Medikamenteninduzierte Überempfindlichkeitsreaktionen (MIHR) stellen ein breites Spektrum immunvermittelter Erkrankungen dar, die nach Exposition gegenüber Arzneimitteln auftreten und verschiedene Organsysteme betreffen können. Zu den wichtigsten Entitäten zählen das DRESS-Syndrom („drug reaction with eosinophilia and systemic symptoms“), das Stevens-Johnson-Syndrom (SJS) und die toxische epidermale Nekrolyse (TEN, Lyell-Syndrom), die Serumkrankheit, medikamenteninduzierte eosinophile Pneumonien sowie Hypersensitivitätspneumonitiden. Pulmonale Manifestationen sind dabei häufig und können den klinischen Verlauf entscheidend beeinflussen. Dieser Übersichtsartikel erläutert die immunologischen Grundlagen, die Pathogenese und Ätiologie der medikamenteninduzierten Hyperreaktivitätserkrankungen, beschreibt ihre klinische Relevanz und stellt die typischen thorakalen Manifestationen und Bildmuster in der Computertomographie (CT) des Thorax dar. Im Vordergrund steht die Bedeutung der radiologischen Diagnostik bei der frühzeitigen Erkennung und Differenzierung dieser potenziell lebensbedrohlichen Krankheitsbilder.

PURPOSE:To adapt and validate a linear binning technique, developed for hyper-polarized 129Xe MRI, for functional lung MRI with matrix-pencil decomposition (MP)-MRI. METHODS:First, a dedicated normalization was applied to the perfusion-weighted and the ventilation-weighted map histograms. Then, using 154 MP-MRI scans of healthy children, reference bins were defined around the peak of the averaged histogram as normal (±1 SD), high (> 1 SD), low (between -1 SD and -2 SD), and defect (< -2 SD), and subsequently applied to a validation dataset comprising healthy children (HC, N = 41) and children with cystic fibrosis (CF, N = 30) to evaluate the accuracy of the classification and its discriminatory power. Furthermore, a third dataset comprising children with CF pre and post elexacaftor/tezacaftor/ivacaftor (ETI) therapy (N = 24) was binned to evaluate the method's sensitivity to treatment effects. Standard outcome parameters, computed with a median threshold, served as a comparison. RESULTS:The adapted linear binning resulted in significantly higher defect and low percentages for perfusion and ventilation between children with CF and HC (p values < 0.0001) with high discrimination (AUCs > 0.85). This was comparable to the standard median-threshold method. Two illustrative cases were included to demonstrate the complementary granularity of the linear binning method. Although standard thresholding indicated improvement in both ventilation and perfusion defects pre and post ETI, linear binning showed that perfusion improvement was restricted to the low category. CONCLUSION:A linear binning method for MP-MRI was developed and validated, providing a healthy reference based on a large dataset and advancing functional lung image processing.

Accurate segmentation of interstitial lung disease (ILD) patterns is essential for quantitative disease assessment and longitudinal monitoring. However, existing approaches remain limited by relying on dense annotations and producing static predictions that cannot be refined, motivating interactive approaches. While promptable models show promise in interactive segmentation, their adaptation to ILDs remains largely unexplored. To address this gap, we investigate prompt-guided foundation models for ILD refinement and present, to the best of our knowledge, the first adaptation of MedSAM2 for interactive 3D ILD segmentation on thoracic CT. We investigate three fine-tuning strategies and multiple clinically motivated prompts: bounding-boxes (BBox), point, lasso, and scribble. On a dataset spanning seven ILD patterns and healthy lung tissue, full model fine-tuning performed best, improving the average Dice score by 4.7 percentage points over MedSAM2.While BBox prompts achieve the strongest performance, non-native MedSAM2 interactions such as lasso and scribble prompts also prove effective. Finally, we present and evaluate a proof-of-concept end-to-end workflow in which MedSAM2 is initialized from an automatic segmentation prior and subsequently refined using radiologist prompts. Model weights and plug-ins made available at: https://github.com/AIHNlab/ILD-SemiSegTool.

Drug-induced hypersensitivity reactions (DIHR) represent a broad spectrum of immune-mediated disorders that occur after exposure to medications and can affect various organ systems. The most important entities include drug reaction with eosinophilia and systemic symptoms (DRESS) syndrome, Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN, Lyell's syndrome), serum sickness, drug-induced eosinophilic pneumonias, and hypersensitivity pneumonitis. Pulmonary manifestations are common and can have a decisive impact on the clinical course. This review article explains the immunological foundations, pathogenesis, and etiology of drug-induced hyperreactivity disorders, describes their clinical relevance, and presents the typical thoracic manifestations and imaging patterns in chest computed tomography (CT). The focus is on the importance of radiological diagnostics for the early detection and differentiation of these potentially life-threatening conditions.

Objectives Magnetic resonance imaging (MRI) is a widely utilized, non-invasive imaging modality for evaluating aortic diseases. In this study, we sought to validate a novel, ECG-gated, non-contrast enhanced cardiovascular magnetic resonance sequence (T2-Prepared GRE and Dixon Water-Fat Separation, T2p GD) in patients with aortic disease — both with and without prior aortic surgery — and to compare its diagnostic performance with that of standard contrast-enhanced MR angiography (CE-MRA). Materials and Methods A total of 55 patients undergoing cardiovascular MRI were recruited, including 26 with and 29 without a history of aortic surgery. The signal-to-noise ratio (SNR) was measured in the aorta, pericardial fat, and myocardial septum, while the contrast-to-noise ratios (CNR) between the aorta and septum, and the aorta and fat, were calculated for both imaging methods, stratified by surgical history. Additionally, diameters of the aorta and coronary arteries were assessed. The diagnostic quality of the images was evaluated using a five-point Likert scale. Results While the SNR in the aorta was comparable between sequences, the new sequence provided higher SNR in both the septum and pericardial fat, as well as a significant increase in the aorta-to-fat CNR. Prior aortic surgery did not affect SNR or CNR values overall; however, in the water-only images of the new sequence, CNR between the aorta and fat was significantly lower in patients with prior surgery. Vascular diameters measured with the T2p GD sequence were consistently larger, as CE-MRA only assesses the endoluminal borders, while the T2p GD sequence includes anatomic delineation of the vessel wall. Qualitative image assessment revealed that the new sequence achieved higher overall quality ratings and superior subjective delineation, enabling more accurate measurement of the aortic sinus and the proximal third of the coronary arteries across all readers. Conclusions The novel T2p GD sequence enables detailed anatomic characterization of the aorta and coronary arteries and a higher CNR than conventional CE-MRA, even in patients with a history of aortic surgery.

Purpose:The diagnosis of primary sclerosing cholangitis remains difficult, particularly in the absence of typical bile duct strictures. This study aims to assess the diagnostic performance of MRI T1 and T2 mapping with regard to differentiating PSC patients from matched controls. Materials and Methods:In this IRB-approved study, we retrospectively included all patients with primary sclerosing cholangitis who underwent multiparametric MRI with T1 and T2 mapping (1.5T and/or 3T) between 09/2018-05/2020 at our institution. Sex-, age-, and field strength-matched patients were selected, without focal or diffuse liver disease and without significant liver steatosis, based on a proton density fat fraction <10%. T1 and T2 relaxation times were compared between patients and controls using a paired Wilcoxon test. Receiver operating characteristic (ROC) curve analysis was performed for the T1 and T2 relaxation times at 1.5T and 3T to discriminate between PSC and non-PSC patients. The cut-off corresponding to the highest Youden index was selected as the optimal threshold. Results:A total of 132 MRI exams were included: 66 from PSC patients (29 at 1.5T and 37 at 3T), and 66 age-, sex-, and field strength-matched controls. The T2 relaxation time was significantly higher in PSC patients at 3T (38 ms [35-40 ms] vs. 35 ms [31-38 ms]) and 1.5T (53 ms [51-57 ms] vs. 50 ms [46-54 ms]). The T1 relaxation time was significantly higher in PSC patients at 3T (830 ms [793-898 ms] vs. 808 ms [745-864 ms]; p=.01), but not at 1.5 T (620 ms [576-658 ms] vs. 583 ms [544-645 ms]; p=.23). The highest diagnostic performance was observed with T2-weighted sequences at 1.5T (AUC = 0.69, p = .013), followed by T1-weighted sequences at 3T (AUC = 0.65, p = .024). T1-weighted imaging at 1.5T showed a trend toward significance (AUC = 0.64, p = .070), while T2-weighted imaging at 3T did not demonstrate significant discriminative ability (AUC = 0.55, p = .433). For parametric T2 mapping at 1.5T, a >47 ms cut-off showed high sensitivity (89.7%) and moderate specificity (48.3%). For parametric T1 mapping at 3T, a >869 ms cut-off showed high specificity (89.2%) but lower sensitivity (40.5%). Conclusion:MRI T2 relaxation times were significantly increased in PSC patients both at 3T and 1.5T, while T1 relaxation time was significantly increased at 3T, but not at 1.5T. T2-weighted imaging at 1.5T demonstrated the highest diagnostic performance. Key Points:· T2 relaxation time is increased in PSC patients both at 1.5T and 3T.. · T1 relaxation time is increased in PSC patients at 3T, but not at 1.5T.. · T2 relaxation time at 1.5T offers the best diagnostic performance.. · T2 mapping represents a promising diagnostic tool for PSC patients, complementing T1 mapping.. Citation Format:· Huber A, Erb S, Ardoino M etal. Liver MRI with T1 and T2 parametric mapping to characterize patients with primary sclerosing cholangitis. Rofo 2026; 10.1055/a-2836-6899.

Background/Objectives: The objective of this study was to investigate the association between congestive heart failure (CHF) and T1 mapping in both liver lobes using cardiac MRI. Methods: This retrospective study included patients who underwent cardiac MRI with T1 mapping sequences on a 1.5 T scanner. The liver T1 values were measured in four hepatic regions, utilizing cardiac short axis and four-chamber views. Echocardiographic and laboratory data were collected within 90 days of the cardiac MRI. Comparisons of the liver T1 values and echocardiographic parameters between patients with and without elevated NT-proBNP levels (>125 pg/mL) were conducted using the Mann–Whitney U test. Logistic regression models were employed to adjust for confounding factors. Results: A total of 397 patients were included (with a median age of 56 years; 127 females), of whom 35% (n = 138) exhibited elevated NT-proBNP levels. The patients with elevated NT-proBNP levels showed a larger end-diastolic volume (EDV: 92 vs. 81 mL/m2, p < 0.001) and a lower LVEF level (50% vs. 60%, p < 0.001). The liver T1 was significantly higher in the right liver lobe (670 vs. 596 ms, p < 0.001) and the caudate lobe (664 vs. 598 ms, p < 0.001), but not in the left lobe (571 vs. 568 ms, p = 0.068) or the dome (590 vs. 560 ms, p = 0.1). T1 mapping in the caudate (OR 1.013, 95% CI 1.004–1.023, p = 0.005) and right liver lobes (OR 1.012, 95% CI 1.003–1.021, p = 0.009) remained independently predictive in the logistic regression analysis. Conclusions: Elevated T1 values in the caudate and right liver lobes assessed by cardiac MRI were independently associated with CHF and outperformed T1 measurements in the left liver lobe in predicting disease.

Rationale and objectivesTo examine the impact of the partial volume effect (PVE) on the imaging of spherical objects depending on their size, density and center voxel position.Materials and methodsWe developed an algorithm for calculating the volume of a sphere wrapped by voxels. The algorithm measured the internal volume of each voxel cut by the sphere and automatically attributed the average voxel density. The sphere volume was simulated by the sum of voxels with an average density above the Hounsfield Unit (HU) cutoff level for that object. Various sphere sizes, densities and positions in the voxel grid were examined. The two clinical settings used were nodules (0 HU) in the lung (-1000 HU) and kidney stones (1000 HU) embedded in the renal parenchyma (30 HU).ResultsSmall kidney stones appeared magnified by the PVE when a stone cutoff level of 130 HU was used: the smallest stone simulated with a diameter of 1.4 mm demonstrated a volume that was 231% the size of the ground truth (sphere volume as measured with the classical formula). A hypothetical stone of 10 cm would still have a PVE of 2%. The PVE did not affect lung nodules if the cutoff level for the nodule fraction was set to the exact mean of both the internal and external density (-500 HU). Lung nodules were more affected by the geometrical effect, where tiny nodules appeared smaller because of the greater curvature of smaller spheres, often cutting less than 50% of the volume of a surface voxel.ConclusionsThis study highlights the potential risks associated with inaccurate raw data postprocessing of CT images with objects that are particularly sensitive to the PVE, such as kidney stones and high-density calcifications (Agatston score).

CLINICAL/METHODICAL ISSUE:Sarcoidosis is a systemic inflammatory disease of unknown cause, characterized by the formation of noncaseating epithelioid cell granulomas, which can potentially affect nearly every organ. The lungs and mediastinal lymph nodes are most commonly affected-with granulomatous involvement found in approximately 90% of cases. Typically, younger adults are affected by sarcoidosis. The clinical picture and the course of the disease can vary significantly, which is why it is also referred to as the "great mimicker". The aim of this review is to provide an overview on the imaging findings in patients with sarcoidosis. In addition, the current diagnostic and therapeutic management will be elaborated. STANDARD RADIOLOGICAL METHODS:The conventional chest radiograph has long been considered a fundamental examination in suspected sarcoidosis. Traditionally, staging is performed according to the Scadding classification (stages 0-IV). However, computed tomography (CT) is increasingly used due to its higher accuracy and availability. METHODICAL INNOVATIONS:Recently, seven dedicated radiologic phenotypes of sarcoidosis have been identified on CT, comprising four nonfibrotic and three fibrotic subtypes. Given the superior accuracy of CT and its potential role in therapy monitoring, baseline high-resolution CT (HRCT) is increasingly recommended. PRACTICAL RECOMMENDATIONS:Sarcoidosis is an inflammatory multisystem disease involving the lungs and thoracic lymph nodes in approximately 90% of cases. Current guidelines still recommend chest radiography as the initial imaging modality for diagnosis and staging; however, CT is increasingly being used.

Masked autoencoders (MAEs) have emerged as a powerful approach for pre-training on unlabelled data, capable of learning robust and informative feature representations. This is particularly advantageous in diffused lung disease research, where annotated imaging datasets are scarce. To leverage this, we train an MAE on a curated collection of over 5,000 chest computed tomography (CT) scans, combining in-house data with publicly available scans from related conditions that exhibit similar radiological patterns, such as COVID-19 and bacterial pneumonia. The pretrained MAE is then fine-tuned on a downstream classification task for diffused lung disease diagnosis. Our findings demonstrate that MAEs can effectively extract clinically meaningful features and improve diagnostic performance, even in the absence of large-scale labelled datasets. The code and the models are available here: https://github.com/eedack01/lung_masked_autoencoder.

To evaluate the prevalence and clinical implications of extra-coronary findings in a large cohort of patients undergoing coronary computed tomography angiography (CCTA). This retrospective study analyzed data from 3295 consecutive CCTA examinations at a single tertiary center. Radiology reports were reviewed for potentially significant extra-coronary findings. Three-year follow-up evaluation was performed via hospital records. Prevalences of confirmed significant findings were determined and extrapolated to patients lost to follow-up. Extra-coronary findings were reported in 92.7

Abstract Interstitial lung diseases (ILDs) include over 200 conditions affecting the lung parenchyma. ILDs are classified as non-fibrosing or fibrosing, with many fibrosing forms initially presenting as inflammatory and only progressing to fibrosis over time. Diagnosing ILDs is challenging due to their variety and rarity, with CT scans playing a key role in the process, though final diagnoses are made through multidisciplinary team discussions. Biopsies are generally avoided unless CT results are inconclusive or conflict with clinical findings, emphasizing the importance of a thorough CT protocol and approach. The systematic evaluation of CT scans for suspected ILD involves a three-step approach: first, describing parenchymal abnormalities; second, assessing extraparenchymal findings such as pleural and mediastinal structures; and finally, determining the differential diagnosis based on the distribution of these findings. Parenchymal abnormalities are divided into decreased lung densities (e.g., emphysema, cysts) and increased lung densities (e.g., ground-glass opacities, consolidations). These findings should follow the Fleischner Society’s terminology for consistency. Extraparenchymal features include abnormalities in the airways, pleura, and mediastinum, as well as possible extra-thoracic manifestations in organs such as liver or spleen. The distribution of findings, whether in different lung zones or within the secondary pulmonary lobule, plays a key role in diagnosing specific diseases, such as differentiating Langerhans cell histiocytosis from lymphangioleiomyomatosis based on cyst location.

Abstract This chapter provides an overview of fibrosing interstitial lung disease (ILD), a broad group of over 200 lung conditions, some of which can progress to severe, irreversible fibrosis. ILDs can have various causes, ranging from environmental exposures to autoimmune diseases, and while some may be treatable or reversible, others, such as idiopathic pulmonary fibrosis (IPF), lead to significant lung damage and poor outcomes. Lung fibrosis occurs when abnormal tissue repair leads to scarring, impairing the lungs’ ability to function properly. The chapter emphasizes the importance of chest CT scans in diagnosing and evaluating lung fibrosis, highlighting common imaging patterns such as honeycombing, traction bronchiectasis, and architectural distortion. Identifying these patterns is critical to diagnose different types of ILDs. The chapter also discusses several common fibrosing ILDs, including IPF, nonspecific interstitial pneumonia (NSIP), sarcoidosis, and hypersensitivity pneumonitis (HP). Each disease has distinct clinical and imaging features, though some, such as chronic HP and IPF, can appear similar, making diagnosis challenging. A systematic, multidisciplinary approach, involving specialists such as radiologists and pulmonologists, is key to accurate diagnosis and effective management of these diseases.

To evaluate a deep learning sequence-adaptive liver multiparametric MRI (mpMRI) assessment with validation in different populations using total and segmental T1 and T2 relaxation time maps. A neural network was trained to label liver segmental parenchyma and its vessels on noncontrast T1-weighted gradient-echo Dixon in-phase acquisitions on 200 liver mpMRI examinations. Then, 120 unseen liver mpMRI examinations of patients with primary sclerosing cholangitis or healthy controls were assessed by coregistering the labels to noncontrast and contrast-enhanced T1 and T2 relaxation time maps for optimization and internal testing. The algorithm was externally tested in a segmental and total liver analysis of previously unseen 65 patients with biopsy-proven liver fibrosis and 25 healthy volunteers. Measured relaxation times were compared to manual measurements using intraclass correlation coefficient (ICC) and Wilcoxon test. Comparison of manual and deep learning-generated segmental areas on different T1 and T2 maps was excellent for segmental (ICC = 0.95 ± 0.1; p < 0.001) and total liver assessment (0.97 ± 0.02, p < 0.001). The resulting median of the differences between automated and manual measurements among all testing populations and liver segments was 1.8 ms for noncontrast T1 (median 835 versus 842 ms), 2.0 ms for contrast-enhanced T1 (median 518 versus 519 ms), and 0.3 ms for T2 (median 37 versus 37 ms). Automated quantification of liver mpMRI is highly effective across different patient populations, offering excellent reliability for total and segmental T1 and T2 maps. Its scalable, sequence-adaptive design could foster comprehensive clinical decision-making. The proposed automated, sequence-adaptive algorithm for total and segmental analysis of liver mpMRI may be co-registered to any combination of parametric sequences, enabling comprehensive quantitative analysis of liver mpMRI without sequence-specific training.

Background:With chronic liver disease (CLD) rising globally, noninvasive methods are needed to stratify early, intermediate, and advanced CLD with and without clinically significant portal hypertension (CSPH). Purpose:To analyze the combined diagnostic value of intravoxel incoherent motion (IVIM) and liver stiffness (LS) from magnetic resonance elastography (MRE) for CLD stage discrimination vs secondary single-parameter analyses. Materials and methods:This retrospective cross-sectional study included 185 patients who underwent 3T liver MRI, including MRE and IVIM, between March 2016 and November 2023. Patients with CLD were grouped based on their liver fibrosis degree into early CLD (F0-F1; n = 21), intermediate CLD (F2; n = 19), advanced CLD (F3-F4, n = 20), and advanced CLD with CSPH (n = 22). CSPH was defined as splenomegaly (>120 mm) with thrombocytopenia (<100 × 109/L), ascites, or portosystemic collaterals. Patients without CLD (n = 103) served as negative controls. IVIM parameters (tissue diffusivity D, perfusion fraction f, and pseudo-diffusion coefficient D*) and MRE LS were analyzed. Statistical analysis included the Kruskal-Wallis test and both univariate and multivariate regression. Results:In total, 185 patients (median age: 55 years, interquartile range 25%-75%: 45-63 years; 94 men) were evaluated. All parameters differed significantly between all groups (P < .001). f and D* decreased with disease progression, while LS increased. D initially decreased in patients with CLD but increased in those with CSPH. Consequently, higher D-values indicated the presence of CSPH in advanced stages (odds ratio [OR] 1.09, 95% CI 1.03-1.17, P = .009). Elevated LS values showed strong associations with the presence of CLD (OR 5.01, CI 2.25-12.65, P < .001). Combining D and LS further improved diagnostic differentiation between disease stages, especially for differentiation between advanced CLD and advanced CLD with CSPH (OR 2.41, CI 1.33-5.44, P = .01). Conclusion:IVIM and MRE are useful for characterizing CLD and CSPH. Combining D from IVIM with LS from MRE improves diagnostic accuracy compared to MRE alone.

To determine whether the nonfunctional liver volume (NFLV) is an indicator of chronic liver disease (CLD). Multiparametric 3T abdominal MRI examinations enhanced with gadobenate dimeglumine of 51 patients were included in the study and divided into two groups: patients with (n=20) and without (n=31) CLD. Pre- and postcontrast T1 relaxation times of the liver and aorta were measured in the T1 mapping sequences. Total and segmental liver volumes (Lvol) were determined using a convolutional neuronal network. The functional liver fraction (FLF) defined as [(1/T1liver postcontrast − 1/T1liver precontrast) ÷ (1/T1blood pool postcontrast − 1/T1blood pool precontrast)] × (1 − hematocrit) and the nonfunctional liver volume (NFLV) defined as (1 − FLF) × Lvol were calculated for the whole liver, segments I–III, and IV–VIII. Volumes, FLF, and NFLV were compared between the groups using the Mann-Whitney U test and receiver operation characteristics (ROC) analysis. Volumes were significantly higher in patients with CLD than without CLD for the whole liver (p<.01), segments I–III (p<.001), and segments IV–VIII (p<.01). No significant difference was found regarding FLF (p=.20–31). NFLV of the whole liver (p<.01), segments I–III (p<.001), and IV–VIII (p<.01) were significantly increased in patients with CLD. The highest AUCs were observed for Lvol (AUC=.80; p<.001) and NFLV (AUC=.78; p<.001), both in segments I–III. The optimal NFLV cutoff values for CLD were 745 ml for the whole liver (77 % sensitivity; 75% specificity), 174 ml for segments I–III (85% sensitivity; 70% specificity), and 573 ml for segments IV–VIII (77% sensitivity; 75% specificity). MRI-derived nonfunctional liver volume (NFLV) is helpful for early detection of imaging changes in CLD. NFLV is highly associated with CLD, notably when measured in the liver segments I–III.

Background/Objectives: The risk of hemorrhage during CT-guided lung biopsy has not been systematically studied in cases where ground-glass opacities (GGO) are present in the access route or when biopsies are performed in highly perfused, dependent lung areas. While patient positioning has been studied for pneumothorax prevention, its role in minimizing hemorrhage risk remains unexplored. This study aimed to determine whether GGOs in the access route and biopsies in dependent lung areas are risk factors for pulmonary hemorrhage during CT-guided lung biopsy. Methods: A retrospective analysis was conducted on 115 CT-guided lung biopsies performed at a single center (2020–2023). Patients were categorized based on post-interventional hemorrhage exceeding 2 cm (Grade 2 or higher). We evaluated the presence of GGOs in the access route and biopsy location (dependent vs. non-dependent areas) using chi square, Fisher’s exact, and Mann–Whitney U tests. Univariate and multivariate logistic regression analyses were conducted to evaluate risk factors for pulmonary hemorrhage. Results: Pulmonary hemorrhage beyond 2 cm occurred in 30 of 115 patients (26%). GGOs in the access route were identified in 67% of these cases (p < 0.01), and hemorrhage occurred more frequently when biopsies were performed in dependent lung areas (63% vs. 40%, p = 0.03). Multivariable analysis showed that GGOs in the access route (OR 5.169, 95% CI 1.889–14.144, p = 0.001) and biopsies in dependent areas (OR 4.064, 95% CI 1.477–11.186, p < 0.001) independently increased hemorrhage risk. Conclusions: GGOs in the access route and dependent lung area biopsies are independent risk factors for hemorrhage during CT-guided lung biopsy.

The COVID-19 pandemic has significantly impacted healthcare practices worldwide, including the use of computed tomography (CT) in chest examinations. This study aims to analyse the trends in CT chest examinations before, during, and after the pandemic. Data were retrospectively collected from 10 public hospitals over a period spanning from pre-pandemic years (2019 until February 2020) through the pandemic (March 2020–April 2023) (as defined by the World Health Organisation) and into the post-pandemic phase (May 2023–December 2024). Patient exposure information from more than 240,000 CT examinations was collected and analysed using a commercial dose management system. Statistical analysis was performed with descriptive and inferential statistics employed to compare the number of CT chest examinations and patient radiation exposure across different periods. The results indicate a marked increase in CT chest examinations during the pandemic, which reached a plateau and remained stable post-pandemic (p-value < 0.001). Importantly, the average radiation exposure per patient has decreased with the evolution of technology, indicating improved dose management practices. The results also showed a shift between protocols used for CT chest examinations during the pandemic, with the move towards more dedicated procedures. These findings highlight the sustained demand for CT chest examinations and the effective management of patient radiation exposure during and after the pandemic. The study underscores that those possible practices obtained during the pandemic became the norm after the end of the pandemic. Question This study analyses trends in CT chest examinations and patient radiation exposure before, during, and after the COVID-19 pandemic. Findings CT chest examinations rose sharply during the pandemic, then stabilised, while average patient radiation exposure remained consistent throughout. Clinical relevance The findings demonstrate ongoing demand for CT chest scans and radiation management, emphasizing the need for sustainable imaging practices in future healthcare planning.

Medical image segmentation is a crucial element of computer-aided diagnosis (CAD) systems. Segmentation maps are used to calculate imaging features, such as quantitative disease distribution and radiomic features. Since their introduction in 2015, UNets have become the state-of-the-art segmentation tools. However, since that time, many new methods for image processing have been introduced, such as vision transformers and multi-layer-perceptron-mixers (MLP-Mixers). Alongside baseline UNets, we have now investigated the application of such MLP-Mixers for medical image segmentation, as part of a CAD system for the diagnosis of interstitial lung diseases (ILDs). Furthermore, we have investigated the effect of 2D and 3D data representations on segmentation and the final CAD results. We have evaluated the performance of the baseline segmentation methods and the MLP-Mixer primary on the overall diagnostic performance of the CAD system - as well as on the accuracy of segmentation as an intermediate step. In addition to network and data representation variations, we have investigated two different techniques for selecting features, an agnostic method and an alternative approach which selects features tailored to a specific segmentation map and diagnosis task. Finally, the CAD’s performance was compared with that of four independent specialists in chest radiology. Among the 105 test cases, the diagnostic accuracy was 77.2±1.6% for the AI-approaches and 79.0±6.9% for the radiologists, indicating that the proposed systems perform comparably well to human readers in most of the cases. For the task of ILD pattern segmentation, similar results were obtained with 3D data and 2D tomography slices.

Antifibrotic therapy with nintedanib is the clinical mainstay in the treatment of progressive fibrosing interstitial lung disease (ILD). High-dimensional medical image analysis, known as radiomics, provides quantitative insights into organ-scale pathophysiology, generating digital disease fingerprints. Here, we performed an integrative analysis of radiomic and proteomic profiles (radioproteomics) to assess whether changes in radiomic signatures can stratify the degree of antifibrotic response to nintedanib in (experimental) fibrosing ILD. Unsupervised clustering of delta radiomic profiles revealed 2 distinct imaging phenotypes in mice treated with nintedanib, contrary to conventional densitometry readouts, which showed a more uniform response. Integrative analysis of delta radiomics and proteomics demonstrated that these phenotypes reflected different treatment response states, as further evidenced on transcriptional and cellular levels. Importantly, radioproteomics signatures paralleled disease- and drug-related biological pathway activity with high specificity, including extracellular matrix (ECM) remodeling, cell cycle activity, wound healing, and metabolic activity. Evaluation of the preclinical molecular response-defining features, particularly those linked to ECM remodeling, in a cohort of nintedanib-treated fibrosing patients with ILD, accurately stratified patients based on their extent of lung function decline. In conclusion, delta radiomics has great potential to serve as a noninvasive and readily accessible surrogate of molecular response phenotypes in fibrosing ILD. This could pave the way for personalized treatment strategies and improved patient outcomes.