|Acquisition Program: || Objective: ||Sound propagation through the human skull will be modeled. The model will include parameters such as skull dimensions, type of head covering, and noise environment that can be input in order to predict bone conduction communication system effectiveness for individuals in various circumstances. The model will be used to improve bone conduction communication systems to a level where they will be able to be used in military environments.
|| Description: ||The perception of sound through bone conduction involves the transmission of sound to the inner ear through the bones of the skull. Bone conduction offers several potential advantages over air conduction (normal hearing through the ears) as a means of communication in military environments. For example, the ears do not have to be covered which allows the surrounding acoustic environment to be monitored, and bone conduction communication systems are less sensitive to ambient noise and may be a better alternative to noise-canceling devices in high noise environments. Several commercial bone conduction microphones and speakers currently are available for civilian applications, but the technology is not yet mature enough to provide high fidelity devices required for military operations across a variety of environments. In particular, speech intelligibility is not very good with the devices (e.g., Acker, Houtsma, and Ahroon, 2005). The sound quality and intelligibility of speech are dependent on the technical parameters of the bone conduction transducers and the placement of the transducer on the head of the communicating person. It is believed that signals transmitted to or received from specific locations on the head will have better speech intelligibility than signals to and from other locations. Fundamental research assessing the nature of vibration and acoustic characteristics of the human skull is necessary in order to improve bone-conduction communication systems.
Computer models that capture the behavior of the skull and vocal tract are needed to predict the regions on the skull most sensitive to mechanisms of both transmission and reception of sound. This research will fill an important gap in our knowledge regarding head vibration patterns and will be used to predict how well individuals will be able to use bone conduction systems based on their personal anthropomorphic characteristics.
|| ||PHASE I: Determine if a model of the acoustic properties of the human head can predict speech intelligibility using bone conduction communication systems. Develop a model that will capture skull vibration patterns in response to external, localized vibration input (i.e., a bone conduction speaker). The vibration input will include a frequency range from about 500 to 6000 Hz at several different intensity levels. The model will provide predictions for ideal and non-ideal bone conduction speaker placement. Using stimuli recorded from an acoustic microphone, the model’s predictions will be tested by measuring speech intelligibility (e.g., the Diagnostic Rhyme Test as specified in ANSI S3.2-1989, R1999) for the predicted ideal speaker placements. Speech intelligibility also will be measured for a non-ideal speaker location. The model’s adequacy will be determined by the degree to which the model predicts speech intelligibility based on the location of the speaker on the skull (i.e., the predicted ideal location results in high degrees of speech intelligibility and the non-ideal location results in poor speech intelligibility). Investigators must obtain IRB approval from the local institution and from the U.S. Army Medical Research and Material Command for all human speech intelligibility testing.
|| ||PHASE II: Include the vocal tract as a parameter in the model developed in Phase I. Map skull vibration patterns in response to vocal tract vibrations. Determine ideal bone conduction microphone placement. Select ideal and non-ideal locations from which to record speech stimuli. Evaluate speech intelligibility for the recorded stimuli. The model’s adequacy will be determined as in Phase I, except listeners will use regular acoustic headphones to listen to the bone conduction microphone stimuli. (In Phase I, listeners will use bone conduction speakers to listen to stimuli recorded with acoustic microphones.) Finally, evaluate the effect of skull anthropomorphics, helmet type (to be provided), and environmental noise on skull vibration patterns. Incorporate as variables to be input by the user of the model. Investigators must obtain IRB approval from the local institution and from the U.S. Army Medical Research and Material Command for all human speech intelligibility testing.
Develop a software program that will allow the user to input variables important to bone conduction effectiveness. The output will include some metric that will indicate if bone conduction speakers/microphones should be used by a particular individual in a particular environment. Evaluation will include assessing how well the output predicts speech intelligibility for the given person/environment.
|| ||PHASE III: Market the software to companies that manufacture bone conduction communication systems. Predicting effectiveness for an individual and improving bone conduction transducers will allow the systems to be used effectively in a variety of civilian and military environments.
|| References: ||
Acker-Mills, B., Houtsma, A., and Ahroon, W. (2005). Speech Intelligibility with Acoustic and Contact Microphones. Paper presented at the RTO HFM Symposium on “New Directions for Improving Audio Effectiveness”,
held in Amersfoort, The Netherlands, 11-13 April 2005, and published in RTO-MP-HFM-123.
American National Standards Institute. Method for measuring the intelligibility of speech over communication systems. 1989; Melville, NY: Standards Secretariat, Acoustical Society of America. ANSI S3.2-1989 (R1999).|
|Keywords: ||Sound propagation, noise, bone conduction, microphone, speaker, speech intelligibility, computer model|