 97%, 90%, and 93%, respectively. No significant differences were found regarding the diagnostic confidence (P = 0.18) and clot burden score (P = 0.071). No significant HU differences were found among vessels with HAS in VNC (56 ± 7HU) and TNC (57 ± 8HU) (P = 0.691) images. CONCLUSIONS: Virtual noncontrast images derived from DECT enable an accurate detection and characterization of HAS.

Frequency and causes of delayed diagnosis of visceral artery pseudoaneurysms with CT: Lessons learned.

OBJECTIVE: Visceral artery pseudoaneurysms (VAPA) are associated with a high morbidity and mortality, but sometimes are missed in initial computed tomography (CT) examinations. The aims of this study were to determine the frequency and causes of misdiagnoses of VAPA with CT. MATERIALS AND METHODS: We retrospectively identified 77 patients with VAPA in our database who underwent contrast-enhanced CT. The frequency of delayed diagnosis was determined and the reasons were noted. We identified the etiology of VAPA, measured size, and noted the affected vessels. RESULTS: Forty-five of the 77 patients (58 %) had a delayed diagnosis of VAPA. There was no difference in the rate of missed VAPA in symptomatic compared to asymptomatic patients (p = 0.255). The majority of VAPA were associated with previous surgery or interventions (n = 48/62 %). The major affected vessel was the hepatic (n = 31) followed by the splenic artery (n = 17). The main reasons for misdiagnosis were a missed arterial phase in CT (n = 16/36 %), artifacts masking the aneurysm (n = 9/20 %), overlooked pseudoaneurysm (n = 19/42 %), and misinterpretation by attending radiologists (n = 1/2 %). Missed VAPA were smaller (median 8 mm) than those VAPA that were initially diagnosed (median 13 mm, p < 0.01), but occurred with a similar frequency in larger and smaller visceral arteries (p = 0.601). CONCLUSIONS: Our study showed that 58 % of VAPA were diagnosed with delay, with the following four reasons for misdiagnosis: Lack of an arterial contrast phase in CT, no techniques for artifact reduction, and lack of awareness of the radiologists. Avoiding delayed diagnosis will most probably improve outcome of patients with VAPA.

Detection and localization of distal radius fractures: Deep learning system versus radiologists.

PURPOSE: To evaluate a deep learning based image analysis software for the detection and localization of distal radius fractures. METHOD: A deep learning system (DLS) was trained on 524 wrist radiographs (166 showing fractures). Performance was tested on internal (100 radiographs, 42 showing fractures) and external test sets (200 radiographs, 100 showing fractures). Single and combined views of the radiographs were shown to DLS and three readers. Readers were asked to indicate fracture location with regions of interest (ROI). The DLS yielded scores (range 0-1) and a heatmap. Detection performance was expressed as AUC, sensitivity and specificity at the optimal threshold and compared to radiologists' performance. Heatmaps were compared to radiologists' ROIs. RESULTS: The DLS showed excellent performance on the internal test set (AUC 0.93 (95% confidence interval (CI) 0.82-0.98) - 0.96 (0.87-1.00), sensitivity 0.81 (0.58-0.95) - 0.90 (0.70-0.99), specificity 0.86 (0.68-0.96) - 1.0 (0.88-1.0)). DLS performance decreased on the external test set (AUC 0.80 (0.71-0.88) - 0.89 (0.81-0.94), sensitivity 0.64 (0.49-0.77) - 0.92 (0.81-0.98), specificity 0.60 (0.45-0.74) - 0.90 (0.78-0.97)). Radiologists' performance was comparable on internal data (sensitivity 0.71 (0.48-0.89) - 0.95 (0.76-1.0), specificity 0.52 (0.32-0.71) - 0.97 (0.82-1.0)) and better on external data (sensitivity 0.88 (0.76-0.96) - 0.98 (0.89-1.0), specificities 0.66 (0.51-0.79) - 1.0 (0.93-1.0), p < 0.05). In over 90%, the areas of peak activation aligned with radiologists' annotations. CONCLUSIONS: The DLS was able to detect and localize wrist fractures with a performance comparable to radiologists, using only a small dataset for training.

Determination of mammographic breast density using a deep convolutional neural network.

OBJECTIVE: High breast density is a risk factor for breast cancer. The aim of this study was to develop a deep convolutional neural network (dCNN) for the automatic classification of breast density based on the mammographic appearance of the tissue according to the American College of Radiology Breast Imaging Reporting and Data System (ACR BI-RADS) Atlas. METHODS: In this study, 20,578 mammography single views from 5221 different patients (58.3 ± 11.5 years) were downloaded from the picture archiving and communications system of our institution and automatically sorted according to the ACR density (a-d) provided by the corresponding radiological reports. A dCNN with 11 convolutional layers and 3 fully connected layers was trained and validated on an augmented dataset. The model was finally tested on two different datasets against: i) the radiological reports and ii) the consensus decision of two human readers. None of the test datasets was part of the dataset used for the training and validation of the algorithm. RESULTS: The optimal number of epochs was 91 for medio-lateral oblique (MLO) projections and 94 for cranio-caudal projections (CC), respectively. Accuracy for MLO projections obtained on the validation dataset was 90.9% (CC: 90.1%). Tested on the first test dataset of mammographies (850 MLO and 880 CC), the algorithm showed an accordance with the corresponding radiological reports of 71.7% for MLO and of 71.0% for CC. The agreement with the radiological reports improved in the differentiation between dense and fatty breast for both projections (MLO = 88.6% and CC = 89.9%). In the second test dataset of 200 mammographies, a good accordance was found between the consensus decision of the two readers on both, the MLO-model (92.2%) and the right craniocaudal-model (87.4%). In the differentiation between fatty (ACR A/B) and dense breasts (ACR C/D), the agreement reached 99% for the MLO and 96% for the CC projections, respectively. CONCLUSIONS: The dCNN allows for accurate classification of breast density based on the ACR BI-RADS system. The proposed technique may allow accurate, standardized, and observer independent breast density evaluation of mammographies. ADVANCES IN KNOWLEDGE: Standardized classification of mammographies by a dCNN could lead to a reduction of falsely classified breast densities, thereby allowing for a more accurate breast cancer risk assessment for the individual patient and a more reliable decision, whether additional ultrasound is recommended.

Qualitative versus quantitative lumbar spinal stenosis grading by machine learning supported texture analysis-Experience from the LSOS study cohort.

PURPOSE: To investigate and compare the reproducibility and accuracy of qualitative ratings and quantitative texture analysis (TA) in detection and grading of lumbar spinal stenosis (LSS) in magnetic resonance imaging (MR) scans of the lumbar spine. MATERIALS AND METHODS: From a nationwide multicenter and multidisciplinary lumbar stenosis outcome study (LSOS) register 82 patients, undergoing MR scans of the lumbar spine due to clinical indication of spinal claudication, with a single level central or lateral severe LSS were included. In total 343 transaxial T2-weighted images of the lumbar spine were included from one to five levels (L1 to S1) per patient. One expert radiologist serving as reference standard rated LSS grade according to a standard four-point (normal to severe) as well as to an eight-point Schizas grading scale. DICOM data were then rescaled to a defined pixel size. Two independent readers performed qualitative ratings analogous to expert reader in addition to TA of spinal canals by manually placing two regions of interest (ROI) per image reflecting qualitative scales: (1) dural sac only (2) inner contour of the spinal canal including epidural fat and bilateral recesses. Interreader agreements of qualitative and quantitative parameters were assessed by Cohen's Kappa (κ) and intraclass correlation (ICC), respectively. TA feature reduction was performed by ICC threshold > 0.75. Remaining features were analyzed with machine learning algorithms (Weka 3 tool) for correlation with LSS grades using 10-fold cross validation. RESULTS: Qualitative ratings showed only moderate reproducibility for both LSS classification systems but high correlation with cut-off cross-sectional area (CSA) <130mm² for severe spinal stenosis. In quantitative TA of both ROIs, machine learning analysis with a decision tree classifier revealed higher performances for LSS grading compared to qualitative assessments using the reference CSA cut-off, respectively. CONCLUSION: Qualitative LSS grading independent of classification system shows moderate reproducibility. TA with machine learning offers highly reproducible quantitative parameters that increase accuracy for severe LSS detection with minor impact of grading score and CSA border definition.

Vessel suppressed chest Computed Tomography for semi-automated volumetric measurements of solid pulmonary nodules.

OBJECTIVE: To evaluate whether vessel-suppressed computed tomography (VSCT) can be reliably used for semi-automated volumetric measurements of solid pulmonary nodules, as compared to standard CT (SCT) MATERIAL AND METHODS: Ninety-three SCT were elaborated by dedicated software (ClearRead CT, Riverain Technologies, Miamisburg, OH, USA), that allows subtracting vessels from lung parenchyma. Semi-automated volumetric measurements of 65 solid nodules were compared between SCT and VSCT. The measurements were repeated by two readers. For each solid nodule, volume measured on SCT by Reader 1 and Reader 2 was averaged and the average volume between readers acted as standard of reference value. Concordance between measurements was assessed using Lin's Concordance Correlation Coefficient (CCC). Limits of agreement (LoA) between readers and CT datasets were evaluated. RESULTS: Standard of reference nodule volume ranged from 13 to 366 mm3. The mean overestimation between readers was 3 mm3 and 2.9 mm3 on SCT and VSCT, respectively. Semi-automated volumetric measurements on VSCT showed substantial agreement with the standard of reference (Lin's CCC = 0.990 for Reader 1; 0.985 for Reader 2). The upper and lower LoA between readers' measurements were (16.3, -22.4 mm3) and (15.5, -21.4 mm3) for SCT and VSCT, respectively. CONCLUSIONS: VSCT datasets are feasible for the measurements of solid nodules, showing an almost perfect concordance between readers and with measurements on SCT.

Modified Dual-Energy Algorithm for Calcified Plaque Removal: Evaluation in Carotid Computed Tomography Angiography and Comparison With Digital Subtraction Angiography.

OBJECTIVES: Computed tomography angiography (CTA) is a valuable tool for the assessment of carotid artery stenosis. However, blooming artifacts from calcified plaques might result in an overestimation of the stenosis grade. The aim of this study was to investigate a new dual-energy computed tomography (DECT) technique with a modified 3-material decomposition algorithm for calcium removal in extracranial carotid artery stenosis. MATERIALS AND METHODS: In this retrospective, institutional review board-approved study, 30 calcified carotid plaques in 22 patients (15 men; mean age, 73 ± 10 years) with clinical suspicion of stroke were included. Dual-energy computed tomography image data were obtained using second-generation dual-source CT with tube voltages at 80 and 140Sn kVp. Conventional CTA and virtual noncalcium (VNCa) images using the modified DECT algorithm were reconstructed. By assessing spectral characteristics, the modified DECT algorithm allows for a selective removal of calcium independent of blooming. Two independent and blinded readers evaluated subjective image quality, blooming artifacts, amount of (residual) calcification, and performed stenosis measurements according to the North American Symptomatic Carotid Endarterectomy Trial (NASCET) criteria. Differences were tested using a pairwise sign test. Paired sample t tests with Bonferroni correction (P < 0.017) and Bland-Altman analyses were used to test for differences in carotid stenosis measurements between VNCa and conventional CTA using digital subtraction angiography (DSA) as the standard of reference. RESULTS: Subjective image quality was similar among conventional CTA and VNCa image data sets (P = 0.82), whereas blooming artifacts were significantly reduced in VNCa images compared with conventional CTA (P < 0.001). Residual calcifications in VNCa images were absent in 11 (37%), minor in 12 (40%), medium sized in 2 (7%), and large in 5 (17%) arteries. Stenosis measurements differed significantly between VNCa (mean NASCET stenosis: 27% ± 20%) and conventional CTA images (mean NASCET stenosis: 39% ± 16%; P < 0.001) and between conventional CTA and DSA (23% ± 16%, P < 0.001). No significant differences in stenosis measurements were observed between VNCa and DSA (P = 0.189), with narrow limits of agreement (mean difference ±1.96 standard deviations: -4.7%, -35.1%, and 25.7%). CONCLUSIONS: A modified 3-material decomposition DECT algorithm for calcium removal was introduced, which allows for an accurate removal of calcified carotid plaques in extracranial carotid artery disease. The algorithm might overcome the problem of overestimation of calcified stenosis due to blooming artifacts in conventional CTA.