A Direct Brain-to-Speech Generator for use in Humans

描述（由申请人提供）：长期目标：这项工作旨在为闭锁的受试者提供一种直接的说话方式。这将允许个人和护理人员之间的快速沟通，以及用语音识别软件控制计算机和控制受试者的环境。这种新发现的沟通能力将为受试者的健康提供一个数量级的改善。

DESCRIPTION (provided by applicant): Long term objectives: This work seeks to provide locked-in subjects with a direct means of producing speech. This will allow rapid communication between the individual and caregivers as well as control of computers with speech recognition software and control of the subject's environment. This newfound ability to communicate will provide an order of magnitude improvement in the subject's well being. Short term aims of this proposal: In the present time frame, we expect to record neural signals from a locked-in subject's speech area, and use these signals to produce a few phonemes. First, we will identify the subject's speech area using functional Magnetic Resonance Imaging. We will then implant the subject with FDA approved Neurotrophic Electrodes with a view to providing a basic switch for computer control and thus enhancing communication using standard software. If the subject wishes, we will then train him or her to drive a speech generator to produce individual phonemes. With training, we expect the subject to control combinations of phonemes or words. Exactly how we produce these phonemes is proprietary information. In subsequent work, we expect to extend the number of neural signals available to the subject so that up to 56 phonemes will be available for control. This should provide a full range of phonemic possibilities for word Production in English speakers patients. There 30,000 ALS patients in the USA and at least 10 times that number worldwide. There are an estimated 50,000-brainstem stroke patients in the USA and well as many other conditions that leave people in the locked-in state.