ASR TO ENHANCE COMPLIANCE WITH ANTIHYPERTENSIVE DRUGS

描述：（改编自申请人的摘要）本发明的目的是 应用程序是一个互动的，计算机介导的，电话监控 一种使用自动语音识别（ASR）来改善血液 提高医生对高血压病人的压力控制 监测和患者对降压药物方案的依从性。的 拟议的系统将整合SOLILOQUY（TM），艾默生和斯特恩的演讲 识别软件，与波士顿大学的电话连接计算机 系统TLC。 SOLILOQUY是一个独立于扬声器，电话兼容 一种能够处理灵活词汇的连续语音识别器， 自然语言语法 TLC，一个互动的门诊病人 监控系统，产生合成语音输出并接受按键音 输入. TLC仅限于数值响应，并分配 是/否按钮，因为使用电话很难拼写单词 键盘。 第一阶段的目标包括创建一个有限的词汇/短语集， 药物副作用;从代表处收集语音数据 人口;整合两个系统以确定可行性;以及 与工作人员进行测试。 在第二阶段，词汇/短语集将 扩展和系统将被移植到“一个盒子”，将测试一个 患者人群。 自2000年以来，还没有人创造过这样的系统。 当时没有技术，也没有人做过必要的语言工作， 对这个群体的研究。 明显的商业应用包括 其他医疗监测环境或电话询问各种 专家系统

DESCRIPTION: (Adapted From The Applicant's Abstract) The goal of this application is an interactive, computer-mediated, telephone monitoring system that uses automatic speech recognition (ASR) to improve blood pressure controlling hypertensive patients by improving physician monitoring and patient compliance with antihypertensive drug regimens. The proposed system will integrate SOLILOQUY (TM), Emerson and Stern's speech recognition software, with Boston University's Telephone-Linked Computer System the TLC. SOLILOQUY is a speaker-independent, telephone-compatible continuous speech recognizer capable of handling flexible vocabularies and natural language phraseology. The TLC, an interactive ambulatory patient monitoring system, produces synthesized speech output and accepts touchtone input. The TLC has been limited to numerical responses and assigned buttons for Yes/No, since it is difficult to spell words using a telephone keypad. Phase I aims include creating a limited vocabulary/phrase set including medication side effects; collecting speech data from a representative population; integrating the two systems to establish feasibility; and live testing with staff. In phase II the vocabulary/phrase sets will be expanded and the system will be ported to "one box" that will be tested one patient population. No one has created such a system before since the technology was not available, and no one has done the necessary linguistic research for this population. Obvious commercial applications include other medical monitoring environments or telephone inquiries to various expert systems.