A Smartphone Framework for Mild Cognitive Impairment Assessment

我的课程与新的数学和统计模型、理论、方法和分析工具的发展有关，这些工具有可能导致未来在人类健康方面的应用——比如在与年龄相关的认知健康问题的诊断和治疗方面。具体来说，我的研究是通过严肃游戏构建认知健康评估的计算机技术。我的Discovery Grant （DG）主题是使用机器学习（ML）设计和开发用于轻度认知障碍（MCI）评估的原创手机严肃游戏。******我的研究是一个计算机工程项目，探索在简单的严肃游戏中认知功能、衰退、“记忆失忆”、“套路丧失”和无法制定或记住策略之间的关系。我将制作一个强大的软件平台，以评估玩家识别、应用和记住简单手机游戏统计获胜策略的能力，并开发分析方法来评估记忆衰退。这项工作的长期目标或挑战将是提供具有吸引力和娱乐性的机制，从而在诊断出更典型和明显的认知衰退症状之前，从玩家的表现中推断出认知变化。******目标1：开发和优化手机游戏形式的新技术中介工具，用于MCI评估。***目标2：开发和优化ML引擎和技术，以聚类和分类玩家行为。***目的3：探索额外的传感器融合，为MCI评估提供额外的特征。******总的来说，这些定义研究计划的目标将进一步确立简单的手机游戏在认知评估中的作用。一个强大的框架，将有助于建立我的猜想的可信度，为探索“手机游戏的MCI评估”提供一个平台。这个框架将帮助我们和其他人更好地理解其概念深度以及严肃游戏在认知健康评估方面的局限性。该项目还将为6个新的HQP提供令人兴奋的机会。******使用智能手机游戏进行明确的认知健康评估的概念是独特的，尽管研究人员意识到它的潜力，但尚未在临床实践中使用。我的游戏是不同的，因为它不是简单地计算机化更多基于纸张的评估方法，也不是依赖于像空间认知这样的困难过程，而是简单而熟悉的过程，比如订购卡片，回忆简单的策略。我的游戏是迭代的，游戏速度相当快，可以在20分钟左右的时间内收集到大量的手工数据。最简单的数据本质上是阳性、假阳性、阴性和假阴性，它们构成了ML方法的输入。这里申请的DG部分是为了开发处理这类数据的ML方法。对加拿大的好处是认知健康评估工具的变革性改进。

My program is related to the development of new mathematical and statistical models, theories, methodologies, and analytical tools with potential to lead to future applications in human health—such as in the diagnosis and treatment of age related cognitive health concerns. Specifically, my research is in building computer technology for Cognitive Health Assessment through serious games. My Discovery Grant (DG) theme is designing and developing original mobile serious games for Mild Cognitive Impairment (MCI) assessment using machine learning (ML). ******My research is a computer engineering program exploring relationships between cognitive function, decline, “memory slips”, “loss of set” and failure to develop and/or remember strategies in simple serious games. I will produce a robust software platform to assess a player's ability to recognize, apply, and remember statistically winning strategies for simple mobile games and develop analytics to assess memory decline. The long term objectives or challenges of this work will be to provide engaging and entertaining mechanisms by which cognitive changes may be inferred from player performance before more typical and obvious symptoms of cognitive decline are diagnosable. ******Objective 1: Develop and optimize new technology-mediated tools in the form of mobile games for application to MCI assessment. ***Objective 2: Develop and optimize ML engines and techniques to cluster and classify player behaviors.***Objective 3: Explore additional sensor fusion to provide additional features for MCI assessment.******Overall, these objectives that define the research program will further establish the role simple mobile games can play in cognitive assessment. A robust framework, will help establish the credibility of my conjectures providing a platform to explore “Mobile Games for MCI assessment”. This framework will help us and others better understand its conceptual depth as well as limitations of serious games for cognitive health assessment. The program will also provide exciting opportunities for 6 new HQP.******The notion of using Smartphone games for explicit cognitive health assessment is unique in that is not in used in clinical practice although researchers are aware of its potential. My game is different in that it is not simply computerizing a more paper based assessment method nor is it relying on difficult processes like spatial cognition, but rather simple and familiar processes such as ordering cards, and recalling simple strategies. My game is iterative, with fairly rapid play, and can collect considerable hand by hand data within 20 minutes or so. The simplest data is essentially positives, false positives, negatives, and false negatives, which form the input into ML methods. The DG applied for here is in part to develop the ML methods for processing this type of data. The benefit to Canada is a transformative improvement in tools for cognitive health assessment.