摘要
傳統語言治療上，音聲障礙患者需要語言治療師個別進行治療，在同一時間內較難同時對多位病人治療，以致於病人需要時程等候復健而形成治療成效不佳。此外音聲治療需要病人花費大量時間練習，容易使病人感到倦怠，導致治療的品質受到影響。
基於上述原因，本研究開發一套音聲復健遊戲系統，藉由遊戲的方式進行復健，用遊戲的方式提高受測者對於語音復健的意願。此系統能利用麥克風收錄一段短時語音後由電腦做即時運算，計算出此段語音的特定參數，如：基頻、音量。此系統擷取上述提到的參數結合遊戲控制器來操作遊戲動畫，利用遊戲動畫給予受測者練習發音的視覺回饋。經由適當的遊戲難度設定以及訓練的結果，治療師可以清楚掌握患者的復健狀況，根據發音表現擬定練習課程。根據治療師制定的訓練清單，患者可以自行在課餘時間利用電腦或是筆記型電腦反覆練習，達到更好的訓練成效。
為了評估本研究遊戲系統的適用性，除了對五位一般發音正常受測者進行實驗外，同時與衛生福利部桃園醫院復健科的語言治療團隊合作，徵求六位分別有嗓音不佳、嗓音異常等症狀的醫院受測者參與為期三週，每週一天，一天進行六次時間共約30分鐘的發音練習實驗 (4次音高遊戲以及2次音量遊戲)。根據不同受測者的發音表現，將給予合適的難度，並隨著實驗的進行逐步調整遊戲的難度，完成實驗後，我們將受測者第一次與最後一次的實驗數據進行比較，判斷受測者經本系統練習後發音進步與否。
為了瞭解醫院受測者與一般受測者在發音上是否有所不同，我們使用曼-惠特尼檢定 (Mann-Whitney U test) 來探討兩群受測者之間的差異。結果發現醫院受測者比一般受測者音量較小且音域廣度較窄，兩者間具有顯著差異 (p<0.05)。我們進一步討論醫院受測者發音的進步情況。比較第一次及最後一次的實驗結果發現平均音量增強為原本大小的12.6% (由70dB上升至78.8dB)，平均發音時間增長為原本的21% (由1.4秒上升至1.7秒)，而且平均發音基頻上限上升了10% (從162.8Hz上升到179.2Hz)，平均發音基頻下限下降約12% (從132Hz下降至116.4Hz)，最後平均音域變為原來的2倍 (由30.8Hz變寬至62.8Hz)，經由上述實驗結果可知，醫院受測者完成實驗後，音量以及音域的表現皆有明顯的進步，這意味著經由本系統練習後，能有效幫助受測者提升發音的表現。

關鍵字：基頻、音量、音聲障礙、音聲治療、嚴肅遊戲。

Abstract
For traditional rehabilitation of speech disorder, speech therapists need to assess and treat patients with voice disorder individually. This individual therapy process leads to a long waiting for patients to have next rehabilitation. In addition, patients have to spend a lot of time practicing, and this practice would bore them and make the efficacy of treatment poor.
Therefore, there is a need to develop a rehabilitation program for voice disorder patients to achieve their goal of recovery. The main purpose of this study was to design a computer game system for voice disorder rehabilitation. By playing game, we motivated the patients for speech rehabilitation. Our game system could record a short speech signal and calculated its volume (in decibel, dB) and fundamental frequency (in Hertz, Hz) immediately. These two parameters were used to control the animated character and provide visual feedback for patients when they practiced. With appropriate level setting, speech therapists would not only diagnosis the condition of treatment but also formulate rehabilitation schedule for the patients from their game performance. Based on each rehabilitation schedule, the patients might practice with the game system on the computer or notebook in their free time to achieve better efficacy of rehabilitation.
To evaluate the feasibility of the system, we collaborated with a speech therapist in Taoyuan General Hospital and design an experiment for five normal subjects (Normal subjects) and six subjects with voice disorder or poor voice (Hospital subjects). This experiment was reviewed by the Internal Review Board (IRB) and carried out for three weeks (one day per week and six times per day with a total time of 30 minutes). Each experiment included 4 pitch and 2 volume games. In terms of their performance, the subjects were trained with our game system in appropriate levels which were evaluated after each test. After finishing all the tests, we evaluated their performance by comparing the results of the first and last experiments.
The performance of the normal and hospital subjects were analyzed with the Mann-Whitney U test. The result showed the improved ranges of pitch and volume for hospital subjects were smaller. There was significant difference between these two groups of subjects (p<0.05). We investigated further to check if the hospital subjects improved their performance after 3 weeks training. We found that the average improvements of the hospital subjects were listed as follow: the volume was increased 10% (from 70 to 78.8 dB); the voicing duration was extended 21% (from 1.4s to 1.7s); the upper fundamental frequency was increased 10% (from 162.8 to 179.2 Hz); lower fundamental was decreased 12 % (from 132 to 116.4 Hz). As a result, the range of pitch was increased two times (from 30.8 Hz to 62.8Hz). These results indicated that all subjects attended this experiment improve their performance. This means that our game system is feasible for patients with voice disorder to improve their voicing after practicing.

Keywords: Fundamental Frequency, Volume, Voice disorder, Voice therapy, Serious game.

英文參考資料:
Kent, R. D. & Read, C. (2002). “The acoustic analysis of speech”, Thomson Learning: Albany, NY, USA.

Dahnial Syauqy. (2014). “Feature Based Scoring on Visible Speech Diagnostic and Rehabilitation System.” National Central University: Taiwan.

Brunella Botte, Claudia Matera, Marta Sponsiello. (2009). "Serious Games between simulation and game. A proposal of taxonomy," Journal of e-Learning and Knowledge Society-English Version , Vol. 5, pp11-21.

Boersma, P., & Weenink, D. (2009). “ Praat – Doing Phonetics by Computer.” The Institute of Phonetic Sciences, University of Amsterdam, Netherland.

Milenkovic, P. (2004). “TF32 [computer Program].” Madison, WI: University of Wisconsin – Madison, Department of Electrical Engineering.

Markus Krause, Jan Smeddinck, Ronald Meyer. (2013). “A Digital Game to Support Voice Treatment for Parkinson’s Disease.” CHI 2013:Changing Perspectives.

Suzanne N. King, Larry Davis, Jeffrey J. Lehman, Bari Hoffman Ruddy. (2011). “A Model for Treating Voice Disorders in School-Age Children within a Video Gaming Environment.” Journal of Voice, Vol. 26, No. 5, pp. 656 – 663.

D Umanski, D Kogovšek, M Ozbič, N O Schiller. (2010). “Development of a voice-based rhythm game for training speech motor skills of children with speech disorders.” Proc. 8th Intl Conf. Disability, Virtual Reality & Associated Technologies

A.A. Navarro-Newball, D. Loaiza , C. Oviedo , A. Castillo , A. Portilla , D. Linares , G. Alvarez. (2014). “Talking to Teo: Video game supported speech therapy.” Entertainment Computing, Vol. 5, issue 4, pp. 401– 412.

Avinash Parnandi, Virendra Karappa, Youngpyo Son, Mostafa Shahin, Jacqueline McKechnie, Kirrie Ballard, Beena Ahmed, Ricardo Gutierrez-Osuna. (2013).” Architecture of an automated therapy tool for childhood apraxia of speech.” Proceedings of the 15th International ACM SIGACCESS Conference on Computers and Accessibility, No.5.

Denice Michelle Edeal and Christina Elke Gildersleeve-Neumann. (2011).” The Importance of Production Frequency in Therapy for Childhood Apraxia of Speech.” American Journal of Speech-Language Pathology, Vol. 20, pp. 95 – 110.

Rupal Patel, Heather Kember, Sara Natale. (2014). “Feasibility of augmenting text with visual prosodic cues to enhance oral reading” Vol. 65, pp.109 – 118.

中文參考資料:
行政院內政部統計處網站，民國103年資料:
http://www.stat.gov.tw/mp.asp?mp=4

王淑娟、高韋樺，民國101年，國內特定型語言障礙相關研究之初探：http://www.ntcu.edu.tw/spc/aspc/6_ebook/pdf/10101/1.pdf

蘇宗柏、陳思遠、王亭貴、王顏和、連倚南，民國99年，復健醫療服務之疾病分類研究：國內某醫學中心近期經驗：
http://www.ntuh.gov.tw/PMR/Lists/List14/Attachments/188/10253009-201012-201101150006-201101150006-229-236.pdf

鄭靜宜，民國100年，語音聲學-說話聲音的科學，心理出版社股份有限公司，台灣台北市。

蘇培豪，民國101年，電腦輔助語言學習中提供個人化發音訓練之對話遊戲架構，國立台灣大學電信工程學研究所，碩士論文。

吳炯德，民國96年，融入語音辨識技術的視訊遊戲設計之研究-以英文單字背誦和記憶之應用為例，南台科技大學多媒體與電腦娛樂科學研究所，碩士論文。

謝秉寰，民國103年，可見式語音診斷與復健系統，國立中央大學電機工程研究所，碩士論文。

林寶貴、錡寶香，民國95年，語言障礙學生輔導手冊：
http://163.32.59.168/2/org/Spedu/03.pdf

林寶貴，民國83年，語言障礙與矯治，五南圖書出版有限公司，台北市，台灣。

王小川，民國98年，語音訊號處理，修定二版，全華圖書股份有限公司，新北市，台灣。

曾進興，民國94年，語言病理學基礎 — 第一卷，心理出版社股份有限公司，台北市，台灣。

楊東華，民國103年，Unity跨平台3D全方位遊戲設計，碁峰資訊股份有限公司，台北市，台灣。

沈明來，民國96年，實用無母數統計學 第二版，九州圖書文物有限公司，台北市，台灣。
李金泉，民國96年，精通SPSS統計分析實務與應用，全華圖書股份有限公司，新北市，台灣。

網頁參考資料:
數位典藏與學習電子報,2015:
http://newsletter.teldap.tw/index.php?language=&lid=
Dr.speech software group 2015: http://www.drspeech.com/index.html
Family Caregiver Alliance, San Francisco,C.A., USA, 2015：https://caregiver.org/
Gerry Beauregard, 2015 : https://gerrybeauregard.wordpress.com/