Bladder cancer accounts for nearly 600,000 new cases and over 200,000 deaths annually worldwide. Approximately 25% of diagnoses correspond to muscle-invasive disease, and up to 50% of patients undergoing radical cystectomy experience recurrence within the first two years, with a 5-year overall survival reaching 50%-60%. Despite the use of neoadjuvant chemotherapy, clinical trials have failed to attain a considerable reduction in the risk of locoregional recurrence, which remains a major clinical challenge due to the limited and largely ineffective salvage treatment options. In this context, adjuvant radiotherapy (ART) has re-emerged as a potential strategy for reducing locoregional recurrence and improving metastasis-free survival, supported by advances in delivery techniques and a reassessment of safety concerns following the BART trial. Simultaneously, perioperative immunotherapy is reshaping the therapeutic landscape of muscle-invasive bladder cancer, with recent studies, such as CheckMate 274 and NIAGARA, establishing a new standard of care. The novelty of this review lies in the integration of the evolving role of ART within the immunotherapy era, with critical examination of its complementary value, toxicity profile and patient selection in light of modern systemic strategies. This narrative review provides an updated synthesis of current evidence and ongoing trials and offers a perspective on how ART can be optimally incorporated into multimodal management of high-risk bladder cancer.
View classification in echocardiographic imaging is essential for accurate analysis and diagnosis. Most automated tools assume that all frames in a sequence belong to the same anatomical view (pure sequence). However, clinical sequences often contain transitions between views (sweep sequence), presenting challenges for reliable measurement and interpretation. While some prior studies briefly acknowledge sweeps [1, 2], none proposes dedicated methods for detecting them. This study elaborates on current artificial intelligence (AI) view classifiers to explore sweep detection in apical sequences by combining frame prediction with temporal signal analysis. A state-of-the-art AI model [2], trained for frame classification on a clean cohort of 3042 pure apical sequences, was applied to 1042 apical echocardiographic sequences to generate view probability curves (Figure 1). The sequences were annotated 5 times as sweep or pure (ground truth: majority voting) by a cohort of 22 sonographers (mean experience 5±3y). Based on a time-series analysis of the probability curves, a sweep detection algorithm was developed using a forward stepwise grid search to optimize thresholds on the following handcraft engineered features: number of class transitions, ratio of dominant class, sequence length and number of distinct classes. The first 5 frames along with frames with low-confidence predictions (highest probability < 0.6) were discarded. The performance is reported via confusion matrix. The agreement between manual annotations was calculated as reference. The model optimization determined two discriminant features, being sweep classified as sequences larger than 15 frames with at least one class transition. The proposed algorithm correctly identified most of the pure sequences (specificity: 83%) and half of the sweep ones (sensitivity: 58%). The inter-annotator agreement was low for sweep sequences, both correctly and incorrectly classified, suggesting that these sequences are intrinsically challenging to interpret and justifying model compromised performance (Figure 2). Misclassified pure sequences had moderate expert agreement, indicating some level of uncertainty, while true pure sequences showed high agreement. The overlapping distributions between sweep and pure sequences across the different features suggest a gradual spectrum, rather than a clear separation, making strict classification inherently difficult (Figure 2). This work establishes a proof of concept for automated sweep detection in echocardiography, a poorly addressed challenge in the literature. While performance remains moderate, it is reasonable given the inherent complexity of the task, often overlooked and highlighted in this study by the limited expert agreement. Finally, the proposed approach, combining AI-based frame classification with temporal signal analysis, is directly transferable to alternative AI frame classifiers.Figure 1 Figure 2
BackgroundPredictive medicine relies on algorithms to determine clinical treatments tailored to each patient’s individual characteristics. Predictive models based on artificial intelligence have shown promise in identifying atrial fibrillation episodes; however, they rarely focus on short-term dynamic prediction. ObjectiveThis study aimed to evaluate the use of an artificial intelligence model and remote monitoring data extracted from pacemaker devices to predict the onset or worsening of arrhythmias in the short term. MethodsThis was a multicenter prospective observational study in which data from 314 patients were analyzed. A total of 65,243 data sequences were collected, of which 55,532 (85.1%) were used to train the algorithm. This model used 31-day records to predict whether the number of arrhythmic episodes would increase, decrease, or remain the same in the following 14 days. ResultsThe sensitivity and specificity of the generated predictions were calculated from 9711 prediction-observation pairs. The global sensitivity was 66.4% (95% CI 64.3%-68.3%), and specificity was 77.4% (95% CI 76.4%-78.4%). For patients with baseline arrhythmia, sensitivity was 76.8% (95% CI 74.6%-78.8%), and specificity was 39.6% (95% CI 35.8%-43.5%). The prediction for patients with no baseline arrhythmia showed a sensitivity of 39% (95% CI 35.1%-43%) and a specificity of 81% (95% CI 80.0%-81.9%). The analysis for the patient subgroup without history of atrial fibrillation (232/314, 73.9%) yielded a 69% sensitivity (95% CI 66.5%-71.5%) and an 80% specificity (95% CI 79.3%-81.3%). ConclusionsThis model was capable of predicting short-term increases or decreases in arrhythmic episodes with reasonable sensitivity and specificity using data collected through remote monitoring of implantable devices. The model’s performance is expected to improve progressively as more data samples become available, including demographic data and clinical records.
