This chapter presents a framework for designing therapeutic sonic interaction technologies, with a focus on movement sonification: the real-time conversion of bodily motion into sound that serves as feedback during motor rehabilitation. Despite growing evidence for their effectiveness, technologies implementing movement sonification are yet to be systematically adopted as part of clinical practice, potentially due to a lack of standardized development methodologies as well as inadequate integration of clinical stakeholder perspectives into interaction design. The framework addresses these barriers through three interconnected contributions. The first is a conceptual reframing of the design task as the calibration of sonic variability to the perceptual affordances of the listener and the demands of the clinical context. The second is a practical design platform inspired by professional audio mixing workflows, which imposes a structured and learnable signal-flow architecture on the interaction design process and enables rapid iterative exploration. The third is a user-centered development methodology adapted from healthcare intervention science, which grounds design decisions in engagement with the clinicians and patients who will use the resulting systems. The HearWalk biofeedback system for hemiparetic gait rehabilitation illustrates the framework, and the chapter concludes by examining where large language models and AI tools can meaningfully assist each stage of this design process, as well as where human clinical and perceptual expertise remains irreplaceable.

翻译：本章提出一个设计疗愈性声音交互技术的框架，重点关注运动声化：即在运动康复过程中将身体动作实时转化为声音反馈的技术。尽管已有越来越多的证据表明其有效性，但实现运动声化的技术尚未被系统地纳入临床实践，这可能归因于缺乏标准化开发方法，以及临床利益相关者视角在交互设计中的整合不足。该框架通过三个相互关联的贡献来解决这些障碍。首先是概念性重构：将设计任务定义为校准声音变异性，以适配听者的感知可供性及临床情境的需求。其次是实用设计平台：借鉴专业音频混音工作流程，在交互设计过程中引入结构化且可学习的信号流架构，并支持快速迭代探索。第三是以用户为中心的方法论：改编自医疗干预科学，将设计决策建立在与最终使用系统的临床医生和患者的互动基础上。以针对偏瘫步态康复的HearWalk生物反馈系统为例说明该框架，本章最后探讨了大语言模型和AI工具如何在设计过程的每个阶段提供有意义的辅助，以及人类临床与感知专业知识在哪些方面仍不可替代。