Assistive technologies for characterizing subjective experience

无法移动和说话的人依靠辅助技术向他人传达他们的主观体验。这些技术将来自自主神经系统和大脑的生理信号转化为与认知活动和情绪状态相关的输出。虽然传达主观体验的系统仍处于初期阶段，但将当前最先进的技术转化为可用于现实世界环境的实用系统需要克服模式识别、信号处理和生理信号分类方面的重大技术挑战。应对这些挑战将能够检测个人的主观体验水平及其认知和情感内容。此外，开发这些技术将促进我们对意识，认知和情感的生理相关性的理解。 为了表征主观经验，本研究提出开发新的技术和算法来分类脑电图（EEG）和自主神经信号。一个由2名博士生和2名硕士生组成的团队，由10名本科暑期学生协助，将i）通过应用图论原理比较从麻醉和非麻醉个体收集的EEG来评估主观体验水平; ii）通过开发可用于认知测试的单次试验事件相关电位分类算法来确定主观体验的认知内容;以及iii）通过从生理信号的大数据集开发精确的独立于用户的情感分类系统来确定主观体验的情感内容。为追求这些目标而开发的算法和软件将导致神经生理学的知识进步、生理信号处理的技术进步以及与我们的医疗保健系统直接相关的商业产品的开发 这一研究方案对加拿大具有相当重要的意义，因为我们的社会需要新的辅助技术来照顾越来越多的因严重残疾、痴呆和创伤而无反应的人。这些技术将使我们能够了解这些人的主观体验，并可能降低护理人员的倦怠和遗弃率，改善治疗，并减少照顾这些弱势群体的重大成本。更直接的是，这项研究为学员提供了一个令人兴奋的计划，以获得具有广泛学术和工业相关性的技术跨学科专业知识，通过生物医学工程职业道路为加拿大社会做出贡献。

Individuals who are unable to move and speak rely on assistive technologies to convey their subjective experience to others. Such technologies translate physiological signals from the autonomic nervous system and the brain into outputs related to cognitive activity and emotional state. While systems to convey subjective experience are in their nascent stages, translating current state-of-the-art into practical systems that can be used in real-world contexts requires overcoming significant technical challenges in pattern recognition, signal processing and physiological signal classification. Addressing these challenges will enable the detection of the level of an individual’s subjective experience, and its cognitive and affective content. Moreover, developing these technologies will advance our understanding of the physiological correlates of consciousness, cognition and emotion. To characterize subjective experience, this research proposes to develop novel techniques and algorithms to classify electroencephalographic (EEG) and autonomic signals. A team of 2 doctoral and 2 masters students, assisted by 10 undergraduate summer students will i) assess levels of subjective experience by applying graph theoretical principles to compare EEG collected from individuals with and without anesthesia; ii) determine the cognitive content of subjective experience by developing single-trial event-related potential classification algorithms that can be used for cognitive testing; and iii) determine the affective content of subjective experience by developing an accurate user-independent emotion classification system from a large dataset of physiological signals. The algorithms and software developed in pursuit of these goals will lead to knowledge advances in neurophysiology, technical advances in physiological signal processing and the development of commercial products that have immediate relevance to our health care system This research program is of considerable importance to Canada, given our society’s need for novel assistive technologies to care for the rising number of unresponsive individuals due to severe disability, dementia and traumatic injury. Such technologies will enable us to understand the subjective experience of these individuals, and will likely reduce the rates of caregiver burnout and abandonment, improve treatment, and reduce the significant cost of caring for this vulnerable population. More immediately, this research provides an exciting program for trainees to gain interdisciplinary expertise in techniques with broad academic and industrial relevance, enabling contributions to Canadian society through biomedical engineering career paths.