Browsing by Subject "Computer Simulation"
Now showing 1 - 3 of 3
- Results Per Page
- Sort Options
Item Clinical Applications of a Computational Voice Simulator(2016-01-19) Rao, Ashwin; Mau, TedBACKGROUND: Given input vocal fold tissue geometries and mechanical properties, a simulator can produce a virtual voice of a defined pitch and sound pressure level. A simulator gives us a tool to systematically investigate the effects of changes to vocal fold geometry or tissue properties on the voice. OBJECTIVE:To create a user-friendly graphical interface for the NCVS voice simulator, and to use it to determine gender differences in vocal fold tissue properties. METHODS: Graphical user interface (GUI): A GUI was programmed in MATLAB, with windows that separately allow user control of (1) vocal fold tissue geometries and mechanical properties, (2) ranges of inputs for "brute force" simulations in which a range of properties are investigated, and (3) optimization parameters that allow "smart" simulations with target pitch and sound pressure levels. Geometric and tissue properties: A literature search was performed to determine reasonable values for vocal fold superficial layer, the vocal ligament, and vocalis muscle to use in the simulator. Gender differences in tissue properties: Two groups of simulations were carried out, one using typical female geometries with the average female speaking voice frequency of 200 Hz, the other using typical male geometries and 100 Hz as the desired target. Assessment of outcome: Each group of simulations comprised of several thousand simulations, with each simulation corresponding to a specific combination of frequency, sound pressure level, and corresponding tissue properties that generated those specific voice qualities. Clusters of solutions were identified with cluster analysis. RESULTS: Solutions visualized in 2D plots showing a strong linear dependence of voice frequency on the longitudinal shear modulus of the vocal ligament, consistent with the general understanding of vocal physiology. There was no clear dependence of frequency on the other tissue properties. Cluster analysis for the male simulation showed two groups of solutions. The first group had the superficial layer longitudinal shear moduli (μ'1) roughly equal to that of the muscle (μ'3). The second group had a ratio of μ'1/μ'3 between 10-20. This indicates that, for a male voice output of 100 Hz, tissue parameters are most optimal if the superficial layer and muscle have similar stiffness, or if the superficial layer is about 10-20 times stiffer than the muscle. The translayer ratio of μ'2/μ'3 also showed two main clusters: one close to 1 and another at a ratio of around 100. This indicated that, for an average male voice, the vocal ligament and muscle should have comparable stiffness, or the ligament is about 100 times stiffer than the muscle. On the other hand, cluster analysis for the female simulation only showed one solution group, with both μ'1/μ'3 and μ'2/μ'3 ratios of around 1. This indicated that, for a female voice output around 100 Hz, the superficial layer, muscle, and ligament layers all should have comparable stiffness levels. CONCLUSIONS: We have shown that simulations for female geometric parameters calculate different clusters than those for male parameters, given the same target pitch and SPL. By creating a GUI for the NCVS voice simulator, we are now enabling clinician users explore how various morphological defects and changes can affect the human voice, ultimately leading to improved clinical and surgical outcomes.Item A Developmental Algorithm for Synapse-Specific Wiring of the Drosophila Visual Map(2017-08-11) Agi, Egemen; Terman, Jonathan R.; Hiesinger, Peter Robin; Krämer, Helmut; Huber, Kimberly M.During brain development, genetic information and environmental input drive neural circuit assembly that requires matching of correct pre- and post-synaptic partners. In cases when environmental input has no instructive role in synaptic partner selection, genetic information alone must suffice to specify synapses in neural circuits. However, how a limited amount of genetic information is translated into developmental algorithms for synapse specification is unclear. A major thrust of the field has been the quest to identify guidance cues and molecular matchmaking codes underlying brain wiring. In this work, I present a complementary approach, in which the characterization of the developmental algorithm based on simple rules is the primary focus, and the molecules executing these rules secondary. I propose that simple rules underlying developmental algorithms can be sufficient to establish seemingly complex wiring diagrams without an elaborate matchmaking code between synaptic partners. I used Drosophila visual map, which is a genetically encoded neural circuit, as a model system to test my hypothesis. During visual map formation, around 4800 photoreceptors simultaneously project to their correct target layer 'lamina' in the brain to find their correct synaptic partners. I developed a 2-photon microscopy-based, intravital imaging technique with which I could observe the development of individual photoreceptor growth cones at the spatiotemporal resolution of filopodial dynamics over 24 hours during visual map formation. Based on these imaging data, I spearheaded a group effort to formulate and computationally test simple rules that are sufficient for photoreceptors to sort to their correct partners without a requirement for precise matchmaking codes. A key prediction of the model was that the post-synaptic partners may not act as target cues for the pre-synaptic photoreceptors. In the second part of my thesis, I tested this hypothesis by ablating and blocking membrane dynamics of post-synaptic partners. My findings indicate that indeed post-synaptic partners of photoreceptors do not act as target cues for photoreceptors, but are necessary during a preceding step in the developmental algorithm to ensure correct wiring. In brief, results I presented in this work support the idea that correct synaptic partner selection can be achieved through a developmental algorithm based on simple rules that sorts correct cells together prior to synapse formation.Item [Southwestern News](2002-06-05) Carter, Wayne