Background: Conventional radiography faces high error rates (3-10%) due to heavy clinical workloads. While AI has emerged as a supportive tool, there is an evidence gap regarding the clinical utility of integrated AI systems in detecting both skeletal and thoracic abnormalities. Objectives: This large-scale, international multicenter study aims to validate the performance of a unified radiographic AI suite across an expanded diagnostic scope while confirming its continued robustness. Methods: A retrospective performance evaluation was conducted using 21,581 adult and pediatric X-rays collected from 20 countries. The reference standard was established through independent review by two expert readers, with adjudication of a third radiologist in cases of discordance. Diagnostic metrics, including Area Under the Curve (AUC), sensitivity, and specificity, were calculated for all 18 pathologies. Subgroup analysis was performed by patients' age, sex, and country of acquisition. Results: For the nine findings within the expanded scope, AUC values exceeded 96.1%, with sensitivity and specificity ranges from 94.5 to 98.8% and 86.6 to 96.1%, respectively. Similarly, for the nine historically validated findings, AUCs remained above 96.1%, with sensitivity and specificity localized between 94.5 and 97.8% and 84.6 and 89.4%, respectively. Consistency was maintained across subgroups. Conclusions: The results confirm the potential of deep learning to transition from narrow, task-specific tools to a unified, high-performance diagnostic system.
Background Over the past two decades, the proportion of very old intensive care patients (VIPs; aged ≥ 80 years) admitted to intensive care units has significantly increased. Despite advances in critical care, outcomes for this vulnerable population remain variable and poorly understood across different eras in medical practice. Objectives This study aimed to compare the one-year outcomes of VIPs admitted to the ICU between 2005–2009 (05–09) and 2015–2019 (15–19), assess shifts in demographics and case mix, and evaluate outcomes for time-sensitive admission diagnoses. Design and Setting Single-center retrospective cohort study conducted in a 12-bed mixed medical-surgical ICU at Sint-Blasius General Hospital, Belgium. Methods All ICU admissions of VIPs during the two periods were analyzed. Patients with missing data or non-retrievable patient files were excluded. For patients with repeated ICU admissions within one year, only data from the last admission were considered. Propensity Score Matching (PSM) was applied to adjust for age, sex, SAPS II score, and APACHE IV admission diagnosis. ICU Length of Stay (LOS) was additionally explored as an indirect surrogate for end-of-life decision-making. Kaplan-Meier survival analysis and statistical tests were used to compare outcomes up to 1 year after ICU admission. Results A total of 885 VIPs were admitted between 05–09 versus 1267 between 15–19. After exclusion, 747 and 1066 ICU admissions from the 05–09 and 15–19 periods, respectively, were analyzed. ICU admissions of VIPs increased among the 15–19 group, with older patients exhibiting lower SAPS II scores. Over time, the top 25 APACHE IV diagnoses showed significant changes. Medical admissions were predominant, whereas planned surgical admissions decreased in 15–19. Despite these shifts, the overall mortality rate remained unchanged. However, there was a notable improvement in ICU LOS in 15–19. PSM was employed to adjust for biases and create two well-balanced cohorts of 506 patients, revealing no significant differences in overall mortality. For time-sensitive conditions, such as AMI, CVA, and sepsis, ICU mortality and LOS improved significantly in 15–19, whereas outcomes for less time-sensitive conditions, such as CHF, COPD exacerbation, and pneumonia, remained unchanged. The reduction in ICU LOS in the later period was primarily driven by shorter ICU stays among survivors (PSM analysis), as suggested by an exploratory surrogate analysis. Conclusion Despite an older and more complex ICU population in 15–19, adjusted survival outcomes remained stable. Improvements in early recognition and standardized treatment protocols have likely contributed to better outcomes for specific acute conditions. However, missing frailty data and residual confounding limit the ability to draw definitive conclusions. Further prospective studies are warranted.
ObjectiveIn two randomized controlled trials, the outcomes of endovascular treatment of complex femoropopliteal arterial lesions were compared with bypass surgery and considered a valid alternative treatment. The aim of this study was to compare both endovascular treatment options with the hypothesis that implantation of heparin-bonded self-expanding covered stents (Viabahn, SECS) or drug-eluting stents (ZilverPTX, DES) are related to similar clinical outcomes at one-year follow-up.MethodsIn a post-hoc analysis, the SuperB trial and Zilverpass databases were merged. Patients in the endovascular treatment arms were included and data was analyzed in an intention-to-treat (ITT) and a per-protocol (PP) fashion. Data included baseline and lesion characteristics, procedural details, and follow-up data. The primary endpoint of this study was primary patency at one-year follow-up. The secondary endpoints were secondary patency, target lesion revascularization (TLR), limb loss, and all-cause mortality.ResultsA total of 176 patients were included; 63 in the SECS arm and 113 in the DES arm. Through 1-year follow-up there were no significant differences in primary patency (ITT 63.4% vs 71.1%: p=0.183 and PP 60.8% vs 71.1%; p=0.100). Secondary patency rates were not significantly different in the ITT analysis (86.5% vs 95.1%; p=0.054), but in the PP analysis, there was a significant difference in favor of the DES group (SECS 85.6% versus DES 95.1%; p=0.038). There was no significant difference in freedom from TLR between groups (79.6% vs 77.0%, p=0.481). No major amputations were performed in the SECS group and two in the DES group (1.8%). Survival rate was 98.2% in the SECS group, and 91.3% in the DES group after one-year follow-up (p=0.106). Based on diagnosis (IC versus CLTI) no differences between IC and CLTI patients were observed in primary, secondary patency and freedom from TLR.ConclusionsTreatment of complex femoropopliteal arterial disease with the heparin-bonded Viabahn endoprosthesis and the Zilver PTX drug-eluting stent are related to similar primary and secondary patency, and TLR rates at one-year, except for secondary patency in the PP analysis. This study further supports the endovascular treatment of long complex lesions in the femoropopliteal artery.
Successful endovascular revascularisation strongly depends on efficient lesion crossing for which a large variety of guidewires (in combination with a variety of catheters or balloons) is available. Hence, knowledge on guidewire types and their characteristics is key for accurate guidewire selection and technique. However, no standardisation of guidewire selection for peripheral arterial disease (PAD) is available, and there is high physician dependent variation in lesion crossing technique, complicating structured education and training. This edutorial provides an overview of guidewire types, properties, applications, and techniques to guide interventionists in optimal guidewire selection and to provide a backbone for endovascular training. A Thin Thread and a Leg Do Not Make an Intervention: Share Your Recipes!European Journal of Vascular and Endovascular SurgeryPreviewIn this issue of the European Journal of Vascular and Endovascular Surgery, Ünlü et al. publish an interesting educational article on a topic that may appear trivial at first glance: guidewire selection and crossing strategies.1 However, it becomes obvious that the choice of proper guidewires and techniques for advanced lesion crossing has not been comprehensively covered by most clinical practice guidelines on vascular diseases, while there are few multidisciplinary expert statements to help guide clinicians. Full-Text PDF
