Stress assessment using multimodal learning in an augmented reality environment

复杂的儿科手术前的焦虑或压力对孩子的健康有害。它还可能延长恢复期，从而增加医疗保健系统的成本。沙夫茨伯里开发二公司正在开发 ** 一个积极的分心娱乐系统（PDES），可以动态地提供游戏/媒体内容，以 ** 减轻用户的压力，最大限度地积极的结果。它可以作为一种手段，非药物 ** 焦虑减少儿童在一个复杂的儿科程序。PDES系统根据用户当前的压力水平动态调整 ** 媒体内容（例如，交互式虚拟现实（VR）游戏），该压力水平可以通过心率、EEG等生物医学信号进行评估。目前，沙夫茨伯里正在研究如何 ** 将当前系统扩展到增强现实（AR）头戴设备中，如微软Hololens，这将 ** 提供与多个用户开发协作体验的机会。此类设备上的摄像头输入还提供了将行为信号（例如用户的手势，面部表情）与生物医学信号结合在一起的机会，以进一步加强压力水平和参与评估。在 ** 这个研究项目中，我们将采用最先进的机器学习技术来融合多模态 **（行为和生物医学）信号，以准确评估用户的压力水平。我们提出的系统将 ** 依赖于非常少的训练数据，由于新的信号变换和迁移学习 ** 技术的实施。该系统将与沙夫茨伯里的PDES完全集成，以便游戏/媒体内容可以动态调整，以减少医疗程序前的压力。

Anxiety or stress before a complex pediatric procedure is detrimental to a child's health. It also adds cost to the**healthcare system by potentially prolonging the recovery period. Shaftesbury Development II Inc is developing**a a Positive Distraction Entertainment System (PDES) that can dynamically deliver game/media content to**alleviate the user's stress for maximum positive outcome. It can be used as a mean for non-pharmacological**anxiety reduction in children before a complex pediatric procedure. The PDES system dynamically adjusts the**media content (e.g. an interactive Virtual Reality (VR) game) based on the user's current stress level which can**be assessed through biomedical signals such as heart-rate, EEG. Currently, Shaftesbury is investigating how to**extend the current system into Augmented Reality (AR) headsets such as Microsoft Hololens, which will**provide the opportunity to develop collaborative experiences with multiple users. The camera input on such**devices also opens up the opportunity to incorporate behavioral signals (e.g. user's hand gesture, facial**expression) alongside biomedical signals to further strengthen the stress level and engagement assessment. In**this research project, we will employ state-of-the-art machine learning techniques to fuse multimodal**(behavioral and biomedical) signals to accurately assess the stress level of a user. Our proposed system will**depend on very little training data due to implementation of novel signal transformation and transfer learning**techniques. The system will be fully integrated with Shaftesbury's PDES so that game/media content can be**adjusted dynamically to reduce stress before a medical procedure.