云端融合的神经系统疾病多通道辅助诊断研究

发布时间：2019-01-15 22:00

【摘要】：对于神经系统疾病,其主要的生理特性可以反映在患者日常的人机交互行为中.对交互过程中产生的相关生理信息进行获取和分析,有助于神经系统疾病辅助诊断与早期预警.传统的神经系统疾病自动检测系统常常只分析单一交互通道中包含的信息,会造成其他通道中包含的重要信息的缺失.为此,本文提出了一个云端融合的神经系统疾病多通道辅助诊断系统.首先,提出一种基于多通道信息的疾病自动诊断方法,获取并分析了用户在多个交互通道中释放的生理信息,并将结果进行融合,提升自动诊断的精确性与鲁棒性;其次,提出了一个云端融合的框架,将从不同地区、不同时刻提取的用户生理信息在云端存储,降低了数据采集的地域限制与时间限制;再次,基于云计算强大的运算能力,系统亦可以实时地对用户在自然交互过程中产生的多通道生理信息进行分析,做出实时、精确的自动诊断;最后,通过语音与笔的交互混合诊断系统实例,验证基于云端融合的神经系统疾病多通道辅助诊断方法的有效性.
[Abstract]:For nervous system diseases, its main physiological characteristics can be reflected in the patient's daily human-machine interaction behavior. The acquisition and analysis of the physiological information generated in the process of interaction is helpful to the diagnosis and early warning of nervous system diseases. Traditional automatic detection systems for neurological diseases often only analyze the information contained in a single interactive channel, which will lead to the absence of important information contained in other channels. Therefore, a cloud fusion multi-channel auxiliary diagnosis system for nervous system diseases is proposed in this paper. Firstly, an automatic disease diagnosis method based on multi-channel information is proposed. The physiological information released by users in multiple interactive channels is obtained and analyzed, and the results are fused to improve the accuracy and robustness of automatic diagnosis. Secondly, a framework of cloud fusion is proposed, in which the physiological information extracted from different regions and different times is stored in the cloud, which reduces the regional and time limitation of data collection. Thirdly, based on the powerful computing power of cloud computing, the system can also analyze the multi-channel physiological information generated by users in the process of natural interaction in real time and make real-time and accurate automatic diagnosis. Finally, the effectiveness of the multi-channel auxiliary diagnosis method for neurological diseases based on cloud fusion is verified by an example of the interactive hybrid diagnosis system of voice and pen.
【作者单位】：中国科学院软件研究所人机交互北京市重点实验室;中国科学院软件研究所计算机科学国家重点实验室;
【基金】：国家重点研发计划(批准号:2016YFB1001405) 国家自然科学基金(批准号:61232013,61422212) 中国科学院前沿科学重点研究计划(批准号:QYZDY-SSW-JSC041)资助项目
【分类号】：R741.04;TP399-C8

【相似文献】