Rule based reasoning for anomaly detection and prediction of heart/stroke attack using wearable sensors.

无线和传感器技术的最新进展是可穿戴技术广泛应用的重要驱动力，可穿戴技术越来越多地用于真实的实时监测各种生物医学信号。从简单的心率监测器到先进的EKSO Bionics和OMSignal智能衬衫，每天都有许多可穿戴设备推出。根据市场研究人员的预测，到2025年，可穿戴技术市场将达到700亿美元，拥有30亿个可穿戴传感器。 许多因素促成了这一增长。主要是，将传感和无线通信技术与新的能量收集和存储技术、高效的电源管理电路和低功耗计算相结合，导致形状因素变得越来越灵活、时尚和无形。这些传感器在无缝收集消费者的医疗数据方面变得越来越先进，并能够将其传输到附近的无线接入点。而不是单独观察每个传感器，这些数据必须在很长一段时间内集中存储和分析，以得出有用的结论。 然而，随着设备使用量的增加，数据量也在增加。这些大量的数据需要适当地融合和处理，可能使用云计算环境来开发算法，以准确预测即将发生的健康状况。这将需要准确的量化风险因素和异常检测。可以开发基于规则的推理（RBR）模型用于诸如心脏病发作或中风发作的致命状况的早期预警系统。这个项目将调查统计数据分析方案，建模和分类技术在噪声和缺失数据的存在。该项目预计将生成算法，这些算法将发出警报并可能挽救生命，减少急诊室的等待时间并降低医疗成本。

Recent advances in wireless and sensor technologies have been a vital driver for the widespread adaptation of wearable technology that is increasingly used to monitor various biomedical signals in real time. From simple heart rate monitors to advanced EKSO Bionics and OMSIGNAL smart shirts, numerous wearable devices are being introduced every day. According to market researchers, the wearable technology market to reach $70 billion by 2025 with 3 billion wearable sensors. Number of factors contributes this boost. Mainly, incorporation sensing and wireless communications technologies alongside new energy harvesting and storage techniques, efficient power management circuits and low power computing have resulted in form factors that are increasingly flexible, fashionable and invisible. These sensors are becoming more advanced in seamlessly collecting medical data about the consumers and be able to transmit it to nearby wireless access points. Rather than individually observing each sensor, this data has to be collectively stored and analyzed over a long period to make useful conclusions. However, with increased usage of the devices is also the amount of data. This massive amount of data needs to be appropriately fused and processed possibly using a cloud computing environment for developing algorithms to accurate prognosis of impending health conditions. This will require accurate quantification risk factors and anomaly detection. Use of Rule based reasoning (RBR) models for early warning systems for fatal conditions such as heart attack or stroke attacks can be developed. This project will investigate statistical data analysis schemes, modeling and classification techniques in the presence of noisy and missing data. This project is expected to generate algorithms that will alarm and potentially save lives, reduce waiting time at emergency rooms and reduce healthcare costs.