Imaezue Gerald C, Tchernichovski Ofer, Goral Mira
Department of Communication Sciences and Disorders, University of South Florida.
Department of Psychology, Hunter College, City University of New York.
Aphasiology. 2024;38(5):838-861. doi: 10.1080/02687038.2023.2239511. Epub 2023 Jul 24.
Previous studies have demonstrated that people with nonfluent aphasia (PWNA) improve their language production after repeating personalized scripts, modeled by speech-language pathologists (SLPs). If PWNA could improve by using their own self-feedback, relying less on external feedback, barriers to aphasia treatment, such as a dearth of clinicians and mobility issues, can be overcome. Here we examine whether PWNA improve their language production through an automated procedure that exposes them to playbacks of their own speech, which are updated recursively, without any feedback from SLPs.
We tested if recursive self-feedback could improve speech fluency in two persons with chronic nonfluent aphasia. We compared two treatments: script production with recursive self-feedback (a new technique) and a non-self-feedback training. We administered the treatments remotely to the participants through their smartphones using two versions of a mobile app we developed. Each participant engaged in each treatment for about three weeks. We estimated clinical improvements of script production through a quantitative trend analysis and nonoverlap of all pairs.
Recursive self-feedback improved speaking rate and speech initiation latency of trained and untrained scripts in both participants. The control (non-self-feedback) training was also effective, but it induced a somewhat weaker improvement in speaking rate, and improved speech initiation latency in only one participant.
Our findings provide preliminary evidence that PWNA can improve their speaking rate and speech initiation latency during production of scripts via fully automated recursive self-feedback. The beneficial effects of recursive self-feedback training suggest that speech unison and repeated exposures to written scripts may be optional ingredients of script-based treatments for aphasia.
先前的研究表明，非流畅性失语症患者（PWNA）在重复由语言病理学家（SLP）制定的个性化脚本后，其语言表达能力有所提高。如果PWNA能够通过利用自身的自我反馈来改善，减少对外部反馈的依赖，那么就可以克服失语症治疗的障碍，如临床医生短缺和行动不便等问题。在此，我们研究PWNA是否能通过一种自动化程序来提高其语言表达能力，该程序让他们接触自己语音的回放，并进行递归更新，且无需SLP的任何反馈。
我们测试了递归自我反馈是否能提高两名慢性非流畅性失语症患者的言语流畅性。我们比较了两种治疗方法：采用递归自我反馈的脚本生成（一种新技术）和非自我反馈训练。我们通过一款我们开发的移动应用的两个版本，通过智能手机对参与者进行远程治疗。每位参与者每种治疗方法都进行了约三周。我们通过定量趋势分析和所有配对的非重叠性来评估脚本生成的临床改善情况。
递归自我反馈提高了两名参与者中经过训练和未经训练脚本的说话速度和言语起始潜伏期。对照（非自我反馈）训练也有效，但在提高说话速度方面效果稍弱，且仅在一名参与者中改善了言语起始潜伏期。
我们的研究结果提供了初步证据，表明PWNA可以通过完全自动化的递归自我反馈在脚本生成过程中提高其说话速度和言语起始潜伏期。递归自我反馈训练的有益效果表明，言语同步和反复接触书面脚本可能是基于脚本的失语症治疗的可选要素。
