Radiologia Medica最新文献

Purpose: To differentiate interstitial lung diseases (ILDs) with fibrotic and inflammatory patterns using high-resolution computed tomography (HRCT) and a radiomics-based artificial intelligence (AI) pipeline.

Materials and methods: This single-center study included 84 patients: 50 with idiopathic pulmonary fibrosis (IPF)-representative of fibrotic pattern-and 34 with cellular non-specific interstitial pneumonia (NSIP) secondary to connective tissue disease (CTD)-as an example of mostly inflammatory pattern. For a secondary objective, we analyzed 50 additional patients with COVID-19 pneumonia. We performed semi-automatic segmentation of ILD regions using a deep learning model followed by manual review. From each segmented region, 103 radiomic features were extracted. Classification was performed using an XGBoost model with 1000 bootstrap repetitions and SHapley Additive exPlanations (SHAP) were applied to identify the most predictive features.

Results: The model accurately distinguished a fibrotic ILD pattern from an inflammatory ILD one, achieving an average test set accuracy of 0.91 and AUROC of 0.98. The classification was driven by radiomic features capturing differences in lung morphology, intensity distribution, and textural heterogeneity between the two disease patterns. In differentiating cellular NSIP from COVID-19, the model achieved an average accuracy of 0.89. Inflammatory ILDs exhibited more uniform imaging patterns compared to the greater variability typically observed in viral pneumonia.

Conclusion: Radiomics combined with explainable AI offers promising diagnostic support in distinguishing fibrotic from inflammatory ILD patterns and differentiating inflammatory ILDs from viral pneumonias. This approach could enhance diagnostic precision and provide quantitative support for personalized ILD management.

Purpose: This study aims to compare palliative cTACE, DEB-TACE and DSM-TACE in patients affected by HCC in intermediate BCLC stage in terms of efficacy and patient tolerance.

Materials and methods: Patients treated with palliative TACE were prospectively enrolled in two centers during 9 months. Procedures were performed superselectively in all patients. Inclusion criteria were: HCC diagnosis, intermediate BCLC stage, portal tree patency, preserved hepatic-renal-coagulation functions, palliative procedural aim, follow-up available up to 6-month post-TACE intervention. Exclusion criteria were: previous TACE treatments, alone or in combination with ablation in the same session, ascites, bilirubin > 2mg/dL, age < 18years, bridge to transplant procedural aim, concomitant infectious diseases. Primary endpoint was to compare efficacy and patients tolerance among the 3 different TACE techniques; secondary endpoint was to compare post-procedural complications occurrence.

Results: Seventy patients were included and divided into three groups according to the TACE technique: 24 were treated with cTACE, 25 with DEB-TACE, 21 with DSM-TACE. According to mRECIST criteria at 1-, 3- and 6-month follow-up, DEB-TACE presented better local response rates but without statistically significant differences. Patients treated with DSM-TACE showed significantly better tolerance, considering post-procedural transaminases and INR values together with clinical adverse events occurrence monitored up to 7 days. There were no differences in post-procedural complications and no major complications occurred.

Conclusions: In this study, in patients with intermediate-stage HCC undergoing palliative treatments, no significant differences emerged comparing cTACE, DEB-TACE and DSM-TACE in terms of procedural efficacy; however, patients treated with DSM-TACE showed significant better procedural tolerance.

Cerebrovascular accident is a leading cause of death and disability. Early detection of cerebrovascular diseases is crucial for timely treatment. This study introduces a novel method for the simultaneous generation of color-coded arteriography, venography, and dynamic angiography derived from dynamic contrast-enhanced magnetic resonance angiography and computed tomography perfusion. By realigning source images into time series volume data, this approach enables the classification of five dynamic phases, allowing for the creation of detailed angiographic images and facilitating a comprehensive evaluation of cerebrovascular accidents in the emergency room. The method enables rapid assessment of ischemic strokes, improving patient selection for recanalization therapy, and aids in the early diagnosis of other cerebrovascular diseases, including cerebral venous thrombosis and arteriovenous shunts. We demonstrate the clinical applications of this technique, highlighting its potential to enhance the accuracy and speed of cerebrovascular imaging, making it a valuable first-line diagnostic tool for stroke patients.

Recent advances in molecular genetics have revolutionized the classification of pediatric-type high-grade gliomas in the 2021 World Health Organization central nervous system tumor classification. This narrative review synthesizes current evidence on the following four tumor types: diffuse midline glioma, H3 K27-altered; diffuse hemispheric glioma, H3 G34-mutant; diffuse pediatric-type high-grade glioma, H3-wildtype and IDH-wildtype; and infant-type hemispheric glioma. We conducted a comprehensive literature search for articles published through January 2025. For each tumor type, we analyze characteristic clinical presentations, molecular alterations, conventional and advanced magnetic resonance imaging features, radiological-molecular correlations, and current therapeutic approaches. Emerging radiogenomic approaches utilizing artificial intelligence, including radiomics and deep learning, show promise in identifying imaging biomarkers that correlate with molecular features. This review highlights the importance of integrating radiological and molecular data for accurate diagnosis and treatment planning, while acknowledging limitations in current methodologies and the need for prospective validation in larger cohorts. Understanding these correlations is crucial for advancing personalized treatment strategies for these challenging tumors.

Purpose: Management of colorectal cancer (CRC) is determined by the stage of the disease and molecular features, such as microsatellite instability (MSI). MSI-high/deficient mismatch repair (MSI-H/dMMR) tumors respond better to immunotherapy but poorly to 5-FU-based treatments. With increasing use of neoadjuvant chemotherapy there is interest in developing non-invasive, radiomics models based on preoperative contrast-enhanced CT scans to predict MSI status and support personalized therapy.

Material and methods: Adult patients diagnosed with CRC who underwent pre-treatment staging with contrast-enhanced CT and had known MSI status were retrospectively analyzed. Portal venous phase images were assessed. Two radiologists, blinded to MSI status, manually segmented tumor regions on CT images. Radiomic features and statistical modeling were used to develop a predictive model for identifying the MSI-H phenotype.

Results: Analysis was conducted on 54 adult CRC patients who had undergone staging CT scans with known MSI status. Two different models were built considering different brands of CT machines. Twenty statistically significant radiomic features from the portal venous phase of CT images able to differentiate MSI from microsatellite stable (MSS) patients were selected for each model. LASSO regression was applied, selecting features for model construction. The best model's performance demonstrated an area under the ROC curve of 0.844 (95% CI = 0.73-0.96 DeLong, p < 0,05).

Conclusion: The results demonstrate the potential of the radiomics model as a non-invasive, cost-effective tool for MSI evaluation, guiding CRC therapy. It aids in identifying patients who would benefit from immunotherapy or chemotherapy, supporting the therapeutic shift from postoperative to preoperative treatment.

Metastatic involvement (MB) of the breast from extramammary malignancies is rare, with an incidence of 0.09-1.3% of all breast malignancies. Due to its variable clinical and radiological presentation, MB often mimics primary breast cancer (BC), leading to potential misdiagnosis and impacting treatment decisions. This narrative review analysed MB cases based on dissemination pathways: hematogenous (HM), lymphatic (LM), or direct contiguous (DC) spread. HM was the most frequent, particularly in melanoma, lung, renal, and gastrointestinal carcinomas, presenting as well-circumscribed, non-calcified nodules without axillary lymph node involvement, distinguishing them from BC. LM spread, common in HM malignancies, caused diffuse breast oedema, skin thickening, and a "peau d'orange" appearance, resembling inflammatory BC. DC spread, though rarer, was observed in advanced lung cancer, with infiltrative lesions extending from the chest wall. Multimodal imaging (Mammography (DM), Ultrasound (US), Magnetic Resonance Imaging (MRI), Computer Tomography (CT), and Positron Emission Tomography (PET)) was critical for detecting MB, while histopathological and immunohistochemical analysis confirmed extramammary origin. Due to the rarity and heterogeneity of MB, diagnosis requires a multidisciplinary approach integrating oncological history, imaging, and pathology. Recognising distinct imaging patterns can aid early diagnosis, avoid unnecessary surgery, and guide appropriate systemic therapy based on the primary malignancy. Early identification of the metastatic pattern may influence clinical management decisions and improve patient outcomes.

The male breast is predisposed to be affected by many of the same pathological processes as the female breast is. The diagnosis of male breast pathologies is generally achievable when clinical evaluation is combined with standard breast imaging methods such as mammography and ultrasound. Magnetic resonance imaging is also a valuable tool in diagnosing the main pathologies affecting the male breast, especially for evaluating pre- and post-surgical treatments and follow-up. However, although this technique has been sufficiently regulated and adopted by many breast radiologists for female breast imaging, its application in the diagnosis of male breast pathologies remains limited to a few specialized centers. This article, based on a retrospective analysis of the experience of the University of Verona, explores various aspects of male breast diseases, including benign conditions such as gynecomastia and breast implant ruptures in transgender women as well as malignant entities such as male breast cancer. Emphasis is placed on the distinctive morphological features, enhancement patterns and kinetics observed in male breast lesions on dynamic contrast-enhanced MRI. This article provides a comprehensive overview of the application of MRI in male breast disease assessment, highlighting the potential role of MRI as a complementary tool to traditional breast imaging techniques.

Background: Preserved ratio impaired spirometry (PRISm) and chronic obstructive pulmonary disease (COPD) are progressive respiratory disorders associated with accelerated pulmonary function decline and systemic comorbidities. This multicenter study aimed to develop a three-category classification model that integrates clinical variables with thoracic computed tomography (CT) radiomics to distinguish normal pulmonary function, PRISm, and COPD.

Methods: A total of 1018 participants from three centers (A, B, C) who underwent chest CT and pulmonary function tests (PFTs) within a 2-week interval were retrospectively analyzed. After applying inclusion and exclusion criteria, 797 individuals were included for analysis (Center A: 667 [training/internal test = 534:133]; Centers B, C: 130 external test). CT images were preprocessed via resampling and intensity normalization, followed by semi-automated segmentation of the airway tree and whole lung parenchyma using Mimics Research. PyRadiomics extracted 2436 radiomic features (1218 per region). Feature selection combined maximum relevance minimum redundancy with least absolute shrinkage and selection operator regression, employing tenfold cross-validation. Five models were developed using multinomial logistic regression: (1) clinical model, (2) airway model, (3) lung model, (4) airway fusion model, and (5) lung fusion model. Performance metrics included accuracy, sensitivity, specificity, positive predictive value, negative predictive value, and the area under the receiver operating characteristic curve (AUC), with DeLong tests comparing model efficacy.

Results: 35 airway tree and 48 lung radiomic features were ultimately selected. The best performing model was the lung fusion model, which integrated three clinical predictors (age, gender, and BMI) with selected lung radiomic features. In external test set, it achieved superior performance with AUCs of 0.939 (95% CI 0.898-0.979) for PFT-normal, 0.830 (0.758-0.902) for PRISm, and 0.904 (0.841-0.966) for COPD, with an overall accuracy of 83.59%. DeLong tests indicated that across all three datasets, the lung fusion model outperformed the other four models.

Conclusion: Combining age, gender, BMI, and lung radiomic features significantly improves detection of PRISm and COPD compared to alternative models. These findings underscore the potential of CT-based radiomics for the early identification and risk stratification of abnormal pulmonary function.

Purpose: To evaluate the diagnostic performance of ⁶⁸Ga-DOTATATE PET, contrast-enhanced CT, combined DOTATATE PET/CT, and MRI in detecting neuroendocrine liver metastases on a per-lesion basis and to assess the added value of diagnostic contrast-enhanced CT to PET interpretation as well as influence of lesion size.

Materials and methods: This retrospective study evaluated patients with histologically-confirmed gastroenteropancreatic neuroendocrine tumors who underwent both contrast-enhanced MRI and ⁶⁸Ga-DOTATATE PET/CT within 12 weeks between August 2017 and December 2023. Three readers evaluated in consensus MRI, ⁶⁸Ga-DOTATATE PET, contrast-enhanced CT, and combined PET/CT in separate sessions. Lesions were stratified by size. Diagnostic performance metrics were calculated using generalized estimating equations with reference standard of imaging follow-up or histopathology.

Results: The study included 36 patients (mean age 66.4 ± 10.7 years, 55.6% male) with 720 lesions, of which 582 were metastases. MRI demonstrated superior performance (sensitivity 94.0%, specificity 94.2%) compared to all other modalities (all p < .001). Combined PET/CT showed superior sensitivity (68.1%) compared to PET alone (61.1%) (p < .001). The addition of diagnostic contrast-enhanced CT to PET interpretation identified 41 additional metastases. Size-stratified analysis revealed superior detection of subcentimeter lesions by MRI (sensitivity 83.3% for ≤ 5 mm, 99.6% for 6-10 mm) compared to PET/CT (34.8% and 78.6%, respectively; p < .001). For lesions > 10 mm, both MRI and PET/CT achieved 100% sensitivity, while PET alone reached 94.1% (p = .019).

Conclusion: MRI outperforms ⁶⁸Ga-DOTATATE PET/CT in detecting individual neuroendocrine liver metastases, particularly for subcentimeter lesions. While diagnostic CT improves PET performance, the combination of MRI with ⁶⁸Ga-DOTATATE PET/CT provides the most comprehensive assessment.

Purpose: This study evaluates the efficacy of a previously published [18F]-FDG PET radiomic signature in predicting locoregional failure locations post-reirradiation in head and neck cancer (HNC) patients, using an independent cohort from a different institution.

Materials and methods: Among the 66 patients reirradiated for recurrent HNC at Massachusetts General Hospital between 2012 and 2022, 31 underwent pre-reirradiation PET, constituting the external cohort for this analysis. These patients were characterized using the same radiomic features as the original model (Intensity_histogram_min, Kurtosis, Correlation, and Contrast), projected as a supplementary individual onto the published first principal component, and assigned to one of two groups using the published cutoff. The cutoff was then optimized for the external cohort to determine the loss of performance due to technical or population shifts.

Results: Among the 31 patients, 22 experienced a second locoregional failure, distributed between 12 "in-field" and 10 "outside" recurrences. With the original cutoff, the model achieved a BA of 70% and a positive predictive value (PPV) of 86% for detecting "in-field" recurrences. After recalibrating the cutoff, the model achieved a BA of 78% and a PPV of 89%, close to the 84.5% BA obtained in the original article.

Conclusion: The study validates the ability of the previously established PET radiomic signature to predict "in-field" relapses following reRT with a high PPV. These results support the potential of PET radiomics in identifying patients who may benefit from "in-field" dose escalation in reRT schemes. The model is freely available through the user-friendly LIFEx software.