Radiologia Medica最新文献

Purpose: To develop and evaluate an unsupervised artificial intelligence (AI)-based method for the automated segmentation and quantitative assessment of fluid content on T2 short tau inversion recovery (STIR) magnetic resonance lymphography (MRL) in patients with lymphedema and lipolymphedema.

Methods: A heterogeneous cohort of 20 patients with lymphedema or lipolymphedema was retrospectively selected from our clinical database. Image segmentation was performed using a K-means clustering algorithm based on pixel intensity, with single-slice manual segmentation serving as the reference standard. The clustering algorithm was optimized by maximizing the Dice similarity coefficient up to a fixed threshold. The selected clusters were then saved for transfer learning and applied to the test set, where segmentation performance was evaluated. The procedure was then applied iteratively in a multi-slice setting, enabling three-dimensional analysis. The distribution of edema was visualized using color maps and plots depicting its spatial pattern along the limb. Additionally, an interactive interface was developed to overlay graphs from multiple examinations, facilitating longitudinal comparisons.

Results: The model effectively segmented lymphedema regions, achieving a satisfactory Dice similarity coefficient of at least 0.8 on the training set compared with the manual reference standard. In the test set, the model achieved a Dice score of 0.74 ± 0.05, demonstrating good agreement with manual annotations. The stacked line plots allowed clear visualization of edema distribution along the limb, providing a volumetric representation of edema and enabling the tracking of longitudinal changes in edema patterns over time.

Conclusion: This unsupervised AI-based method shows promise for automated segmentation and quantification of edema in T2-STIR MRL. It offers potential for objective assessment of lymphedema and lipolymphedema, aiding in diagnosis, staging, and treatment monitoring.

Transjugular Intrahepatic Portosystemic Shunt (TIPS) is a well-established procedure for managing portal hypertension and its complications that are refractory to medical and endoscopic therapy. However, complex cases may render the conventional approach ineffective. We describe the cross-sight (CS) technique, which combines ultrasound and fluoroscopic guidance to allow simultaneous puncture of the hepatic and portal veins through a single transhepatic needle pass. Following portal access, stent graft deployment is performed via the jugular route, while the percutaneous tract is closed with a vascular plug to minimize bleeding risk. In our experience, this approach enabled safe and effective shunt creation after failed standard TIPS, with no procedure-related complications. The CS technique provides enhanced procedural control and avoids the limitations of alternative strategies such as DIPS or gun-sight methods. By reducing puncture attempts and access points, it represents a feasible salvage option for complex TIPS cases, with potential immediate application in interventional radiology practice.

Purpose: To evaluate the safety and the efficacy of thoracic endovascular aortic repair (TEVAR) in patients with traumatic thoracic aortic injuries (TTAIs), analyzing the mid-long-term outcomes.

Materials and methods: In this retrospective and monocentric study, 20 patients (46 ± 17.0 years old; mean ± SD) underwent TEVAR for TTAI from February 2012 to April 2023. All patients were subjected to computed tomography angiography (CTA) before discharge; afterward, the follow-up was set up by CTA or magnetic resonance imaging (MRI). Technical success, clinical success, safety, and mid-long-term follow-up were evaluated.

Results: Both technical and clinical success were 100%. No procedure-related death was reported. Safety was 95%. In four (20%) patients the left subclavian artery (LSA) was covered; in one out of these four (25%), revascularization was necessary through carotid-subclavian bypass. In a patient with an anatomic variant of aberrant right subclavian artery (ARSA) a stent placement was required. The follow-up's median duration was 17 ± 79.5 months (median ± IQR) and in three cases there were minimal complications: a stable type IA endoleak (EL) < 1 cm, a minimal irregularity of device's links, and a millimeter bird beak sign.

Conclusions: TEVAR for TTAI was found to be safe (3 minimal complications) and effective both in the short and mid-long term. Patients' adherence to follow-up and contrast-induced kidney damage remains a challenge, but the use of MRI may be a valid alternative, avoiding ionizing radiation and the use of iodinated contrast media.

Purpose: The aim of this study was to develop a new contrast-enhanced ultrasound (CEUS)-based scoring system to distinguish between benign and malignant lesions and to evaluate its potential in obviating unnecessary ultrasound-guided breast biopsies.

Methods: Between November 2019 and June 2023, women with newly diagnosed breast lesions (BI-RADS 4a-c and 5) scheduled for ultrasound-guided biopsy and who underwent CEUS were retrospectively included. Two independent readers (one highly experienced breast radiologist with > 10 years of experience in breast imaging and one in training with > 3 years of experience) assessed all CEUS studies qualitatively in terms of time and intensity of enhancement compared with surrounding tissue, enhancement pattern and size increase after contrast administration, assigning a CEUS score. Further, a combined BI-RADS and CEUS score (CB score) was obtained. Descriptive statistics, AUC (area under the curve), receiver operating characteristic (ROC) analysis, sensitivity and specificity were used to investigate the diagnostic performance of this combined approach.

Results: A total of 294 lesions in 289 women (mean age 59.8 years; age range 23-88) were analyzed. The average lesion size was 16.6 ± 13.7 mm. Of these, 149 were malignant and 146 benign. CB score showed the highest diagnostic performance compared to CEUS score alone (average AUC = 0.935 vs. 0.890, respectively; p < 0.0001). For the more experienced reader, CB score yielded an AUC of 0.947; for the less experienced reader, AUC was 0.922. Finally, CB score would have obviated up to 40% of unnecessary biopsies.

Conclusion: The combined CEUS and BI-RADS scoring system (CB score) improved diagnostic performance for breast cancer classification and showed high accuracy and consistency across readers with different experience, reducing the rate of unnecessary biopsies.

Purpose: Functional and myocardial tissue modifications can occur in treatment-naïve cancer patients, driven by uncontrolled inflammation and neuro-hormonal activation. These mechanisms may result in myocardial alterations that serve as subclinical imaging markers of potential cardiac dysfunction. This study aims to evaluate the role of cardiac magnetic resonance in investigating cancer-associated immune activation and identifying early biomarkers of myocardial damage in treatment-naive cancer patients.

Methods: This prospective study enrolled 100 participants, including 50 treatment-naive cancer patients affected by diffuse large B cell lymphoma and 50 age- and gender-matched healthy controls. All participants underwent comprehensive cardiac magnetic resonance imaging before starting chemotherapy. Correlation analysis was conducted to identify differences in myocardial tissue characterization and functional parameter between the two groups.

Results: Native T1 values were significantly higher in the cancer cohort compared to controls (1007 ± 29 ms vs. 976 ± 29 ms; p < 0.001), while global longitudinal strain was significantly reduced (- 14 ± 2% vs. - 22 ± 6%; p < 0.001). Correlation analysis identified native T1 as an independent predictor of global longitudinal strain impairment (R = - 0.3; p < 0.05). No significant differences in left ventricular mass were observed, suggesting that myocardial tissue changes may precede structural remodeling.

Conclusion: Treatment-naive cancer patients exhibit subclinical myocardial alterations, characterized by elevated native T1 and reduced global longitudinal strain. Cardiac magnetic resonance emerges as a valuable tool for identifying early biomarkers of myocardial dysfunction, offering opportunities for risk stratification and proactive management in cardio-oncology.

Purpose: This study aimed to investigate the effect and potential mechanism of plaque morphology on primary patency after endovascular therapy for femoropopliteal occlusion (FPO).

Materials and methods: A total of 355 FPO patients were divided into four groups based on the chronic total occlusion crossing approach and plaque cap morphology (CTOP) classification. The volume of the plaque cap, calcification score and hemodynamic simulation were calculated on the basis of the CT angiography images, and the effect of cap morphology on primary patency was assessed at 12 months after therapy.

Results: Proximal concave cap lesions (CTOPs I & II) were associated with fewer retrograde punctures and flow-limiting dissections (p < 0.05). Although there was no difference in primary patency or freedom from CD-TLR among the CTOP classifications, the primary patency of proximal concave cap was significantly greater than that of convex cap (81.7 vs. 72.9%; p = 0.04); the freedom from CD-TLR of proximal concave cap was significantly higher than that of convex cap (91.4 vs. 81.4%; p = 0.006). Moreover, the proximal concave cap had a smaller plaque volume than the convex cap (26.00 ± 25.77 mm³ vs. 51.26 ± 43.56 mm³, p < 0.001), and lower severe calcification was detected in the proximal concave cap (4 vs. 28%, p = 0.001). Computer simulation analysis confirmed that the proximal concave cap presented a greater oscillatory shear index (0.35 + 0.20 vs. 0.07 + 0.06, p = 0.002).

Conclusions: The proximal concave plaque cap of the FPO indicates a smaller cap volume and lower degree of severe calcification, which helps to determine endovascular therapy strategy before procedure.

Purpose: To assess the perceived work-life balance, sources of job dissatisfaction, and career aspirations in the new generation of residents and young radiologists.

Materials and methods: A national survey was conducted among young members of the Italian Society of Medical and Interventional Radiology (SIRM) including a total of 15 questions related to job satisfaction, work-life balance, work preferences, career challenges, as well as use and challenges of artificial intelligence. Responses were compared according to the different career stages.

Results: A total of 204 responses were collected, of which 113 (55.4%) reported to be resident, while 91 (44.6%) were practicing radiologists. Senior radiologists reported the lowest job satisfaction (p < 0.001). The most common factors contributing to job dissatisfaction were inadequate compensation (73.0%), inadequate work-life balance (39.2%), and excessive workload (36.3%). Overall, young residents expressed a preference for private practice in future careers (34.0%), public hospitals were preferred by senior residents (38.3%) and young radiologists (42.9%), while senior radiologists expressed a preference for university hospitals (42.9%, p = 0.001). The most commonly reported career challenges were lack of AI training (44.6%), lack of guidance/mentorship (38.2%), and lack of fellowship programme (35.8%). Particularly, most of the responders (37.4%) used AI only rarely, learned AI on their own (31.4%), and felt to be only "somewhat prepared" to face the AI challenges (48.5%).

Conclusion: Early-career professionals face several challenges with difficulty of achieving a sustainable work-life balance and limited job satisfaction.

Objectives: To compare segmental (s-) and subsegmental (ss-) transarterial chemoembolization (TACE) for small hepatocellular carcinoma (HCC), in terms of early complete response (CR), local tumor progression (LTP) and impact on hepatic function (ALBI score).

Materials and methods: A single-center retrospective study was conducted on consecutive patients who underwent s-TACE or ss-TACE as exclusive treatment for small (< 3 cm) HCC between 2021 and 2023. The primary endpoints were 1-month CR and LTP rate during follow-up. The effect of the treatments on hepatic function, as assessed by the ALBI score, was analyzed as a secondary endpoint. Propensity score matching (PSM), based on both baseline and procedural data, was applied to minimize selection bias.

Results: Eighty-nine patients with a total of 114 lesions were enrolled in a per-lesion analysis. No significant differences were found in terms of 1-month CR (74% vs.83%, p = 0.24). Twenty-five transplanted patients were censored at the date of transplantation. During a median follow-up of 16 months (range 7-28 months), ssTACE showed a significantly higher LTP rate compared to sTACE (38% vs. 7.5%, p = 0.001). The results were confirmed after applying PSM, with a hazard ratio of 4 (95% CI, 2-10) for LTP in the ssTACE group. The median time to LTP was similar for both groups (6.7 months vs. 5.6 months, p = 0.92). No differences were observed regarding the worsening of liver function.

Conclusions: Despite similar early CR rates, ssTACE showed a significantly higher LTP rate compared to sTACE, with no significant differences in liver function deterioration between the two approaches.

Purpose: This meta-analysis evaluates the diagnostic performance of computed tomography (CT)-based artificial intelligence (AI) models versus radiologists for preoperative microvascular invasion (MVI) detection in hepatocellular carcinoma (HCC).

Methods: A systematic literature search was conducted in PubMed, Embase, and Web of Science to identify studies published up to February 2025 focusing on the diagnostic accuracy of CT-based AI models for the preoperative detection of MVI in HCC, compared with the diagnostic performance of radiologists. A bivariate random-effects model was employed to calculate the pooled sensitivity, specificity, and area under the curve (AUC), all presented with 95% confidence intervals (CIs). Heterogeneity among studies was assessed using the I² statistic. The methodological quality of included studies was evaluated using a modified version of the Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) tool.

Results: Of 918 identified studies, 32 studies with 3,709 cases were included. For the internal validation set, the pooled sensitivity, specificity, and AUC for detecting MVI in HCC were 0.83 (95% CI 0.79-0.87), 0.81 (95% CI 0.76-0.86), and 0.89 (95% CI 0.86-0.92), respectively. Radiologists achieved a sensitivity of 0.82 (95% CI 0.63-0.93), specificity of 0.65 (95% CI 0.45-0.81), and AUC of 0.80 (95% CI 0.77-0.84).

Conclusions: CT-based AI may have the potential to outperform radiologists in predicting MVI in HCC. However, existing evidence is limited by study heterogeneity and limited number of the direct comparison between AI and radiologists. Prospective multicenter studies are needed to validate its clinical utility.