In vivo characterization of urinary calculi on dual-energy CT: going a step ahead with sub-differentiation of calcium stones.

BACKGROUND: The role of dual-energy computed tomography (DECT) in characterization of urinary calculi is evolving and literature regarding differentiation of calcium calculi is sparse and confounding. PURPOSE: To evaluate the capability of DECT in assessing the urinary calculi composition in vivo, especially in differentiating various types of calcium calculi. MATERIAL AND METHODS: One hundred and twenty patients underwent DECT for characterization of urinary calculi. Seventy patients with 114 calculi, including 93 calcium stones, were retrospectively analyzed. DE ratios and attenuation differences were compared using ANOVA and receiver-operating-characteristic (ROC) analysis was done to predict cut-off values, in particular for detecting calcium-oxalate-monohydrate (COM) stones. RESULTS: DE ratio ≤1.14 accurately detected uric acid calculi, ≥1.29 was definitive for calcium and intermediate values were characteristic of cystine stones. DE ratios were significantly different between group 1 (COM [n = 32]; mean 1.376 ± 0.041), group 2 ([calcium oxalate dihydrate (COD) + COM] [n = 51]; 1.416 ± 0.048), and group 3 ([carbonate apatite (CaP) + COD + COM] [n = 10]; 1.468 ± 0.038) (group 1 vs. 2, P = 0.001; 1 vs. 3, P = 0.000; 2 vs. 3, P = 0.004). More importantly, pure COM calculi (group 1) had significantly lower DE ratio compared with mixed calcium calculi (groups 2 and 3) (P = 0.000). Attenuation differences (between low and high kV images) could not distinguish between COM and mixed calculi. ROC analysis for detection of COM calculi yielded AUC of 0.770 with cut-off DE ratio 1.385 (sensitivity 65.6%, specificity 82%) and value <1.335 was seen only with COM calculi (100% specificity). CONCLUSION: DECT can be employed for in vivo differentiation of various types of calculi and for detection of relatively lithotripsy-resistant COM calculi.

Acute prevertebral calcific tendinitis: a source of non-surgical acute cervical pain.

BACKGROUND: Prevertebral calcific tendinitis results from calcium hydroxyapatite crystal deposits in the longus colli muscles, which induce symptoms similar to some surgically-treated conditions, such as retropharyngeal abscesses. Imaging techniques are critical for accurate diagnosis. PURPOSE: To describe the computed tomography (CT) findings associated with prevertebral calcific tendinitis. MATERIAL AND METHODS: Retrospective analysis performed in an 18-month period, searching for patients with neck CT and reports with diagnosis of "calcific longus collis tendinitis" or "prevertebral calcific tendinitis". CT images and clinical data available in the medical records were analyzed. RESULTS: One hundred and thirty-four examinations were performed in the period studied. Nine patients who fulfilled inclusion criteria were identified and their CT imaging characteristics are presented. Six presented with calcific deposits in the right longus colli muscle. CT matched the clinical pain lateralization in all cases. Eight patients had no significant enhancement post injection of contrast media. CONCLUSION: Prevertebral calcific tendinitis is a cause of acute cervical pain that clinically mimics a retropharyngeal abscess, however on neck CT has a characteristic appearance. Correct identification of this pathologic condition will help avoiding unnecessary invasive procedures.

Prediction of Coronary Artery Calcium and Cardiovascular Risk on Chest Radiographs Using Deep Learning.

PURPOSE: To assess the ability of deep convolutional neural networks (DCNNs) to predict coronary artery calcium (CAC) and cardiovascular risk on chest radiographs. MATERIALS AND METHODS: In this retrospective study, 1689 radiographs in patients who underwent cardiac CT and chest radiography within the same year, between 2013 and 2018, were included (mean age, 56 years ± 11 [standard deviation]; 969 radiographs in women). Agatston scores were used as ground truth labels for DCNN training on radiographs. DCNNs were trained for binary classification of (a) nonzero or zero total calcium scores, (b) presence or absence of calcium in each coronary artery, and (c) total calcium scores above or below varying thresholds. Results from classification of test images were compared with established 10-year atherosclerotic cardiovascular disease (ASCVD) risk scores in each cohort. Classifier performance was measured using area under the receiver operating characteristic curve (AUC) with attention maps to highlight areas of decision-making. RESULTS: Binary classification between zero and nonzero total calcium scores reached an AUC of 0.73 on frontal radiographs, with similar performance on laterals (AUC, 0.70; P = .56). Performance was similar for binary classification of absolute total calcium score above or below 100 (AUC, 0.74). Frontal radiographs that tested positive for a predicted nonzero CAC score correlated with a higher 10-year ASCVD risk of 17.2% ± 10.9 compared with 11.9% ± 10.2 for a negative test, indicating predicted CAC score of zero (P < .001). Multivariate logistic regression demonstrated the algorithm could predict a nonzero calcium score independent of traditional cardiovascular risk factors. Performance was reduced for individual coronary arteries. Heat maps primarily localized to the cardiac silhouette and occasionally other cardiovascular findings. CONCLUSION: DCNNs trained on chest radiographs had modest accuracy for predicting the presence of CAC correlating with cardiovascular risk.Keywords: Coronary Arteries, Cardiac, Calcifications/Calculi, Neural NetworksSee also the commentary by Gupta and Blankstein in this issue.©RSNA, 2021.

Automated Segmentation and Volume Measurement of Intracranial Internal Carotid Artery Calcification at Noncontrast CT.

PURPOSE: To develop and evaluate a fully-automated deep learning-based method for assessment of intracranial internal carotid artery calcification (ICAC). MATERIALS AND METHODS: This was a secondary analysis of prospectively collected data from the Rotterdam study (2003-2006) to develop and validate a deep learning-based method for automated ICAC delineation and volume measurement. Two observers manually delineated ICAC on noncontrast CT scans of 2319 participants (mean age, 69 years ± 7 [standard deviation]; 1154 women [53.2%]), and a deep learning model was trained to segment ICAC and quantify its volume. Model performance was assessed by comparing manual and automated segmentations and volume measurements to those produced by an independent observer (available on 47 scans), comparing the segmentation accuracy in a blinded qualitative visual comparison by an expert observer, and comparing the association with first stroke incidence from the scan date until 2016. All method performance metrics were computed using 10-fold cross-validation. RESULTS: The automated delineation of ICAC reached a sensitivity of 83.8% and positive predictive value (PPV) of 88%. The intraclass correlation between automatic and manual ICAC volume measures was 0.98 (95% CI: 0.97, 0.98; computed in the entire dataset). Measured between the assessments of independent observers, sensitivity was 73.9%, PPV was 89.5%, and intraclass correlation coefficient was 0.91 (95% CI: 0.84, 0.95; computed in the 47-scan subset). In the blinded visual comparisons of 294 regions, automated delineations were judged as more accurate than manual delineations in 131 regions, less accurate in 94 regions, and equally accurate in the rest of the regions (131 of 225, 58.2%; P = .01). The association of ICAC volume with incident stroke was similarly strong for both automated (hazard ratio, 1.38 [95% CI: 1.12, 1.75]) and manually measured volumes (hazard ratio, 1.48 [95% CI: 1.20, 1.87]). CONCLUSION: The developed model was capable of automated segmentation and volume quantification of ICAC with accuracy comparable to human experts.Keywords CT, Neural Networks, Carotid Arteries, Calcifications/Calculi, Arteriosclerosis, Segmentation, Vision Application Domain, Stroke Supplemental material is available for this article. © RSNA, 2021.

Model-based iterative reconstruction and adaptive statistical iterative reconstruction: dose-reduced CT for detecting pancreatic calcification.

BACKGROUND: Iterative reconstruction methods have attracted attention for reducing radiation doses in computed tomography (CT). PURPOSE: To investigate the detectability of pancreatic calcification using dose-reduced CT reconstructed with model-based iterative construction (MBIR) and adaptive statistical iterative reconstruction (ASIR). MATERIAL AND METHODS: This prospective study approved by Institutional Review Board included 85 patients (57 men, 28 women; mean age, 69.9 years; mean body weight, 61.2 kg). Unenhanced CT was performed three times with different radiation doses (reference-dose CT [RDCT], low-dose CT [LDCT], ultralow-dose CT [ULDCT]). From RDCT, LDCT, and ULDCT, images were reconstructed with filtered-back projection (R-FBP, used for establishing reference standard), ASIR (L-ASIR), and MBIR and ASIR (UL-MBIR and UL-ASIR), respectively. A lesion (pancreatic calcification) detection test was performed by two blinded radiologists with a five-point certainty level scale. RESULTS: Dose-length products of RDCT, LDCT, and ULDCT were 410, 97, and 36 mGy-cm, respectively. Nine patients had pancreatic calcification. The sensitivity for detecting pancreatic calcification with UL-MBIR was high (0.67-0.89) compared to L-ASIR or UL-ASIR (0.11-0.44), and a significant difference was seen between UL-MBIR and UL-ASIR for one reader (P = 0.014). The area under the receiver-operating characteristic curve for UL-MBIR (0.818-0.860) was comparable to that for L-ASIR (0.696-0.844). The specificity was lower with UL-MBIR (0.79-0.92) than with L-ASIR or UL-ASIR (0.96-0.99), and a significant difference was seen for one reader (P < 0.01). CONCLUSION: In UL-MBIR, pancreatic calcification can be detected with high sensitivity, however, we should pay attention to the slightly lower specificity.