Radiology. Cardiothoracic imaging最新文献

{"title":"Machine Learning-based Model for Major Adverse Cardiac Event Prediction in Patients with Hypertrophic Cardiomyopathy.","authors":"Thomas Geyer, Christopher McIntosh, Vishesh Sood, João Francisco Matos, Hayley McKee, Kate Hanneman","doi":"10.1148/ryct.250433","DOIUrl":"10.1148/ryct.250433","url":null,"abstract":"<p><p>Purpose To develop a machine learning-based model to identify patients with hypertrophic cardiomyopathy (HCM) at high risk of major adverse cardiac events (MACEs) and key predictors of model performance. Materials and Methods This retrospective cohort study included patients who underwent cardiac MRI for HCM evaluation between September 2015 and December 2022. Cardiac MRI included balanced cine steady-state free precession, native T1 and T2 mapping, and late gadolinium enhancement. MACEs were defined as a composite of cardiovascular death, resuscitated sudden cardiac death, or heart failure hospitalization. A penalized Cox proportional hazards model with elastic net regularization (CoxNet) was developed with 33 clinical, genetic, echocardiography, and cardiac MRI variables. Model training used 200 iterations of stratified subsampling cross-validation (80% training, 20% testing). Performance was evaluated with the Harrell C index and compared with the 2014 European Society of Cardiology sudden cardiac death risk model. Results A total of 604 patients were included (mean age, 52 years ± 15 [SD]; 417 male patients; median follow-up, 3.0 years [IQR, 1.9-4.2 years]). The CoxNet model demonstrated favorable performance for MACE prediction (C index, 0.75; 95% CI: 0.65, 0.83), similar to the European Society of Cardiology model (C index, 0.67; 95% CI: 0.57, 0.75; <i>P</i> = .07). Key predictors included apical aneurysm, left ventricular end-systolic volume indexed to body surface area, extensive late gadolinium enhancement, native T1 <i>z</i> score, and male sex. Conclusion A machine learning-based model comprising routinely available variables showed strong performance for MACE prediction in HCM. Key variables highlight the impact of cardiac MRI features on risk stratification. <b>Keywords:</b> MRI, Machine Learning, Model Training, Cardiac, Heart, Hyperplasia, Hypertrophy © RSNA, 2026 See also commentary by Shiwani in this issue.</p>","PeriodicalId":21168,"journal":{"name":"Radiology. Cardiothoracic imaging","volume":"8 3","pages":"e250433"},"PeriodicalIF":4.2,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148157948","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Estimation of Lung Volume Superimposed on the Mediastinum and Diaphragm on Chest Radiographs Using Upright Multidetector Row CT.","authors":"Yoshitake Yamada, Minoru Yamada, Shotaro Chubachi, Yoichi Yokoyama, Keita Nakatsuka, Naoki Hiraishi, Takayuki Abe, Koichi Fukunaga, Masahiro Jinzaki","doi":"10.1148/ryct.250520","DOIUrl":"10.1148/ryct.250520","url":null,"abstract":"<p><p>Purpose To quantify lung and lobar volumes superimposed on the mediastinum and diaphragm on upright chest radiographs using upright multidetector row CT and to compare them with supine measurements. Materials and Methods In this prospective study, asymptomatic Asian participants underwent both upright and supine multidetector row CT within 2 hours. Lung and lobar volumes were measured in each position. Ray-sum images were reconstructed to delineate mediastinal and diaphragmatic boundaries, and superimposed volumes in the posteroanterior direction were calculated. These CT-derived volumes were considered to approximate lung fields projected over the mediastinum and diaphragm on radiographs. Percentages of superimposed volumes relative to total lung and lobar volumes were determined. Mixed-effects models and multiple linear regression analyses were used. Results The study included 132 participants (mean age ± SD, 47.4 years ± 11.4; 85 female; mean body mass index ± SD, 22.2 ± 3.4 [calculated by dividing weight in kilograms by height in meters squared]). The proportion of lung volume superimposed on the mediastinum and diaphragm was 23.9% ± 3.1 in the upright position, significantly lower than 25.7% ± 3.3 in the supine position (<i>P</i> < .001), despite a 9.8% increase in total lung volume upright. Superimposed percentages of the right upper, middle, and lower lobes and the left upper and lower lobes in the upright position were 10.4% ± 2.5, 7.5% ± 3.2, 28.8% ± 7.9, 12.3% ± 1.9, and 46.0% ± 8.3, respectively. All lobes except the left lower lobe showed significantly lower superimposition upright than supine (all <i>P</i> < .03). Multiple linear regression identified older age, higher body mass index, and male sex as independent predictors of greater superimposition in the upright position (all <i>P</i> < .03). Conclusion Lung volume superimposed on the mediastinum and diaphragm was lower on upright than supine chest radiographs despite greater lung volume in the upright position. University Hospital Medical Information Network Clinical Trials Registry: UMIN000026586 <b>Keywords:</b> CT, Thorax, Lung, Mediastinum, Diaphragm, Anatomy, Comparative Studies, Technology Assessment <i>Supplemental material is available for this article.</i> © The Author(s) 2026. Published by the Radiological Society of North America under a CC BY 4.0 license.</p>","PeriodicalId":21168,"journal":{"name":"Radiology. Cardiothoracic imaging","volume":"8 3","pages":"e250520"},"PeriodicalIF":4.5,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147942938","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Prediction of Anthracycline-induced Cardiotoxicity Using Cardiac MRI Parameters: An Animal Study.","authors":"Na Young Kim, Jinho Park, Heae Surng Park, Yoonjung Kim, Ji Won Kim, Jung Im Jung, Eun-Ju Kang, Kyunghwa Han, Chul Hwan Park, Peder E Z Larson, Yoo Jin Hong","doi":"10.1148/ryct.240525","DOIUrl":"10.1148/ryct.240525","url":null,"abstract":"<p><p>Purpose To evaluate the prognostic utility of cardiac MRI parameters, particularly native T1 and extracellular volume fraction (ECV), for predicting early mortality in a preclinical model of anthracycline-induced cardiotoxicity. Materials and Methods Thirty-one Sprague-Dawley rats received weekly intravenous doxorubicin injections (2 mg/kg) for up to 12 weeks and were followed up until natural death. Cardiac MRI, including parametric mapping sequences, was performed. Myocardial histology was evaluated after death. Cox proportional hazards regression with time-dependent covariates was used to assess the association between MRI variables and early death. Results Of the 31 rats, 17 (55%) died before 12 weeks (early death group), and 14 survived to study completion. In the early death group, native T1 and ECV increased between weeks 4 and 6 (from 1210.51 msec ± 53.75 to 1368.09 msec ± 60.11 and from 17.52% ± 2.05 to 21.63% ± 2.22, respectively; both <i>P</i> < .001). At week 6, both parameters were higher in the early death group than in the survival group (1368.09 msec ± 60.11 vs 1243.10 msec ± 55.71, <i>P</i> = .003; 21.63% ± 2.22 vs 18.47% ± 2.13, <i>P</i> = .045). In univariable analyses, native T1 (hazard ratio, 1.01; 95% CI: 1.00, 1.01; <i>P</i> < .001) and ECV (hazard ratio, 1.17; 95% CI: 1.04, 1.32; <i>P</i> = .01) were statistically significantly associated with early death. A model combining native T1 and ECV achieved a C-index of 0.76 (95% CI: 0.64, 0.92). Conclusion Native T1 and ECV were good predictors of early mortality in rats with anthracycline-induced cardiotoxicity. <b>Keywords:</b> Animal Studies, MR-Imaging, Cardiac, Myocardium, Experimental Investigations <i>Supplemental material is available for this article.</i> © RSNA, 2026.</p>","PeriodicalId":21168,"journal":{"name":"Radiology. Cardiothoracic imaging","volume":"8 3","pages":"e240525"},"PeriodicalIF":4.5,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147842338","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"From Risk Score to Risk Phenotype: Cardiac MRI and Machine Learning in Hypertrophic Cardiomyopathy.","authors":"Hunain Shiwani","doi":"10.1148/ryct.260199","DOIUrl":"https://doi.org/10.1148/ryct.260199","url":null,"abstract":"","PeriodicalId":21168,"journal":{"name":"Radiology. Cardiothoracic imaging","volume":"8 3","pages":"e260199"},"PeriodicalIF":4.2,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148157860","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Dual-Phase 3D Modified Dixon Steady-State MR Angiography in Pediatric Complex Congenital Heart Disease.","authors":"Lucia D Beissel, Christopher Hart, Alexander Isaak, Taraneh Aziz-Safaie, Claus C Pieper, Daniel Kuetting, Shuo Zhang, Christoph Katemann, Julian A Luetkens","doi":"10.1148/ryct.250395","DOIUrl":"10.1148/ryct.250395","url":null,"abstract":"<p><p>Purpose To evaluate the impact of dual-phase three-dimensional (3D) modified Dixon (mDixon) steady-state MR angiography (MRA) on phase-dependent image quality and vessel measurements in pediatric patients with congenital heart disease (CHD). Materials and Methods In this retrospective single-center study, pediatric patients with CHD underwent cardiac MRI at 3 T, including free-breathing dual-phase 3D mDixon steady-state MRA with respiratory navigator gating and interleaved acquisition of diastolic and systolic images. Vessel diameters were measured independently by two readers. Overall image quality, including vessel border delineation and breathing-, motion-, flow-, and susceptibility-related artifacts, was evaluated using a five-point Likert scale (1 = nondiagnostic; 5 = excellent). Paired <i>t</i> tests or Wilcoxon signed rank tests were used for phase comparisons. Results A total of 47 patients (mean age, 6.4 years ± 3.5 [SD]) were included. Arterial vessel diameters were larger in systole than in diastole (eg, main pulmonary artery: mean, 23.2 mm ± 5.3 [SD] vs 19.1 mm ± 3.9; <i>P</i> < .01). Systolic image quality was higher for coronary arteries and pulmonary veins (eg, pulmonary veins: median, 5.00 [IQR, 4.00-5.00] vs 4.00 [IQR, 3.00-4.00]; <i>P</i> < .01), whereas diastolic image quality was higher for most other vessels (eg, main pulmonary artery: median, 5.00 [IQR, 4.00-5.00] vs 3.00 [IQR, 2.00-4.00]; <i>P</i> < .01). In two (4%) of 47 patients with nondiagnostic diastolic image quality due to arrhythmia or high heart rate variability, systolic imaging yielded diagnostic image quality. Conclusion Dual-phase 3D mDixon steady-state MRA enabled phase-dependent assessment of vessel dimensions and image quality in pediatric CHD. Additional systolic imaging improved diagnostic adequacy and visualization of selected structures within a single acquisition. <b>Keywords:</b> Pediatric Congenital Heart Disease, Modified Dixon MR Angiography <i>Supplemental material is available for this article.</i> © RSNA, 2026.</p>","PeriodicalId":21168,"journal":{"name":"Radiology. Cardiothoracic imaging","volume":"8 3","pages":"e250395"},"PeriodicalIF":4.5,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147983039","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Iatrogenic Aerogenous Tumor Spread after CT-guided Lung Biopsy.","authors":"Jiachang Chi, Wangrui Liu","doi":"10.1148/ryct.250300","DOIUrl":"10.1148/ryct.250300","url":null,"abstract":"<p><p>Tumor spread through airspaces (STAS) is a recognized pattern of aerogenous invasion; however, the potential role of biopsy-related hemorrhage in contributing to STAS-like dissemination remains uncertain. The authors present a case in which new ground-glass opacities appeared 20 months after CT-guided biopsy and were spatially aligned with the immediate postbiopsy alveolar hemorrhage tract. Biopsy of these lesions revealed minimally invasive adenocarcinoma with an epidermal growth factor receptor exon 19 deletion identical to that of the index tumor. This radiologic, pathologic, and molecular concordance supports iatrogenic aerogenous spread rather than de novo multifocal disease. This case highlights the importance of hemorrhage-aware biopsy planning, real-time documentation of hemorrhage pathways, and early postbiopsy CT surveillance when hemorrhage extends along bronchovascular bundles. <b>Keywords:</b> Lung Adenocarcinoma, Ground-Glass Opacity, Spread Through Air Spaces, Iatrogenic Complication, CT-Guided Biopsy, Tumor Seeding, Lung Hemorrhage <i>Supplemental material is available for this article.</i> © The Author(s) 2026. Published by the Radiological Society of North America under a CC BY 4.0 license.</p>","PeriodicalId":21168,"journal":{"name":"Radiology. Cardiothoracic imaging","volume":"8 3","pages":"e250300"},"PeriodicalIF":4.5,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147982984","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Uncertainty-Aware, End-to-End Deep Learning for Functional Lung MRI Quantification Using ¹²⁹Xe and ¹H MRI.","authors":"Joshua R Astley, Helen Marshall, Laurie J Smith, Alberto M Biancardi, Martin Brook, Guilhem J Collier, Paul J C Hughes, Laura C Saunders, Jim M Wild, Bilal A Tahir","doi":"10.1148/ryct.250371","DOIUrl":"https://doi.org/10.1148/ryct.250371","url":null,"abstract":"<p><p>Purpose To develop an automatic end-to-end deep learning pipeline for predicting the ventilation defect percentage (VDP) from coregistered functional hyperpolarized xenon 129 (¹²⁹Xe) MRI and structural proton (¹H) MRI scans without manual intervention. Materials and Methods In this retrospective study (2015-2024), ¹²⁹Xe MRI and ¹H MRI scans from healthy participants and patients with a range of pulmonary diseases were used to predict VDP and its associated prediction confidence via an uncertainty-aware convolutional neural network framework. Monte Carlo dropout was used to quantify model uncertainty. Model robustness was assessed using test-time augmentation to simulate test-retest repeatability. The proposed approach was evaluated on a stratified testing set via the median absolute error. Results The dataset comprised 574 paired ¹²⁹Xe MRI and ¹H MRI scans from 47 healthy participants (mean ± SD age, 28.3 years ± 17.3; 28 female participants) and 527 patients with a range of pulmonary pathologies (mean ± SD age, 44.9 years ± 21.9; 295 female patients). The proposed framework produced a median absolute error of 1.01% (IQR, 0.49-2.47) VDP compared with manually corrected, segmentation-derived VDPs; no evidence of difference was found (<i>P</i> = .70). Twenty Monte Carlo dropout iterations were completed, producing VDP prediction distributions that were subsequently clustered into confidence groupings. The proposed approach demonstrated clinical classification accuracy of 91% (95% CI: 68, 94; 32 of 35). Conclusion An uncertainty-aware, end-to-end deep learning approach enabled accurate prediction of VDP without manual segmentation, with performance comparable to segmentation-based methods and quantification of prediction uncertainty. <b>Keywords:</b> Functional Imaging, Lung, Multi-Modal, MRI, Uncertainty-Aware <i>Supplemental material is available for this article.</i> © RSNA, 2026.</p>","PeriodicalId":21168,"journal":{"name":"Radiology. Cardiothoracic imaging","volume":"8 3","pages":"e250371"},"PeriodicalIF":4.5,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148272755","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Radiologic Assessment of Intramural Esophageal Dissection.","authors":"Krunoslav Budimir, Ivana Kuhtić, Pave Markoš, Maja Hrabak-Paar","doi":"10.1148/ryct.250554","DOIUrl":"https://doi.org/10.1148/ryct.250554","url":null,"abstract":"","PeriodicalId":21168,"journal":{"name":"Radiology. Cardiothoracic imaging","volume":"8 3","pages":"e250554"},"PeriodicalIF":4.2,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148219624","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Feasibility of Free-breathing Deep Learning-reconstructed Single-Shot Cine MRI in Participants with Arrhythmia: Comparison with Conventional Segmented Cine MRI.","authors":"Nan Zhang, Fan Du, Caizhong Chen, Xiuzheng Yue, Mengsu Zeng, Hang Jin","doi":"10.1148/ryct.250298","DOIUrl":"10.1148/ryct.250298","url":null,"abstract":"<p><p>Purpose To evaluate the feasibility of retrospective electrocardiographically (ECG) gated single-shot cine using deep learning-enhanced compressed sensing (AI-CS) versus conventional balanced steady-state free precession (bSSFP) cine, focusing on left ventricular (LV) structure and function. Materials and Methods Between September 1, 2023, and September 28, 2024, participants (including those with suspected arrhythmias) were prospectively recruited to undergo short-axis cine imaging with both bSSFP and AI-CS single-shot sequences on a 1.5-T scanner. LV volumetric parameters (LV end-diastolic volume, end-systolic volume, stroke volume, ejection fraction, and mass) and strain parameters (peak strain in the radial, longitudinal, and circumferential directions and the time to peak strain SD) were measured and compared using Wilcoxon signed rank tests. Results Among 25 healthy volunteers (mean age, 37.88 years ± 16.76 [SD]; 18 female) and 45 participants with suspected arrhythmia (mean age, 53.21 years ± 15.45; 20 female), the AI-CS single-shot cine had better image quality compared with bSSFP cine, particularly in participants with arrhythmia (European Cardiovascular Magnetic Resonance Registry score: 0.32 ± 0.68 for bSSFP cine vs 0.05 ± 0.22 for AI-CS single-shot cine; <i>P</i> < .001), with fewer mistrigger events and cardiac motion artifacts. AI-CS showed good to excellent agreement with bSSFP for biventricular volume and LV mass measurements and provided comparable ejection fraction values to those at echocardiography in cases in which bSSFP failed (37.50% ± 5.28 vs 31.70% ± 6.43; <i>z</i> = -1.864; <i>P</i> = .06). Scan time was significantly reduced with AI-CS (10 seconds ± 2 vs 132 seconds ± 8; <i>P</i> < .001). Conclusion AI-CS single-shot cine demonstrated greater image quality and clinical feasibility compared with bFFSP cine in healthy participants and participants with suspected arrhythmias. <b>Keywords:</b> Artificial Intelligence-assisted Compressed SENSE, Arrhythmias, Left Ventricular Structure, Left Ventricular Function, Cardiac MRI, Balanced Steady-State Free Precession Cine Sequences <i>Supplemental material is available for this article.</i> © RSNA, 2026.</p>","PeriodicalId":21168,"journal":{"name":"Radiology. Cardiothoracic imaging","volume":"8 2","pages":"e250298"},"PeriodicalIF":4.5,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147514392","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"<i>Pneumocystis jirovecii</i> Pneumonia: Typical and Atypical CT Findings in a Large Cohort of Immunocompromised Patients with and without HIV.","authors":"Aaron M Pulsipher, Kealy Ham, Holenarasipur R Vikram, Kyle J Henry, Emily R Thompson, Michael B Gotway","doi":"10.1148/ryct.250171","DOIUrl":"10.1148/ryct.250171","url":null,"abstract":"<p><p><i>Pneumocystis jirovecii</i>, an atypical fungus, is a recognized cause of severe pneumonia in individuals with HIV. However, <i>P jirovecii</i> pneumonia (PJP) is now increasingly encountered among immunocompromised patients without HIV, including those receiving treatment for hematologic or solid organ malignancies, solid organ or bone marrow transplant recipients, patients with autoimmune diseases treated with immunosuppressive therapy, and patients requiring prolonged corticosteroid therapy; many of whom may not receive prophylaxis. This article summarizes characteristic and atypical chest CT findings of PJP in immunocompromised patients to support early recognition. Diffuse pulmonary ground-glass opacity (GGO) is the most common chest CT manifestation and should strongly suggest PJP in immunocompromised patients, although variant patterns, particularly peripheral and upper lobe predominant distributions, are frequently observed; midlung and basilar distributions are rare. Consolidation is also common, especially in more severely ill patients, but neither GGO nor consolidation distribution reliably differs across underlying causes of immunocompromise. Interlobular septal thickening and mosaic attenuation are additional CT findings, and, although lymphadenopathy and pleural effusion may occur, these often reflect the underlying condition causing immunocompromise or comorbid disease rather than PJP itself. Cyst formation may be observed and can predispose to pneumothorax. CT abnormalities often improve quickly after initiation of PJP-directed therapy, but not in all patients. Atypical CT manifestations, including nodules (with solid nodules potentially representing \"granulomatous\" PJP), masses, and cavitation, should raise suspicion for coinfection or other superimposed processes. Complex or mixed CT patterns in patients with PJP may result from the disease predisposing to immunocompromise or from its treatment, underscoring the importance of clinical context in interpreting chest CT findings. <b>Keywords:</b> CT, Pulmonary, Thorax, Lung, Infection, Transplantation, <i>Pneumocystis jirovecii</i> Pneumonia, Immunosuppression, Ground-Glass Opacity <i>Supplemental material is available for this article.</i> © RSNA, 2026.</p>","PeriodicalId":21168,"journal":{"name":"Radiology. Cardiothoracic imaging","volume":"8 2","pages":"e250171"},"PeriodicalIF":4.5,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147639632","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}