Search

Rationale: Chronic obstructive pulmonary disease (COPD) and emphysema are associated with endothelial damage and altered pulmonary microvascular perfusion. The molecular mechanisms underlying these changes are poorly understood in patients, in part because of the inaccessibility of the pulmonary vasculature. Peripheral blood mononuclear cells (PBMCs) interact with the pulmonary endothelium. Objectives: To test the association between gene expression in PBMCs and pulmonary microvascular perfusion in COPD. Methods: The Multi-Ethnic Study of Atherosclerosis (MESA) COPD Study recruited two independent samples of COPD cases and controls with ⩾10 pack-years of smoking history. In both samples, pulmonary microvascular blood flow, pulmonary microvascular blood volume, and mean transit time were assessed on contrast-enhanced magnetic resonance imaging, and PBMC gene expression was assessed by microarray. Additional replication was performed in a third sample with pulmonary microvascular blood volume measures on contrast-enhanced dual-energy computed tomography. Differential expression analyses were adjusted for age, gender, race/ethnicity, educational attainment, height, weight, smoking status, and pack-years of smoking. Results: The 79 participants in the discovery sample had a mean age of 69 ± 6 years, 44% were female, 25% were non-White, 34% were current smokers, and 66% had COPD. There were large PBMC gene expression signatures associated with pulmonary microvascular perfusion traits, with several replicated in the replication sets with magnetic resonance imaging ( n  = 47) or dual-energy contrast-enhanced computed tomography ( n  = 157) measures. Many of the identified genes are involved in inflammatory processes, including nuclear factor-κB and chemokine signaling pathways. Conclusions: PBMC gene expression in nuclear factor-κB, inflammatory, and chemokine signaling pathways was associated with pulmonary microvascular perfusion in COPD, potentially offering new targetable candidates for novel therapies.

Purpose: Radiology reports mostly contain free-text, which makes it challenging to obtain structured data. Natural language processing (NLP) techniques transform free-text reports into machine-readable document vectors that are important for creating reliable, scalable methods for data analysis. The aim of this study is to classify unstructured radiograph reports according to fractures of the distal fibula and to find the best text mining method., Materials & Methods: We established a novel German language report dataset: a designated search engine was used to identify radiographs of the ankle and the reports were manually labeled according to fractures of the distal fibula. This data was used to establish a machine learning pipeline, which implemented the text representation methods bag-of-words (BOW), term frequency-inverse document frequency (TF-IDF), principal component analysis (PCA), non-negative matrix factorization (NMF), latent Dirichlet allocation (LDA), and document embedding (doc2vec). The extracted document vectors were used to train neural networks (NN), support vector machines (SVM), and logistic regression (LR) to recognize distal fibula fractures. The results were compared via cross-tabulations of the accuracy (acc) and area under the curve (AUC)., Results: In total, 3268 radiograph reports were included, of which 1076 described a fracture of the distal fibula. Comparison of the text representation methods showed that BOW achieved the best results (AUC = 0.98; acc = 0.97), followed by TF-IDF (AUC = 0.97; acc = 0.96), NMF (AUC = 0.93; acc = 0.92), PCA (AUC = 0.92; acc = 0.9), LDA (AUC = 0.91; acc = 0.89) and doc2vec (AUC = 0.9; acc = 0.88). When comparing the different classifiers, NN (AUC = 0,91) proved to be superior to SVM (AUC = 0,87) and LR (AUC = 0,85)., Conclusion: An automated classification of unstructured reports of radiographs of the ankle can reliably detect findings of fractures of the distal fibula. A particularly suitable feature extraction method is the BOW model., Key Points: · The aim was to classify unstructured radiograph reports according to distal fibula fractures.. · Our automated classification system can reliably detect fractures of the distal fibula.. · A particularly suitable feature extraction method is the BOW model.., Citation Format: · Dewald CL, Balandis A, Becker LS et al. Automated Classification of Free-Text Radiology Reports: Using Different Feature Extraction Methods to Identify Fractures of the Distal Fibula. Fortschr Röntgenstr 2023; 195: 713 - 719., Competing Interests: The authors declare that they have no conflict of interest., (The Author(s). This is an open access article published by Thieme under the terms of the Creative Commons Attribution-NonDerivative-NonCommercial-License, permitting copying and reproduction so long as the original work is given appropriate credit. Contents may not be used for commercial purposes, or adapted, remixed, transformed or built upon. (https://creativecommons.org/licenses/by-nc-nd/4.0/).)

Background Experience with functional CT in the lungs without additional equipment in clinical routine is limited. Purpose To report initial experience and evaluate the robustness of a modified chest CT protocol and photon-counting CT (PCCT) for comprehensive analysis of pulmonary vasculature, perfusion, ventilation, and morphologic structure in a single examination. Materials and Methods In this retrospective study, consecutive patients with clinically indicated CT for various known and unknown pulmonary function impairment (six subgroups) were included between November 2021 and June 2022. After administration of an intravenous contrast agent, inspiratory PCCT was followed by expiratory PCCT after a delay of 5 minutes. Advanced automated postprocessing was performed, and CT-derived functional parameters were calculated (regional ventilation, perfusion, late contrast enhancement, and CT angiography). Mean intravascular contrast enhancement in the mediastinal vessels and radiation dose were determined. Using analysis of variance, the mean values of lung volumes, attenuation, ventilation, perfusion, and late contrast enhancement were tested for differences between subgroups of patients. Results In 166 patients (mean age, 63.2 years ± 14.2 [SD]; 106 male patients), all CT-derived parameters could be acquired (84.7% success rate; 166 of 196 patients). At the inspiratory examination, mean density was 325 HU in the pulmonary trunk, 260 HU in the left atrium, and 252 HU in the ascending aorta. The mean dose-length product for inspiration and expiration was 110.32 mGy · cm and 109.47 mGy · cm, respectively; the mean CT dose index for inspiration and expiration was 3.22 mGy and 3.09 mGy, respectively (less than the mean total radiation dose of 8-12 mGy, which is diagnostic reference level). Significant differences ( P < .05) between the subgroups were found for all assessed parameters. Visual inspection allowed for voxelwise assessment of morphologic structure and function. Conclusion The proposed PCCT protocol allowed for a dose-efficient and robust simultaneous evaluation of pulmonary morphologic structure, ventilation, vasculature, and parenchymal perfusion in a procedure requiring advanced software but no additional hardware. © RSNA, 2023.

Fibrosis represents a relevant response to the implantation of biomaterials, which occurs not only at the tissue-material interface (fibrotic encapsulation) but also within the void fraction of complex three-dimensional (3D) biomaterial constructions (fibrotic ingrowth). Usual evaluation of the biocompatibility mostly depicts fibrosis at the interface of the biomaterial using semiquantitative scores. Here, the relations between encapsulation and infiltrating fibrotic growth are poorly represented. Virtual pathology and digital image analysis provide new strategies to assess fibrosis in a more differentiated way. In this study, we adopted a method previously used to quantify fibrosis in visceral organs to the quantification of fibrosis to 3D biomaterials. In a proof-of-concept study, we transferred the "Collagen Proportionate Area" (CPA) analysis from hepatology to the field of biomaterials. As one task of an experimental animal study, we used CPA analysis to quantify the fibrotic ingrowth into a filamentous scaffold after subcutaneous implantation. We were able to demonstrate that the application of the CPA analysis is well suited as an additional fibrosis evaluation strategy for new biomaterial constructions. The CPA method can contribute to a better understanding of the fibrotic interactions between 3D scaffolds and the host tissue responses., (© 2021 The Authors. Journal of Biomedical Materials Research Part A published by Wiley Periodicals LLC.)

Objectives: To evaluate the agreement in detecting pulmonary perfusion defects in patients with chronic thromboembolic pulmonary hypertension using dual-energy CT and dynamic contrast-enhanced MRI. Second, to compare both imaging modalities in monitoring lung perfusion changes in these patients after undergoing pulmonary endarterectomy., Methods: 20 patients were examined with CT and MRI before and/or after pulmonary endarterectomy. Estimated perfusion defect percentage from both modalities was compared in a lobe-based analysis. Spatial agreement of perfusion defect maps was also assessed., Results: A significant correlation between CT and MRI based perfusion defect percentage was calculated in all lung lobes (r > 0.78; p < 0.001). In addition, a good spatial agreement between perfusion defect maps was found (mean spatial overlap for the whole lung was 68.2%; SD = 6.9). Both CT and MRI detected improvements in pulmonary perfusion after pulmonary endarterectomy: 8% and 7% decrease in whole lung perfusion defect percentage (p = 0.007 and 0.004), respectively. In a lobe-wise analysis, improvements were statistically significant only in lower lobes using both modalities (reduction in defect percentage ranged from 16-29%; p < 0.02)., Conclusions: Dual-energy CT is an alternative to MRI in monitoring chronic thromboembolic pulmonary hypertension. Both imaging modalities provided comparable estimations of perfusion defects and could detect similar improvement in lung perfusion after pulmonary endarterectomy., Competing Interests: The authors have declared that no competing interests exist.

Objectives: To develop and evaluate a deep learning algorithm for fully automated detection of primary sclerosing cholangitis (PSC)-compatible cholangiographic changes on three-dimensional magnetic resonance cholangiopancreatography (3D-MRCP) images., Methods: The datasets of 428 patients (n = 205 with confirmed diagnosis of PSC; n = 223 non-PSC patients) referred for MRI including MRCP were included in this retrospective IRB-approved study. Datasets were randomly assigned to a training (n = 386) and a validation group (n = 42). For each case, 20 uniformly distributed axial MRCP rotations and a subsequent maximum intensity projection (MIP) were calculated, resulting in a training database of 7720 images and a validation database of 840 images. Then, a pre-trained Inception ResNet was implemented which was conclusively fine-tuned (learning rate 10 -3 )., Results: Applying an ensemble strategy (by binning of the 20 axial projections), the mean absolute error (MAE) of the developed deep learning algorithm for detection of PSC-compatible cholangiographic changes was lowered from 21 to 7.1%. Sensitivity, specificity, positive predictive (PPV), and negative predictive value (NPV) for detection of these changes were 95.0%, 90.9%, 90.5%, and 95.2% respectively., Conclusions: The results of this study demonstrate the feasibility of transfer learning in combination with extensive image augmentation to detect PSC-compatible cholangiographic changes on 3D-MRCP images with a high sensitivity and a low MAE. Further validation with more and multicentric data is now desirable, as it is known that neural networks tend to overfit the characteristics of the dataset., Key Points: • The described machine learning algorithm is able to detect PSC-compatible cholangiographic changes on 3D-MRCP images with high accuracy. • The generation of 2D projections from 3D datasets enabled the implementation of an ensemble strategy to boost inference performance.

Magnetic resonance imaging (MRI) is an emerging tool for diagnosis and treatment monitoring of chronic thromboembolic pulmonary hypertension (CTEPH). The current study aims to identify central pulmonary arterial hemodynamic parameters that reflect clinical, cardiac and pulmonary changes after PEA. 31 CTEPH patients, who underwent PEA and received pre- and postoperative MRI, were analyzed retrospectively. Central pulmonary arterial blood flow, lung perfusion and right heart function data were derived from MRI. Mean pulmonary arterial pressure (mPAP) and 5-month follow-up six-minute walk-distance (6MWD) were assessed. After PEA, mPAP decreased significantly and patients achieved a higher 6MWD. Central pulmonary arterial blood flow velocities, pulmonary blood flow (PBF) and right ventricular function increased significantly. Two-dimensional (2D) phase-contrast (PC) MRI-derived average mean velocity, maximum mean velocity and deceleration volume changes after PEA correlated with changes of 6MWD and right heart ejection fraction (RVEF). Deceleration volume is a novel 2D PC MRI parameter showing further correlation with PBF changes. In conclusion, 2D PC MRI-derived main pulmonary hemodynamic changes reflect changes of RVEF, PBF and 5-month follow-up 6MWD and may be used for future CTEPH patient monitoring after PEA., Competing Interests: The authors have declared that no competing interests exist.

Background: Perfusion-weighted (Qw) noncontrast-enhanced proton lung MRI is a promising technique for assessment of pulmonary perfusion, but still requires validation., Purpose: To improve perfusion-weighted phase-resolved functional lung (PREFUL)-MRI, to validate PREFUL with perfusion single photon emission computed tomography (SPECT) as a gold standard, and to compare PREFUL with dynamic contrast-enhanced (DCE)-MRI as a reference., Study Type: Retrospective., Population: Twenty patients with chronic obstructive pulmonary disease (COPD), 14 patients with cystic fibrosis (CF), and 21 patients with chronic thromboembolic pulmonary hypertension (CTEPH) were included., Field Strength/sequence: For PREFUL-MRI, a spoiled gradient echo sequence and for DCE-MRI a 3D time-resolved angiography with stochastic trajectories sequence were used at 1.5T., Assessment: PREFUL-MRI coronal slices were acquired in free-breathing. DCE-MRI was performed in breath-hold with injection of 0.03 mmol/kg bodyweight of gadoteric acid at a rate of 4 cc/s. Perfusion SPECT images were obtained for six CTEPH patients. Images were coregistered. An algorithm to define the appropriate PREFUL perfusion phase was developed using perfusion SPECT data. Perfusion defect percentages (QDP) and Qw-values were calculated for all methods. For PREFUL quantitative perfusion values (PREFUL Q ) and for DCE pulmonary blood flow (PBF) was calculated., Statistical Tests: Obtained parameters were assessed using Pearson correlation and Bland-Altman analysis., Results: Qw-SPECT correlated with Qw-DCE (r = 0.50, P < 0.01) and Qw-PREFUL (r = 0.47, P < 0.01). Spatial overlap of QDP maps showed an agreement ≥67.7% comparing SPECT and DCE, ≥64.1% for SPECT and PREFUL, and ≥60.2% comparing DCE and PREFUL. Significant correlations of Qw-PREFUL and Qw-DCE were found (COPD: r = 0.79, P < 0.01; CF: r = 0.77, P < 0.01; CTEPH: r = 0.73, P < 0.01). PREFUL Q /PBF correlations were similar/lower (CF, CTEPH: P > 0.12; COPD: P < 0.01) compared to Qw-PREFUL/DCE correlations. PREFUL Q -values were higher/similar compared to PBF-values (COPD, CF: P < 0.01; CTEPH: P = 0.026)., Data Conclusion: The automated PREFUL algorithm may allow for noncontrast-enhanced pulmonary perfusion assessment in COPD, CF, and CTEPH patients comparable to DCE-MRI. Level of Evidence 3 Technical Efficacy Stage 2 J. Magn. Reson. Imaging 2020;52:103-114., (© 2020 The Authors. Journal of Magnetic Resonance Imaging published by Wiley Periodicals, Inc. on behalf of International Society for Magnetic Resonance in Medicine.)

Background: Chronic obstructive pulmonary disease (COPD) is associated with high morbidity and mortality. Identification of imaging biomarkers for phenotyping is necessary for future treatment and therapy monitoring. However, translation of visual analytic pipelines into clinics or their use in large-scale studies is significantly slowed by time-consuming postprocessing steps., Purpose: To implement an automated tool chain for regional quantification of pulmonary microvascular blood flow in order to reduce analysis time and user variability., Study Type: Prospective., Population: In all, 90 MRI scans of 63 patients, of which 31 had a COPD with a mean Global Initiative for Chronic Obstructive Lung Disease status of 1.9 ± 0.64 (μ ± σ)., Field Strength/sequence: 1.5T dynamic gadolinium-enhanced MRI measurement using 4D dynamic contrast material-enhanced (DCE) time-resolved angiography acquired in a single breath-hold in inspiration. [Correction added on August 20, 2019, after first online publication: The field strength in the preceding sentence was corrected.] ASSESSMENT: We built a 3D convolutional neural network for semantic segmentation using 29 manually segmented perfusion maps. All five lobes of the lung are denoted, including the middle lobe. Evaluation was performed on 61 independent cases from two sites of the Multi-Ethnic Study of Arteriosclerosis (MESA)-COPD study. We publish our implementation of a model-free deconvolution filter according to Sourbron et al for 4D DCE MRI scans as open source., Statistical Test: Cross-validation 29/61 (# training / # testing), intraclass correlation coefficient (ICC), Spearman ρ, Pearson r, Sørensen-Dice coefficient, and overlap., Results: Segmentations and derived clinical parameters were processed in ~90 seconds per case on a Xeon E5-2637v4 workstation with Tesla P40 GPUs. Clinical parameters and predicted segmentations exhibit high concordance with the ground truth regarding median perfusion for all lobes with an ICC of 0.99 and a Sørensen-Dice coefficient of 93.4 ± 2.8 (μ ± σ)., Data Conclusion: We present a robust end-to-end pipeline that allows for the extraction of perfusion-based biomarkers for all lung lobes in 4D DCE MRI scans by combining model-free deconvolution with deep learning., Level of Evidence: 3 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2020;51:571-579., (© 2019 International Society for Magnetic Resonance in Medicine.)

Purpose: To evaluate the effect of inhaled 100% oxygen on pulmonary circulation dynamics in healthy volunteers using 2D phase-contrast magnetic resonance imaging (2D PC MRI)., Materials and Methods: Twenty-one healthy volunteers were examined at 1.5T. Through-plane 2D PC MRI measurements were performed in the main pulmonary artery during free-breathing and breath-hold. Acceleration time and volume, maximum and minimum area, area change, average and maximum mean velocity, forward volume, heart rate, as well as blood pressure were determined. At baseline, subjects breathed room air. After application of a closed-fit full face mask, three further measurements were conducted: at room air (control), directly after starting 15 L/min 100% oxygen (wash-in), and after 5 minutes during continuous oxygen supply (saturation). Data were analyzed with a mixed linear model. Skewed distributed variables were rank-transformed. Tukey contrasts with family-wise adjusted P-values were applied for pairwise comparisons., Results: Inhaled oxygen affected several hemodynamic parameters. Average mean velocity (P < 0.01: breath-hold during wash-in and saturation, P = 0.03: free-breathing during saturation) and maximum mean velocity (P < 0.01: breath-hold and free-breathing during saturation) decreased. When obtained during free-breathing, acceleration volume (P = 0.02: saturation), area change (P = 0.02: saturation), and maximum area (P = 0.02: wash-in, P = 0.03: saturation) increased, while minimum area and forward volume did not change., Conclusion: Oxygen alters pulmonary circulation dynamics in the main pulmonary artery of healthy volunteers, which can be reliably detected using 2D phase-contrast MRI., Level of Evidence: 2 Technical Efficacy: Stage 1 J. Magn. Reson. Imaging 2017;46:1698-1706., (© 2017 International Society for Magnetic Resonance in Medicine.)

Background: Whole-slide imaging (WSI) has become more prominent and continues to gain in importance in student teaching. Applications with different scope have been developed. Many of these applications have either technical or design shortcomings., Objective: To design a survey to determine student expectations of WSI applications for teaching histological and pathological diagnosis. To develop a new WSI application based on the findings of the survey., Methods: A total of 216 students were questioned about their experiences and expectations of WSI applications, as well as favorable and undesired features. The survey included 14 multiple choice and two essay questions. Based on the survey, we developed a new WSI application called Pate utilizing open source technologies., Results: The survey sample included 216 students-62.0% (134) women and 36.1% (78) men. Out of 216 students, 4 (1.9%) did not disclose their gender. The best-known preexisting WSI applications included Mainzer Histo Maps (199/216, 92.1%), Histoweb Tübingen (16/216, 7.4%), and Histonet Ulm (8/216, 3.7%). Desired features for the students were latitude in the slides (190/216, 88.0%), histological (191/216, 88.4%) and pathological (186/216, 86.1%) annotations, points of interest (181/216, 83.8%), background information (146/216, 67.6%), and auxiliary informational texts (113/216, 52.3%). By contrast, a discussion forum was far less important (9/216, 4.2%) for the students., Conclusions: The survey revealed that the students appreciate a rich feature set, including WSI functionality, points of interest, auxiliary informational texts, and annotations. The development of Pate was significantly influenced by the findings of the survey. Although Pate currently has some issues with the Zoomify file format, it could be shown that Web technologies are capable of providing a high-performance WSI experience, as well as a rich feature set.