CT texture analysis reliability in pulmonary lesions: the influence of 3D vs. 2D lesion segmentation and volume definition by a Hounsfield-unit threshold.

OBJECTIVE: Reproducibility problems are a known limitation of radiomics. The segmentation of the target lesion plays a critical role in texture analysis variability. This study's aim was to compare the interobserver reliability of manual 2D vs. 3D lung lesion segmentation with and without pre-definition of the volume using a threshold of - 50 HU. METHODS: Seventy-five patients with histopathologically proven lung lesions (15 patients each with adenocarcinoma, squamous cell carcinoma, small cell lung cancer, carcinoid, and organizing pneumonia) who underwent an unenhanced CT scan of the chest were included. Three radiologists independently segmented each lesion manually in 3D and 2D with and without pre-segmentation volume definition by a HU threshold, and shape parameters and original, Laplacian of Gaussian-filtered, and wavelet-based texture features were derived. To assess interobserver reliability and identify the most robust texture features, intraclass correlation coefficients (ICCs) for different segmentation settings were calculated. RESULTS: Shape parameters had high reliability (64-79% had excellent and good ICCs). Texture features had weak reliability levels, with the highest ICCs (38% excellent or good) found for original features in 3D segmentation without the use of a HU threshold. A small proportion (4.3-11.5%) of texture features had excellent or good ICC values at all segmentation settings. CONCLUSION: Interobserver reliability of texture features from CT scans of a heterogeneous collection of manually segmented lung lesions was low with a small proportion of features demonstrating high reliability independent of the segmentation settings. These results indicate a limited applicability of texture analysis and the need to define robust texture features in patients with lung lesions. KEY POINTS: • Our study showed a low reproducibility of texture features when 3 radiologists independently segmented lung lesions in CT images, which highlights a serious limitation of texture analysis. • Interobserver reliability of texture features was low regardless of whether the lesion was segmented in 2D and 3D with or without a HU threshold. • In contrast to texture features, shape parameters showed a high interobserver reliability when lesions were segmented in 2D vs. 3D with and without a HU threshold of - 50.

Perfusion Computed Tomography in Rectal Carcinoma: Influence of Optimization of the Patlak Range on Calculation of Equivalent Blood Volume and Flow Extraction.

PURPOSE: The aim of the study is to assess the influence of manual adjustment of the Patlak range in computed tomography (CT) perfusion analysis of rectal carcinoma compared with default range of the perfusion software. METHODS: This study was approved by the institutional review board and informed consent was obtained. Twenty-one patients (12 male, 9 female; mean age ± SD, 59 ± 11 years) with rectal cancer were included and underwent perfusion CT before preoperative chemoradiotherapy. Equivalent blood volume (BV) and flow-extraction (FE) were calculated using the Patlak plot model. Two perfusion sets were calculated per patient, a perfusion set using the default setting as provided by the software (dBV, dFE) and an optimized perfusion set after manual adaption of the Patlak range (aBV, aFE), which was limited to the intravascular space clearance of contrast to the extravascular space. Perfusion values calculated with both methods were compared for significance in differences using the Wilcoxon test. A P value of 0.05 or less was defined as statistically significant. RESULTS: Adjustment of the Patlak range statistically significantly influenced BV and FE calculation. Median dBV was 23.2 mL/100 mL (interquartile range [IQR], 12.1 mL/100 mL), whereas median aBV was 20.3 mL/100 mL (IQR, 10.9 mL/100 mL). The difference in BV was statistically significant ( P = 0.021). Median dFE was 8.3 mL/min/100 mL (IQR, 4.7 mL/min/100 mL), whereas median aFE was 15.4 mL/min/100 mL (IQR, 5.8 mL/min/100 mL). The difference in FE was statistically significant ( P < 0.001). CONCLUSIONS: Our findings indicate that in perfusion CT of rectal carcinoma, adjustment of the Patlak range may significantly influence BV and FE compared with default setting of the software. This may contribute to standardization in the use of this technique for functional imaging of rectal cancer.

Type B Aortic Dissection CTA Collection with True and False Lumen Expert Annotations for the Development of AI-based Algorithms.

Aortic dissections (ADs) are serious conditions of the main artery of the human body, where a tear in the inner layer of the aortic wall leads to the formation of a new blood flow channel, named false lumen. ADs affecting the aorta distally to the left subclavian artery are classified as a Stanford type B aortic dissection (type B AD). This is linked to substantial morbidity and mortality, however, the course of the disease for the individual case is often unpredictable. Computed tomography angiography (CTA) is the gold standard for the diagnosis of type B AD. To advance the tools available for the analysis of CTA scans, we provide a CTA collection of 40 type B AD cases from clinical routine with corresponding expert segmentations of the true and false lumina. Segmented CTA scans might aid clinicians in decision making, especially if it is possible to fully automate the process. Therefore, the data collection is meant to be used to develop, train and test algorithms.

Association of covert brain infarct phenotype with stroke recurrence in first-ever manifest ischemic stroke according to etiology.

INTRODUCTION: Covert brain infarcts (CBI) are frequent incidental findings on MRI and associated with future stroke risk in patients without a history of clinically evident cerebrovascular events. However, the prognostic value of CBI in first-ever ischemic stroke patients is unclear and previous studies did not report on different etiological stroke subtypes. We aimed to test CBI phenotypes and their association with stroke recurrence in first-ever ischemic stroke patients according to stroke etiology. PATIENTS AND METHODS: This study is a pooled data analysis of two prospectively collected cohorts of consecutive first-ever ischemic stroke patients admitted to the comprehensive stroke centers of Bern (Switzerland) and Graz (Austria). CBI phenotypes were identified on brain MRI within 72 h after admission. All patients underwent a routine follow-up (median: 12 months) to identify stroke recurrence. RESULTS: Of 1577 consecutive ischemic stroke patients (median age: 71 years), 691 patients showed CBI on brain MRI (44%) and 88 patients had a recurrent ischemic stroke (6%). Baseline CBI were associated with stroke recurrence in multivariable analysis (HR 1.9, 95% CI 1.1-3.3). CBI phenotypes with the highest risk for stroke recurrence were cavitatory CBI in small vessel disease (SVD)-related stroke (HR 7.1, 95% CI 1.6-12.6) and cortical CBI in patients with atrial fibrillation (HR 3.0, 95% CI 1.1-8.1). DISCUSSION AND CONCLUSION: This study reports a ≈ 2-fold increased risk for stroke recurrence in first-ever ischemic stroke patients with CBI. The risk of recurrent stroke was highest in patients with cavitatory CBI in SVD-related stroke and cortical CBI in patients with atrial fibrillation.Subject terms: Covert brain infarcts, stroke.

Three dimensional computed tomography texture analysis of pulmonary lesions: Does radiomics allow differentiation between carcinoma, neuroendocrine tumor and organizing pneumonia?

PURPOSE: To investigate whether CT texture analysis allows differentiation between adenocarcinomas, squamous cell carcinomas, carcinoids, small cell lung cancers and organizing pneumonia and between carcinomas and neuroendocrine tumors. METHOD: This retrospective study included patients 133 patients (30 patients with organizing pneumonia, 30 patients with adenocarcinoma, 30 patients with squamous cell carcinoma, 23 patients with small cell lung cancer, 20 patients with carcinoid), who underwent CT-guided biopsy of the lung and had a corresponding histopathologic diagnosis. Pulmonary lesions were segmented in consensus by two radiologists with and without a threshold of -50HU in three dimensions. Groupwise comparisons were performed to assess for differences between all five above-listed entities and between carcinomas and neuroendocrine tumors. RESULTS: Pairwise comparisons of the five entities revealed 53 statistically significant texture features when using no HU-threshold and 6 statistically significant features with a threshold of -50HU. The largest AUC (0.818 [95%CI 0.706-0.930]) was found for the feature wavelet-HHH_glszm_SmallAreaEmphasis for discrimination of carcinoid from the other entities when using no HU-threshold. In differentiating neuroendocrine tumors from carcinomas, 173 parameters proved statistically significant when using no HU threshold versus 52 parameters when using a -50HU-threshold. The largest AUC (0.810 [95%CI 0.728-0,893]) was found for the parameter original_glcm_Correlation for discrimination of neuroendocrine tumors from carcinomas when using no HU-threshold. CONCLUSIONS: CT texture analysis revealed features that differed significantly between malignant pulmonary lesions and organizing pneumonia and between carcinomas and neuroendocrine tumors of the lung. Applying a HU-threshold for segmentation substantially influenced the results of texture analysis.

A pediatric wrist trauma X-ray dataset (GRAZPEDWRI-DX) for machine learning.

Digital radiography is widely available and the standard modality in trauma imaging, often enabling to diagnose pediatric wrist fractures. However, image interpretation requires time-consuming specialized training. Due to astonishing progress in computer vision algorithms, automated fracture detection has become a topic of research interest. This paper presents the GRAZPEDWRI-DX dataset containing annotated pediatric trauma wrist radiographs of 6,091 patients, treated at the Department for Pediatric Surgery of the University Hospital Graz between 2008 and 2018. A total number of 10,643 studies (20,327 images) are made available, typically covering posteroanterior and lateral projections. The dataset is annotated with 74,459 image tags and features 67,771 labeled objects. We de-identified all radiographs and converted the DICOM pixel data to 16-Bit grayscale PNG images. The filenames and the accompanying text files provide basic patient information (age, sex). Several pediatric radiologists annotated dataset images by placing lines, bounding boxes, or polygons to mark pathologies like fractures or periosteal reactions. They also tagged general image characteristics. This dataset is publicly available to encourage computer vision research.

Non-contrast-enhanced CT texture analysis of primary and metastatic pancreatic ductal adenocarcinomas: value in assessment of histopathological grade and differences between primary and metastatic lesions.

PURPOSE: To evaluate the utility of non-contrast-enhanced CT texture analysis (CTTA) for predicting the histopathological differentiation of pancreatic ductal adenocarcinomas (PDAC) and to compare non-contrast-enhanced CTTA texture features between primary PDAC and hepatic metastases of PDAC. METHODS: This retrospective study included 120 patients with histopathologically confirmed PDAC. Sixty-five patients underwent CT-guided biopsy of primary PDAC, while 55 patients underwent CT-guided biopsy of hepatic PDAC metastasis. All lesions were segmented in non-contrast-enhanced CT scans for CTTA based on histogram analysis, co-occurrence matrix, and run-length matrix. Statistical analysis was conducted for 372 texture features using Mann-Whitney U test, Bonferroni-Holm correction, and receiver operating characteristic (ROC) analysis. A p value < 0.05 was considered statistically significant. RESULTS: Three features were identified that differed significantly between histopathological G2 and G3 primary tumors. Of these, "low gray-level zone emphasis" yielded the largest AUC (0.87 ± 0.04), reaching a sensitivity and specificity of 0.76 and 0.83, respectively, when a cut-off value of 0.482 was applied. Fifty-four features differed significantly between primary and hepatic metastatic PDAC. CONCLUSION: Non-contrast-enhanced CTTA of PDAC identified differences in texture features between primary G2 and G3 tumors that could be used for non-invasive tumor assessment. Extensive differences between the features of primary and metastatic PDAC on CTTA suggest differences in tumor microenvironment.

Pediatric radius torus fractures in x-rays-how computer vision could render lateral projections obsolete.

It is an indisputable dogma in extremity radiography to acquire x-ray studies in at least two complementary projections, which is also true for distal radius fractures in children. However, there is cautious hope that computer vision could enable breaking with this tradition in minor injuries, clinically lacking malalignment. We trained three different state-of-the-art convolutional neural networks (CNNs) on a dataset of 2,474 images: 1,237 images were posteroanterior (PA) pediatric wrist radiographs containing isolated distal radius torus fractures, and 1,237 images were normal controls without fractures. The task was to classify images into fractured and non-fractured. In total, 200 previously unseen images (100 per class) served as test set. CNN predictions reached area under the curves (AUCs) up to 98% [95% confidence interval (CI) 96.6%-99.5%], consistently exceeding human expert ratings (mean AUC 93.5%, 95% CI 89.9%-97.2%). Following training on larger data sets CNNs might be able to effectively rule out the presence of a distal radius fracture, enabling to consider foregoing the yet inevitable lateral projection in children. Built into the radiography workflow, such an algorithm could contribute to radiation hygiene and patient comfort.

Diagnosing Pulmonary Embolism With Computed Tomography Pulmonary Angiography: Diagnostic Accuracy of a Reduced Scan Range.

PURPOSE: Computed tomography pulmonary angiography (CT-PA) is frequently used in the diagnostic workup of pulmonary embolism (PE), even in highly radiosensitive patient populations. This study aims to assess CT-PA with reduced z -axis coverage (compared with a standard scan range covering the entire lung) for its sensitivity for detecting PE and its potential to reduce the radiation dose. MATERIALS AND METHODS: We retrospectively analyzed 602 consecutive CT-PA scans with definite or possible PE reported. A reduced scan range was defined based on the topogram, where the cranial slice was set at the top of the aortic arch and the caudal slice at the top of the lower hemidiaphragm. Locations of emboli in relation to the reduced scan range were recorded. RESULTS: We included 513 CT-PA scans with definite acute PE in statistical analysis. Patients' median age was 66 (52 to 77) years, 46% were female. Median dose length product was 270.8 (111.3 to 503.9) mGy*cm. Comparing the original and reduced scan ranges, the mean scan length was significantly reduced by 48.0±8.6% (26.8±3.0 vs. 13.9±2.6 cm, P <0.001). Single emboli outside the reduced range in addition to emboli within were found in 15 scans (2.9%), while only 1 scan (0.2%) had an embolus outside the reduced range and none within it. The resulting sensitivity of CT-PA with reduced scan range was 99.81% (95% confidence interval: 98.74%-99.99%) for detecting any PE. CONCLUSION: A reduced scan length in CT-PA, as defined above, would substantially decrease radiation dose while maintaining diagnostic accuracy for detecting PE.

Phantom study for comparison between computed tomography- and C-Arm computed tomography-guided puncture applied by residents in radiology. / Phantomstudie zum Vergleich zwischen Computertomografie- und C-Arm-Computertomografie-gesteuertem Punktionsverfahren bei Anwendung durch Weiterbildungsassistenten in der Radiologie.

PURPOSE: Comparison of puncture deviation and puncture duration between computed tomography (CT)- and C-arm CT (CACT)-guided puncture performed by residents in training (RiT). METHODS: In a cohort of 25 RiTs enrolled in a research training program either CT- or CACT-guided puncture was performed on a phantom. Prior to the experiments, the RiT's level of training, experience playing a musical instrument, video games, and ball sports, and self-assessed manual skills and spatial skills were recorded. Each RiT performed two punctures. The first puncture was performed with a transaxial or single angulated needle path and the second with a single or double angulated needle path. Puncture deviation and puncture duration were compared between the procedures and were correlated with the self-assessments. RESULTS: RiTs in both the CT guidance and CACT guidance groups did not differ with respect to radiologic experience (p = 1), angiographic experience (p = 0.415), and number of ultrasound-guided puncture procedures (p = 0.483), CT-guided puncture procedures (p = 0.934), and CACT-guided puncture procedures (p = 0.466). The puncture duration was significantly longer with CT guidance (without navigation tool) than with CACT guidance with navigation software (p < 0.001). There was no significant difference in the puncture duration between the first and second puncture using CT guidance (p = 0.719). However, in the case of CACT, the second puncture was significantly faster (p = 0.006). Puncture deviations were not different between CT-guided and CACT-guided puncture (p = 0.337) and between the first and second puncture of CT-guided and CACT-guided puncture (CT: p = 0.130; CACT: p = 0.391). The self-assessment of manual skills did not correlate with puncture deviation (p = 0.059) and puncture duration (p = 0.158). The self-assessed spatial skills correlated positively with puncture deviation (p = 0.011) but not with puncture duration (p = 0.541). CONCLUSION: The RiTs achieved a puncture deviation that was clinically adequate with respect to their level of training and did not differ between CT-guided and CACT-guided puncture. The puncture duration was shorter when using CACT. CACT guidance with navigation software support has a potentially steeper learning curve. Spatial skills might accelerate the learning of image-guided puncture. KEY POINTS: · The CT-guided and CACT-guided puncture experience of the RiTs selected as part of the program "Researchers for the Future" of the German Roentgen Society was adequate with respect to the level of training.. · Despite the lower collective experience of the RiTs with CACT-guided puncture with navigation software assistance, the learning curve regarding CACT-guided puncture may be faster compared to the CT-guided puncture technique.. · If the needle path is complex, CACT guidance with navigation software assistance might have an advantage over CT guidance.. CITATION FORMAT: · Meine TC, Hinrichs JB, Werncke T et al. Phantom study for comparison between computed tomography- and C-Arm computed tomography-guided puncture applied by residents in radiology. Fortschr Röntgenstr 2022; 194: 272 - 280.

Contrast media dose adjustment to allometric parameters of body mass in multiphasic CT of the liver: A comparison of different metrics.

PURPOSE: To assess correlations of lean body weight (LBW) calculated with various formulas, total body weight (TBW), body height (BH), body mass index (BMI), body surface area (BSA) and fat-free mass (FFM) with vascular and parenchymal enhancement in multiphasic CT of the liver. METHOD: Thirty consecutive patients underwent multiphasic CT of the liver using constant iodine dose and flow rate. Contrast enhancement of aorta, portal vein and liver was calculated by measuring mean vascular and parenchymal attenuation in pre-contrast and post-contrast phases. Correlations of TBW, BH, BMI, BSA, FFM, and LBW (calculated with formulas of Boer, Hume, James and Green&Duffull) with enhancement were tested using Spearman's correlation coefficient. The method of Fieller et al. was used to calculate 95 % confidence intervals. A p-value ≤ 0.05 was considered statistically significant. RESULTS: Aortal enhancement correlated strongly with TBW, BSA, LBWBoer and LBWHume and moderately with BH, BMI, FFM, LBWJames and LBWGreen&Duffull. Liver enhancement in the late arterial phase correlated moderately with TBW, FFM, LBWBoer, LBWHume and LBWGreen&Duffull and weakly with BSA. Liver enhancement in the portal venous phase correlated strongly with TBW, BSA, FFM, LBWBoer, LBWHume and LBWGreen&Duffull, whereby overlap of the 95 % CI graphs demonstrated that the differences in the correlation coefficients were not statistically significant. Liver enhancement in the delayed phase correlated moderately with BH but did not correlate significantly with any other parameter. CONCLUSION: Regardless of the form used for calculation, LBW did not correlate statistically significantly stronger than TBW with vascular or parenchymal enhancement of the liver.

The value of MRI compared to conventional radiography in analysing morphologic changes in the spine in axial spondyloarthritis.

BACKGROUND: Imaging of morphologic changes in the vertebral spine in axial spondyloarthritis (SpA) is routinely performed with conventional radiography limited by superposition in the thoracic segments and radiation exposure. The objective was to assess the reliability of MRI compared to conventional radiography in depicting morphologic vertebral lesions in patients with axial SpA. Forty patients diagnosed with axial SpA were included in this cross-sectional study. Patients underwent MRI of the whole spine with T1-weighted and TIRM sequences in the sagittal plane and conventional radiography of the cervical and lumbar spine in lateral projections. Morphologic changes (syndesmophytes and erosions) in the anterior vertebral endplates on MRI and conventional radiography were independently evaluated by two radiologists. Inter-modality and interobserver agreement were calculated using Cohen's Kappa. RESULTS: Inter-modality agreement was low for cervical and lumbar syndesmophytes and erosions (κ ≤ 0.2 ± 0.07-0.1). Interobserver agreement on conventional radiography was highest for cervical and lumbar anterior syndesmophytes/bridging (κ = 0.92 ± 0.02-0.03). Syndesmophytes in thoracic anterior vertebral units were the most frequent MRI finding with a high interobserver agreement (κ = 0.83 ± 0.05). CONCLUSIONS: In imaging morphologic changes in the spine in patients with axial SpA, MRI was shown to be not an equivalent substitute but a complementary imaging modality to conventional radiography. Conventional radiography seems superior to depict morphologic cervical and lumbar vertebral changes compared to MRI, whereas MRI may visualise morphologic lesions in the thoracic spine.