ABSTRACT
BACKGROUND AND OBJECTIVE: Ulcerative colitis (UC) is a chronic disease characterized by recurrent symptoms and significant morbidity. The exact cause of the disease remains unknown. The selection of current treatment options for ulcerative colitis depends on the severity and location of the disease in each patient. Therefore, developing a fully automated endoscopic images for evaluating UC is crucial for guiding treatment plans and facilitating early prevention efforts. METHODS: We propose a network called ulcerative colitis evaluation based on fine-grained lesion learner and noise suppression gating (UCFNNet). UCFNNet contains three novel modules. Firstly, a fine-grained lesion feature learner (FG-LF Learner) is proposed by integrating local features and a Softmax category prediction (SCP) module to improve the feature accuracy in small lesion areas. Subsequently, a graph convolutional feature combiner (GCFC) is developed to connect features across adjacent convolutional layers and to incorporate short connections between input and output, thereby mitigating feature loss during transmission. Thereafter, a noise suppression gating (NS gating) technique is designed by implementing a grid attention mechanism and a feature gating (FG) module to prioritize significant lesion features and suppress irrelevant and noisy regions in the input feature map. RESULTS: We evaluate the performance of the proposed network on both privately-collected and publicly-available datasets. The evaluation of UC achieves excellent results on privately-collected dataset, with an accuracy (ACC) of 89.57 %, Matthews correlation coefficient (MCC) of 85.52 %, precision of 89.26 %, recall of 89.48 %, and F1-score of 89.78 %. The results are also impressive on publicly-available dataset, with ACC of 85.47 %, MCC of 80.42 %, precision of 85.62 %, recall of 84.00 %, and F1-score of 84.53 %, surpassing the performance of state-of-the-art techniques. CONCLUSION: Our proposed model introduces three innovative algorithm modules, which outperform the current state-of-the-art methods and achieve high ACC and F1-score. This indicates that our method has superior performance compared to traditional machine learning and existing deep methods, which means that our method has good application prospects. Meanwhile, it has been verified that the proposed model demonstrates good interpretability. The source code is available at github.com/YinLeRenNB/UCFNNet.
ABSTRACT
Background: Bifurcation percutaneous coronary intervention (PCI) is associated with higher risk of clinical events. Objectives: This study aimed to determine clinical and lesion features that predict adverse outcomes, and to evaluate the differential prognostic impact of these features in patients undergoing PCI for bifurcation lesions. Methods: We analyzed 5,537 patients from the BIFURCAT (comBined Insights From the Unified RAIN and COBIS bifurcAtion regisTries) registry. The primary outcome was major adverse cardiac events (MACE) at 2-year follow-up; secondary outcomes included hard endpoints (all-cause death, myocardial infarction) and lesion-oriented clinical outcomes (LOCO) (target-vessel myocardial infarction, target lesion revascularization). The least absolute shrinkage and selection operator (LASSO) model was used for feature selection. Results: During the 2-year follow-up period, MACE occurred in 492 patients (8.9%). The LASSO model identified 5 clinical features (old age, chronic renal disease, diabetes mellitus, current smoking, and left ventricular dysfunction) and 4 lesion features (left main disease, proximal main branch disease, side branch disease, and a small main branch diameter) as significant features that predict MACE. A combination of all 9 features improved the predictive value for MACE compared with clinical and lesion features (area under the receiver-operating characteristics curve: 0.657 vs 0.636 vs 0.581; P < 0.001). For secondary endpoints, the clinical features had a higher impact than lesion features on hard endpoints, whereas lesion features had a higher impact than clinical features on LOCO. Conclusions: In bifurcation PCI, 9 features were associated with MACE. Clinical features were predominant predictors for hard endpoints, and lesion features were predominant for predicting LOCO. Clinical and lesion features have distinct values, and both should be considered in bifurcation PCI.