Dual-source photon-counting computed tomography for coronary in-stent observation: influence of heart rate and virtual monoenergetic image.

To investigate the effect of heart rate and virtual monoenergetic image (VMI) on coronary stent imaging in dual-source photon-counting detector computed tomography (PCD-CT). A dynamic cardiac phantom was used to vary the heart rate at 50 beats per minute (bpm), 70 bpm, and 90 bpm. Five types of stents (4.0 mm, 3.5 mm, 3.0 mm, 2.75 mm, and 2.5 mm diameter) were scanned at three different locations and reconstructed VMI at 70 keV. In addition, 50% stenosis was assessed for 3.0 mm and 4.0 mm stents. To assess in-stent stenosis, 40 keV, 70 keV, and 100 keV images were compared. Measurable lumen and contrast to noise ratio (CNR) from lumen to stenosis were evaluated quantitatively. A-4-point scale was used for the qualitative image quality of in-stent stenosis. The measurable lumen had no significant differences among heart rates in patent stents (p = 0.55). In-stent stenosis, the residual lumen was significantly larger in 40 keV [27.5% (20.8-32.3%)] than in 70 keV [11.5% (10.0-23.0%), p < 0.05] and 100 keV [0% (0-5.2%), p < 0.05]. The CNR was higher in 40 keV [12.5 (7.5-18.2)] than in 70 keV [5.3 (2.9-7.7), p < 0.05] and 100 keV [1.3 (0.5-2.7), p < 0.05]. The image quality was better in 40 keV (3.4 ± 0.7) than in 70 keV [(2.6 ± 0.8), p < 0.05] and 100 keV [(1.3 ± 0.4), p < 0.05]. Dual-source PCD-CT maintains a measurable lumen even at high heart rates. Adjusting the VMI can be helpful in visualizing the in-stent stenosis.

Photon-Counting Detector CT Radiological-Histological Correlation in Cadaveric Human Lung Nodules and Airways.

OBJECTIVES: The aim of this study was to compare the performances of photon-counting detector computed tomography (PCD-CT) and energy-integrating detector computed tomography (EID-CT) for visualizing nodules and airways in human cadaveric lungs. MATERIALS AND METHODS: Previously obtained 20 cadaveric lungs were scanned, and images were prospectively acquired by EID-CT and PCD-CT at a radiation dose with a noise level equivalent to the diagnostic reference level. PCD-CT was scanned with ultra-high-resolution mode. The EID-CT images were reconstructed with a 512 matrix, 0.6-mm thickness, and a 350-mm field of view (FOV). The PCD-CT images were reconstructed at 3 settings: PCD-512: same as EID-CT; PCD-1024-FOV350: 1024 matrix, 0.2-mm thickness, 350-mm FOV; and PCD-1024-FOV50: 1024 matrix, 0.2-mm thickness, 50-mm FOV. Two specimens per lung were examined after hematoxylin and eosin staining. The CT images were evaluated for nodules on a 5-point scale and for airways on a 4-point scale to compare the histology. The Wilcoxon signed rank test with Bonferroni correction was performed for statistical analyses. RESULTS: Sixty-seven nodules (1321 µm; interquartile range [IQR], 758-3105 µm) and 92 airways (851 µm; IQR, 514-1337 µm) were evaluated. For nodules and airways, scores decreased in order of PCD-1024-FOV50, PCD-1024-FOV350, PCD-512, and EID-CT. Significant differences were observed between series other than PCD-1024-FOV350 versus PCD-1024-FOV50 for nodules (PCD-1024-FOV350 vs PCD-1024-FOV50, P = 0.063; others P < 0.001) and between series other than EID-CT versus PCD-512 for airways (EID-CT vs PCD-512, P = 0.549; others P < 0.005). On PCD-1024-FOV50, the median size of barely detectable nodules was 604 µm (IQR, 469-756 µm) and that of barely detectable airways was 601 µm (IQR, 489-929 µm). On EID-CT, that of barely detectable nodules was 837 µm (IQR, 678-914 µm) and that of barely detectable airways was 1210 µm (IQR, 674-1435 µm). CONCLUSIONS: PCD-CT visualized small nodules and airways better than EID-CT and improved with high spatial resolution and potentially can detect submillimeter nodules and airways.

Three-dimensional iodine mapping quantified by dual-energy CT for predicting programmed death-ligand 1 expression in invasive pulmonary adenocarcinoma.

We examined the association between texture features using three-dimensional (3D) io-dine density histogram on delayed phase of dual-energy CT (DECT) and expression of programmed death-ligand 1 (PD-L1) using immunostaining methods in non-small cell lung cancer. Consecutive 37 patients were scanned by DECT. Unenhanced and enhanced (3 min delay) images were obtained. 3D texture analysis was performed for each nodule to obtain 7 features (max, min, median, mean, standard deviation, skewness, and kurtosis) from iodine density mapping and extracellular volume (ECV). A pathologist evaluated a tumor proportion score (TPS, %) using PD-L1 immunostaining: PD-L1 high (TPS ≥ 50%) and low or negative expression (TPS < 50%). Associations between PD-L1 expression and each 8 parameter were evaluated using logistic regression analysis. The multivariate logistic regression analysis revealed that skewness and ECV were independent indicators associated with high PD-L1 expression (skewness: odds ratio [OR] 7.1 [95% CI 1.1, 45.6], p = 0.039; ECV: OR 6.6 [95% CI 1.1, 38.4], p = 0.037). In the receiver-operating characteristic analysis, the area under the curve of the combination of skewness and ECV was 0.83 (95% CI 0.67, 0.93) with sensitivity of 64% and specificity of 96%. Skewness from 3D iodine density histogram and ECV on dual energy CT were significant factors for predicting PD-L1 expression.

Association between interstitial lung abnormality and mortality in patients with esophageal cancer.

PURPOSE: To investigate the relationship between interstitial lung abnormalities (ILAs) and mortality in patients with esophageal cancer and the cause of mortality. MATERIALS AND METHODS: This retrospective study investigated patients with esophageal cancer from January 2011 to December 2015. ILAs were visually scored on baseline CT using a 3-point scale (0 = non-ILA, 1 = indeterminate for ILA, and 2 = ILA). ILAs were classified into subcategories of non-subpleural, subpleural non-fibrotic, and subpleural fibrotic. Five-year overall survival (OS) was compared between patients with and without ILAs using the multivariable Cox proportional hazards model. Subgroup analyses were performed based on cancer stage and ILA subcategories. The prevalences of treatment complications and death due to esophageal cancer and pneumonia/respiratory failure were analyzed using Fisher's exact test. RESULTS: A total of 478 patients with esophageal cancer (age, 66.8 years ± 8.6 [standard deviation]; 64 women) were evaluated in this study. Among them, 267 patients showed no ILAs, 125 patients were indeterminate for ILAs, and 86 patients showed ILAs. ILAs were a significant factor for shorter OS (hazard ratio [HR] = 1.68, 95% confidence interval [CI] 1.10-2.55, P = 0.016) in the multivariable Cox proportional hazards model adjusting for age, sex, smoking history, clinical stage, and histology. On subgroup analysis using patients with clinical stage IVB, the presence of ILAs was a significant factor (HR = 3.78, 95% CI 1.67-8.54, P = 0.001). Subpleural fibrotic ILAs were significantly associated with shorter OS (HR = 2.22, 95% CI 1.25-3.93, P = 0.006). There was no significant difference in treatment complications. Patients with ILAs showed a higher prevalence of death due to pneumonia/respiratory failure than those without ILAs (non-ILA, 2/95 [2%]; ILA, 5/39 [13%]; P = 0.022). The prevalence of death due to esophageal cancer was similar in patients with and without ILA (non-ILA, 82/95 [86%]; ILA 32/39 [82%]; P = 0.596). CONCLUSION: ILAs were significantly associated with shorter survival in patients with esophageal cancer.

Prediction of solid and micropapillary components in lung invasive adenocarcinoma: radiomics analysis from high-spatial-resolution CT data with 1024 matrix.

PURPOSE: To predict solid and micropapillary components in lung invasive adenocarcinoma using radiomic analyses based on high-spatial-resolution CT (HSR-CT). MATERIALS AND METHODS: For this retrospective study, 64 patients with lung invasive adenocarcinoma were enrolled. All patients were scanned by HSR-CT with 1024 matrix. A pathologist evaluated subtypes (lepidic, acinar, solid, micropapillary, or others). Total 61 radiomic features in the CT images were calculated using our modified texture analysis software, then filtered and minimized by least absolute shrinkage and selection operator (LASSO) regression to select optimal radiomic features for predicting solid and micropapillary components in lung invasive adenocarcinoma. Final data were obtained by repeating tenfold cross-validation 10 times. Two independent radiologists visually predicted solid or micropapillary components on each image of the 64 nodules with and without using the radiomics results. The quantitative values were analyzed with logistic regression models. The receiver operating characteristic curves were generated to predict of solid and micropapillary components. P values < 0.05 were considered significant. RESULTS: Two features (Coefficient Variation and Entropy) were independent indicators associated with solid and micropapillary components (odds ratio, 30.5 and 11.4; 95% confidence interval, 5.1-180.5 and 1.9-66.6; and P = 0.0002 and 0.0071, respectively). The area under the curve for predicting solid and micropapillary components was 0.902 (95% confidence interval, 0.802 to 0.962). The radiomics results significantly improved the accuracy and specificity of the prediction of the two radiologists. CONCLUSION: Two texture features (Coefficient Variation and Entropy) were significant indicators to predict solid and micropapillary components in lung invasive adenocarcinoma.

Usefulness of Three-Dimensional Iodine Mapping Quantified by Dual-Energy CT for Differentiating Thymic Epithelial Tumors.

Background: Dual-energy CT has been reported to be useful for differentiating thymic epithelial tumors. The purpose is to evaluate thymic epithelial tumors by using three-dimensional (3D) iodine density histogram texture analysis on dual-energy CT and to investigate the association of extracellular volume fraction (ECV) with the fibrosis of thymic carcinoma. Methods: 42 patients with low-risk thymoma (n = 20), high-risk thymoma (n = 16), and thymic carcinoma (n = 6) were scanned by dual-energy CT. 3D iodine density histogram texture analysis was performed for each nodule on iodine density mapping: Seven texture features (max, min, median, average, standard deviation [SD], skewness, and kurtosis) were obtained. The iodine effect (average on DECT180s-average on unenhanced DECT) and ECV on DECT180s were measured. Tissue fibrosis was subjectively rated by one pathologist on a three-point grade. These quantitative data obtained by examining associations with thymic carcinoma and high-risk thymoma were analyzed with univariate and multivariate logistic regression models (LRMs). The area under the curve (AUC) was calculated by the receiver operating characteristic curves. p values < 0.05 were significant. Results: The multivariate LRM showed that ECV > 21.47% in DECT180s could predict thymic carcinoma (odds ratio [OR], 11.4; 95% confidence interval [CI], 1.18-109; p = 0.035). Diagnostic performance was as follows: Sensitivity, 83.3%; specificity, 69.4%; AUC, 0.76. In high-risk thymoma vs. low-risk thymoma, the multivariate LRM showed that the iodine effect ≤1.31 mg/cc could predict high-risk thymoma (OR, 7; 95% CI, 1.02-39.1; p = 0.027). Diagnostic performance was as follows: Sensitivity, 87.5%; specificity, 50%; AUC, 0.69. Tissue fibrosis significantly correlated with thymic carcinoma (p = 0.026). Conclusions: ECV on DECT180s related to fibrosis may predict thymic carcinoma from thymic epithelial tumors, and the iodine effect on DECT180s may predict high-risk thymoma from thymoma.

Anatomy-aware self-supervised learning for anomaly detection in chest radiographs.

In this study, we present a self-supervised learning (SSL)-based model that enables anatomical structure-based unsupervised anomaly detection (UAD). The model employs an anatomy-aware pasting (AnatPaste) augmentation tool that uses a threshold-based lung segmentation pretext task to create anomalies in normal chest radiographs used for model pretraining. These anomalies are similar to real anomalies and help the model recognize them. We evaluate our model using three open-source chest radiograph datasets. Our model exhibits area under curves of 92.1%, 78.7%, and 81.9%, which are the highest among those of existing UAD models. To the best of our knowledge, this is the first SSL model to employ anatomical information from segmentation as a pretext task. The performance of AnatPaste shows that incorporating anatomical information into SSL can effectively improve accuracy.

BERT-based Transfer Learning in Sentence-level Anatomic Classification of Free-Text Radiology Reports.

Purpose: To assess whether transfer learning with a bidirectional encoder representations from transformers (BERT) model, pretrained on a clinical corpus, can perform sentence-level anatomic classification of free-text radiology reports, even for anatomic classes with few positive examples. Materials and Methods: This retrospective study included radiology reports of patients who underwent whole-body PET/CT imaging from December 2005 to December 2020. Each sentence in these reports (6272 sentences) was labeled by two annotators according to body part ("brain," "head & neck," "chest," "abdomen," "limbs," "spine," or "others"). The BERT-based transfer learning approach was compared with two baseline machine learning approaches: bidirectional long short-term memory (BiLSTM) and the count-based method. Area under the precision-recall curve (AUPRC) and area under the receiver operating characteristic curve (AUC) were computed for each approach, and AUCs were compared using the DeLong test. Results: The BERT-based approach achieved a macro-averaged AUPRC of 0.88 for classification, outperforming the baselines. AUC results for BERT were significantly higher than those of BiLSTM for all classes and those of the count-based method for the "brain," "chest," "abdomen," and "others" classes (P values < .025). AUPRC results for BERT were superior to those of baselines even for classes with few labeled training data (brain: BERT, 0.95, BiLSTM, 0.11, count based, 0.41; limbs: BERT, 0.74, BiLSTM, 0.28, count based, 0.46; spine: BERT, 0.82, BiLSTM, 0.53, count based, 0.69). Conclusion: The BERT-based transfer learning approach outperformed the BiLSTM and count-based approaches in sentence-level anatomic classification of free-text radiology reports, even for anatomic classes with few labeled training data.Keywords: Anatomy, Comparative Studies, Technology Assessment, Transfer Learning Supplemental material is available for this article. © RSNA, 2023.

Radiologists with and without deep learning-based computer-aided diagnosis: comparison of performance and interobserver agreement for characterizing and diagnosing pulmonary nodules/masses.

OBJECTIVES: To compare the performance of radiologists in characterizing and diagnosing pulmonary nodules/masses with and without deep learning (DL)-based computer-aided diagnosis (CAD). METHODS: We studied a total of 101 nodules/masses detected on CT performed between January and March 2018 at Osaka University Hospital (malignancy: 55 cases). SYNAPSE SAI Viewer V1.4 was used to analyze the nodules/masses. In total, 15 independent radiologists were grouped (n = 5 each) according to their experience: L (< 3 years), M (3-5 years), and H (> 5 years). The likelihoods of 15 characteristics, such as cavitation and calcification, and the diagnosis (malignancy) were evaluated by each radiologist with and without CAD, and the assessment time was recorded. The AUCs compared with the reference standard set by two board-certified chest radiologists were analyzed following the multi-reader multi-case method. Furthermore, interobserver agreement was compared using intraclass correlation coefficients (ICCs). RESULTS: The AUCs for ill-defined boundary, irregular margin, irregular shape, calcification, pleural contact, and malignancy in all 15 radiologists, irregular margin and irregular shape in L and ill-defined boundary and irregular margin in M improved significantly (p < 0.05); no significant improvements were found in H. L showed the greatest increase in the AUC for malignancy (not significant). The ICCs improved in all groups and for nearly all items. The median assessment time was not prolonged by CAD. CONCLUSIONS: DL-based CAD helps radiologists, particularly those with < 5 years of experience, to accurately characterize and diagnose pulmonary nodules/masses, and improves the reproducibility of findings among radiologists. KEY POINTS: • Deep learning-based computer-aided diagnosis improves the accuracy of characterizing nodules/masses and diagnosing malignancy, particularly by radiologists with < 5 years of experience. • Computer-aided diagnosis increases not only the accuracy but also the reproducibility of the findings across radiologists.

Ultrafast cervcial spine MRI protocol using deep learning-based reconstruction: Diagnostic equivalence to a conventional protocol.

PURPOSE: A major drawback of magnetic resonance imaging (MRI) is its limited imaging speed. This study proposed an ultrafast cervical spine MRI protocol (2 min 57 s) using deep learning-based reconstruction (DLR) and compared the diagnostic results to those of conventional MRI protocols (12 min 54 s). METHODS: Fifty patients who underwent cervical spine MRI using both conventional and ultrafast protocols, including sagittal T1-weighted, T2-weighted, short-TI inversion recovery, and axial T2*-weighted imaging were included in this study. The ultrafast protocol shortened the acquisition time to approximately-one-fourth of that of the conventional protocol by reducing the phase matrix, oversampling rate, and number of excitations, and by applying compressed sensing. To compensate for the decreased signal-to-noise ratio caused by acceleration, noise reduction using DLR was performed. For image interpretation, three neuroradiologists graded or classified degenerative changes, including central canal stenosis, foraminal stenosis, endplate degeneration, disc degeneration, and disc hernia. The presence of other pathologies was also recorded. Given the absence of a reference standard, we tested the interchangeability of the two protocols by calculating the 95% confidence interval (CI) of the individual equivalence index. We also assessed the inter-protocol intra-reader agreement using kappa statistics. RESULTS: Except for endplate degeneration, the 95 % CI of the individual equivalence index for all variables did not exceed 5 %, indicating interchangeability between the two protocols. The kappa values ranged from 0.600 to 0.977, indicating substantial to almost perfect agreement. CONCLUSIONS: The proposed ultrafast MRI protocol yielded almost equivalent diagnostic results compared as the conventional protocol.

The evaluation of the reduction of radiation dose via deep learning-based reconstruction for cadaveric human lung CT images.

To compare the quality of CT images of the lung reconstructed using deep learning-based reconstruction (True Fidelity Image: TFI ™; GE Healthcare) to filtered back projection (FBP), and to determine the minimum tube current-time product in TFI without compromising image quality. Four cadaveric human lungs were scanned on CT at 120 kVp and different tube current-time products (10, 25, 50, 75, 100, and 175 mAs) and reconstructed with TFI and FBP. Two image evaluations were performed by three independent radiologists. In the first experiment, using the same tube current-time product, a side-by-side TFI and FBP comparison was performed. Images were evaluated with regard to noise, streak artifacts, and overall image quality. Overall image quality was evaluated in view of whole image quality. In the second experiment, CT images reconstructed using TFI and FBP with five different tube current-time products were displayed in random order, which were evaluated with reference to the 175 mAs-FBP image. Images were scored with regard to normal structure, abnormal findings, noise, streak artifacts, and overall image quality. Median scores from three radiologists were statistically analyzed. Quantitative evaluation of noise was performed by setting regions of interest (ROIs) in air. In first experiment, overall image quality was improved, and noise was decreased in images of TFI compared to that of FBP for all tube current-time products. In second experiment, scores of all evaluation items except for small vessels in images of 25 mAs-TFI were almost the same as that of 175 mAs-FBP (all p > 0.31). Using TFI instead of FBP, at least 85% radiation dose reduction could be possible without any degradation in the image quality.

CT-based radiomics analysis for differentiation between thymoma and thymic carcinoma.

Background: The purpose of our study was to differentiate between thymoma and thymic carcinoma using a radiomics analysis based on the computed tomography (CT) image features. Methods: The CT images of 61 patients with thymic epithelial tumors (TETs) who underwent contrast-enhanced CT with slice thickness <1 mm were analyzed. Pathological examination of the surgical specimens revealed thymoma in 45 and thymic carcinoma in 16. Tumor volume and the ratio of major axis to minor axis were calculated using a computer-aided diagnostic software. Sixty-one different radiomics features in the segmented CT images were extracted, then filtered and minimized by least absolute shrinkage and selection operator (LASSO) regression to select the optimal radiomics features for predicting thymic carcinoma. The association between the quantitative values and a diagnosis of thymic carcinoma were analyzed with logistic regression models. Parameters identified as significant in univariate analysis were included in multiple analyses. Receiver-operating characteristic (ROC) curves were assessed to evaluate the diagnostic performance. Results: Thymic carcinoma was significantly predominant in men (P=0.001). Optimal radiomics features for predicting thymic carcinoma were as follows: gray-level co-occurrence matrix (GLCM)-homogeneity, GLCM-energy, compactness, large zone high gray-level emphasis (LZHGE), solidity, size of minor axis, and kurtosis. Multiple logistic regression analysis of these features revealed solidity and GLCM-energy as independent indicators associated with thymic carcinoma [odds ratio, 14.7 and 14.3; 95% confidence interval (CI): 1.6-139.0 and 3.0-68.7; and P=0.045 and 0.002, respectively]. Area under the curve (AUC) for diagnosing thymic carcinoma was 0.882 (sensitivity, 81.2%; specificity, 91.1%). Multivariate analysis adjusted for sex similarly revealed two features (solidity and GLCM-energy) as independent indicators associated with thymic carcinoma (odds ratio, 14.6 and 23.9; 95% CI: 2.4-89.2 and 1.9-302.8; P=0.004 and 0.014, respectively). Adjusted AUC for diagnosing thymic carcinoma was 0.921 (95% CI: 0.82-0.97): sensitivity, 62.5% and specificity, 100%. Conclusions: Two texture features (GLCM-energy and solidity) were significant predictors of thymic carcinoma.

Quantitative volumetry of ground-glass nodules on high-spatial-resolution CT with 0.25-mm section thickness and 1024 matrix: Phantom and clinical studies.

OBJECTIVES: To compare high-resolution (HR) and conventional (C) settings of high-spatial-resolution computed tomography (CT) for software volumetry of ground-glass nodules (GGNs) in phantoms and patients. METHODS: We placed -800 and -630 HU spherical GGN-mimic nodules in 28 different positions in phantoms and scanned them individually. Additionally, 60 GGNs in 45 patients were assessed retrospectively. Images were reconstructed using the HR-setting (matrix size, 1024; slice thickness, 0.25 mm) and C-setting (matrix size, 512; slice thickness, 0.5 mm). We measured the GGN volume and mass using software. In the phantom study, the absolute percentage error (APE) was calculated as the absolute difference between Vernier caliper measurement-based and software-based volumes. In patients, we measured the density (mean, maximum, and minimum) and classified GGNs into low- and high-attenuation GGNs. RESULTS: In images of the -800 HU, but not -630 HU, phantom nodules, the volumes and masses differed significantly between the two settings (both p < 0.01). The APE was significantly lower in the HR-setting than in the C-setting (p < 0.01). In patients, volumes did not differ significantly between settings (p = 0.59). Although the mean attenuation was not significantly different, the maximum and minimum values were significantly increased and decreased, respectively, in the HR-setting (both p < 0.01). The volumes of both low-attenuation and high-attenuation GGNs were not significantly different between settings (p = 0.78 and 0.39, respectively). CONCLUSION: The HR-setting might yield a more accurate volume for phantom GGN of -800 HU and influence the detection of maximum and minimum CT attenuation.

Deep learning algorithm for detection of aortic dissection on non-contrast-enhanced CT.

OBJECTIVES: To develop a deep learning-based algorithm to detect aortic dissection (AD) and evaluate the diagnostic ability of the algorithm compared with those of radiologists. METHODS: Included in the study were 170 patients (85 with AD and 85 without AD). An AD detection algorithm was developed using a convolutional neural network with Xception architecture. Of the patient data, 80% were used for training and validation and 20% were used for testing. Fivefold cross-validation was performed to evaluate the method. An average of 6688 non-contrast-enhanced CT images (slice thickness, 5 mm) were used for training. A radiologist reviewed both contrast-enhanced and non-contrast-enhanced images and identified the slices of AD. The identified slices were used as ground truth. Receiver operating characteristic curve and area under the curve (AUC) analysis was performed. Five radiologists independently evaluated the images. The accuracy, sensitivity, and specificity of the algorithm and those of the radiologists were compared. RESULTS: The AUC of the developed algorithm was 0.940, and a cutoff value of 0.400 provided accuracy of 90.0%, sensitivity of 91.8%, and specificity of 88.2%. For the radiologists, median (range) accuracy, sensitivity, and specificity were 88.8 (83.5-94.1)%, 90.6 (83.5-94.1)%, and 94.1 (72.9-97.6)%, respectively. There was no significant difference in performance in terms of accuracy, sensitivity, or specificity between the algorithm and the average performance of the radiologists (p > 0.05). CONCLUSIONS: The developed algorithm showed comparable diagnostic performance to radiologists for detecting AD, which suggests the potential of the proposed method to support clinical practice by reducing missed ADs. KEY POINTS: • A deep learning-based algorithm for detecting aortic dissection was developed using the non-contrast-enhanced CT images of 170 patients. • The algorithm had an AUC of 0.940 for detecting aortic dissection. • The accuracy, sensitivity, and specificity of the algorithm were comparable to those of radiologists.