Early identification of atypical infant movement behaviors consistent with underlying neuromotor pathologies can expedite timely enrollment in therapeutic interventions that exploit inherent neuroplasticity to promote recovery. Traditional neuromotor assessments rely on qualitative evaluations performed by specially trained personnel, mostly available in tertiary medical centers or specialized facilities. Such approaches are high in cost, require geographic proximity to advanced healthcare resources, and yield mostly qualitative insight. This paper introduces a simple, low-cost alternative in the form of a technology customized for quantitatively capturing continuous, full-body kinematics of infants during free living conditions at home or in clinical settings while simultaneously recording essential vital signs data. The system consists of a wireless network of small, flexible inertial sensors placed at strategic locations across the body and operated in a wide-bandwidth and time-synchronized fashion. The data serve as the basis for reconstructing three-dimensional motions in avatar form without the need for video recordings and associated privacy concerns, for remote visual assessments by experts. These quantitative measurements can also be presented in graphical format and analyzed with machine-learning techniques, with potential to automate and systematize traditional motor assessments. Clinical implementations with infants at low and at elevated risks for atypical neuromotor development illustrates application of this system in quantitative and semiquantitative assessments of patterns of gross motor skills, along with body temperature, heart rate, and respiratory rate, from long-term and follow-up measurements over a 3-mo period following birth. The engineering aspects are compatible for scaled deployment, with the potential to improve health outcomes for children worldwide via early, pragmatic detection methods.

早期识别与潜在神经运动病理一致的非典型婴儿运动行为可以加快及时注册利用固有神经可塑性促进康复的治疗干预措施。传统的神经运动评估依赖于由受过专门培训的人员进行的定性评估，这些人员大多在三级医疗中心或专业设施中可用。这种方法成本高，需要在地理上靠近先进的医疗资源，并且主要产生定性的洞察力。本文介绍了一种简单、低成本的替代方案，其形式为定制的技术，用于在家庭或临床环境中的自由生活条件下定量捕获婴儿的连续全身运动，同时记录基本生命体征数据。该系统由一个由小型、灵活的惯性传感器组成的无线网络组成，这些传感器放置在整个身体的重要位置，并以宽带和时间同步的方式运行。这些数据作为重建化身形式的 3D 运动的基础，无需视频记录和相关的隐私问题，供专家进行远程视觉评估。这些定量测量也可以以图形格式呈现并使用机器学习技术进行分析，具有使传统运动评估自动化和系统化的潜力。对处于非典型神经运动发育低风险和高风险婴儿的临床实施说明了该系统在粗大运动技能模式的定量和半定量评估中的应用，以及体温、心率、和呼吸频率，来自出生后 3 个月内的长期和后续测量。工程方面与规模部署兼容，有可能通过早期、务实的检测方法改善全球儿童的健康结果。