急性缺血性脑卒中(AIS)是一种严重影响人类健康相关的疾病,其拥有高发病率和高死亡率,并与吸烟、高脂饮食等不良生活习惯相关,人工智能(AI)比如机器学习(ML)和深度学习(DL),可以实现从临床及辅助检查尤其是成像学检查中提取特征数据,经过算法处理,得出可信结果。近几年AI更多地应用于医院系统的工作中,并成为临床工作及科研项目有力的帮手。本文全面综述了AI预测急性缺血性脑卒中(AIS)患者在经过血管内治疗,尤其是经过血栓切除术治疗后的预后情况,从而实现精准有效的临床管理和护理决策。此外,本文还批判性地评估了现有研究的局限性,并且指出了新的研究方向,最终目标是提高AIS患者的生存率。Acute ischemic stroke (AIS) is a serious human health-related disease that is characterized by elevated morbidity and mortality rates. It is often linked to detrimental lifestyle behaviors, including smoking and high-fat dietary intake. The advent of artificial intelligence (AI), encompassing machine learning (ML) and deep learning (DL) methodologies, facilitates the extraction and analysis of feature data derived from clinical and ancillary assessments, particularly imaging studies. These data are processed through sophisticated algorithms to yield reliable outcomes. In recent years, AI has been increasingly integrated into hospital systems, emerging as a formidable tool in both clinical practice and research initiatives. This paper presents a comprehensive analysis of AI applications in predicting the prognosis of acute ischemic stroke (AIS) patients following endovascular interventions, with a particular focus on thrombectomy procedures. The objective is to enhance the accuracy and efficacy of clinical management and care decision-making processes. Furthermore, the study critically examines the limitations inherent in current research and identifies prospective avenues for future investigation, ultimately aiming to improve the survival outcomes of AIS patients.
背景:膝骨关节炎是一种常见的退行性疾病,不仅严重影响患者的生活质量,同时增加社会医疗负担。早期准确诊断膝骨关节炎对于患者的治疗和预后至关重要,传统的诊断方法不仅主观且耗时,还不能保证稳定的高准确率。目的:开发一种基于深度学习的膝骨关节炎自动诊断方法,利用深度学习网络提高诊断的准确性和效率。方法:在YOLOv8n网络基础上采用Efficient-ViT网络替换YOLOv8n的骨干网络以及增加注意力机制的方法,提出了一种新的网络模型YOLOV8-ViT模型,用于自动识别和分类膝骨关节炎的X射线片图像。实验数据集来自广州中医药大学第三附属医院的5078张膝骨关节炎患者的X射线片图像,由3个影像医师根据Kellgren-Lawrence分级标准采用labelme软件来标注膝关节炎部位并进行分类,采用并集结果。评价指标包括Precision、F1分数、mean average precision(mAP)、Recall、val/box_loss、val/cls_loss和val/dfl_loss。结果与结论:实验结果表明,与YOLOv5n、YOLOv8n、YOLOv9n模型比较,YOLOV8-ViT模型的准确率、IoU阈值为0.5的平均精度(mAP50)、IoU阈值为0.5-0.95的平均精度(mAP50-95)、F1分数和Recall均有所提高,val/box_loss、val/cls_loss和val/dfl_loss分别降低了0.496、0.45和0.523,1.037、0.305和0.728,0.267、0654和0.854,验证了该模型具有较高的检测精度。
暂无评论