AIを活用し神経発達症の子どもに運動促進(Hello, ChatGPT: ‘Please generate exercise prompt for neurodivergent child’)

ad

2025-04-10 ミシガン大学

ミシガン大学の研究チームは、神経多様な子どもたち、特に自閉症スペクトラムの児童が家庭で取り組みやすい運動指示を提供するため、ChatGPTを活用しました。パンデミック中に開発された「InPACT at Home」運動ビデオ132本の指示を、分かりやすいステップ形式に再構成。運動をカテゴリー化し、AIが生成した簡潔な説明で言語や感覚の困難を抱える子どもにも配慮しました。視覚支援や一貫した用語使用など、ユニバーサルデザインも取り入れています。

<関連情報>

神経障害児のための家庭での身体活動の強化:AIとユニバーサルデザインによるInPACT at Homeプログラムの適応
Enhancing home-based physical activity for neurodivergent children: adapting the InPACT at Home program with AI and universal design

Tania Sapre,Haylie L. Miller,Anna Schwartz,Leah R. Ketcheson,Alanna Price,Kerry Winkelseth,Jeanne M. Barcelona,Ronald F. Zernicke,,Rebecca E. Hasson
Frontiers in Physiology  Published:07 January 2025
DOI:https://doi.org/10.3389/fphys.2024.1496114

AIを活用し神経発達症の子どもに運動促進(Hello, ChatGPT: ‘Please generate exercise prompt for neurodivergent child’)

Purpose: While it is common practice for schools across the United States to include neurodivergent children in physical education classes, many programs outside of school—such as those at home or in the community—are not effectively tailored to meet their support needs. This gap contributes to lower levels of physical activity among neurodivergent children. Our objective was to address this issue by systematically adapting the InPACT (Interrupting Prolonged sitting with ACTivity) at Home program to enable neurodivergent children to safely engage in physical activity at home.

Methods: The rapid-cycle research adaptation process involved several key steps: (1) sorting and grouping video content based on different types of skills and exercises (problem exploration); (2) assembling an expert team to guide the development of the instructions (knowledge exploration); and (3) using generative artificial intelligence (AI) to create concise instructions and cue words for each skill/exercise (solution development). These outputs were then fine-tuned by the expert team. The refinements were guided by the Universal Design for Learning (UDL) principle of “Representation,” which acknowledges that learners perceive and understand information in diverse ways.

Results: From the 132 InPACT at Home videos, over 500 activities were identified and categorized into main skill groups: jumping, core, lateral, sport, upper body, lower body, and compound movements. Expert meetings emphasized the importance of the “Three C’s”—consistency, conciseness, and clarity—in crafting instructions, along with the use of simple, elementary sight words. AI was employed to generate and refine prompts like “Provide simplified step-by-step instructions for a jumping jack, suitable for a neurodivergent child” and “Condense the step-by-step instructions for a jumping jack, suitable for a neurodivergent child”.

Discussion: The adaptation of the existing InPACT at Home program was guided by dissemination and implementation science frameworks, aiming to increase equitable access to structured youth physical activity opportunities for neurodivergent children. By incorporating AI and UDL principles, we aim to further enhance the program’s accessibility. Our next steps include evaluating the effectiveness of our program adaptations in encouraging participation in the InPACT at Home program and subsequently increasing physical activity levels among neurodivergent children.

教育
ad
ad
Follow
ad
タイトルとURLをコピーしました