http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Information processing in the halteres of dipteran insects
Fox, Jessica L University of Washington 2010 해외박사(DDOD)
소속기관이 구독 중이 아닌 경우 오후 4시부터 익일 오전 9시까지 원문보기가 가능합니다.
Sensory systems acquire, process, and transmit information from the environment to the central nervous system. For many sensory organs, the neural signal sent to higher processing centers is a function of the filtering properties of both the sensory structure and the underlying neurons. I examine a specialized mechanoreceptor, the halteres of dipteran insects (flies), to quantitatively analyze the mechanisms by which sensory structures acquire information and convert it to a meaningful neural signal that can be used for control of behavior. The halteres are reduced hindwings with several fields of sensory cells, known as campaniform sensilla, at their bases. The halteres experience multiple forces, including Coriolis forces, as they oscillate and rotate with the body during flight. As these forces deform the haltere cuticle, the neurons associated with the sensilla fire action potentials that signal to multiple targets, including neck motoneurons, wing-steering motoneurons, and other neurons within the central nervous system. In Chapter 1, I introduce the current state of our understanding of the haltere system. The halteres are crucial mechanoreceptors for the control of fly flight, but there have been few studies on the neural encoding of force information. The halteres provide direct, rapid input to wing and neck motoneurons, and behavioral studies suggest that halteres are used to detect the velocity of body rotations. In Chapter 2 (Fox & Daniel, 2008), I analyze the first electrophysiological recordings of single haltere neurons. I describe their responses to step deflections, frequency sweeps, constant-frequency sine waves, and sine waves of increasing amplitude. I find that haltere neurons respond with low latency (4 ms) and high precision to motion stimuli. I also find that they show greater responses in preferred directions of motion, and that a simple model incorporating haltere position and a realistic delay is sufficient to predict the phase of the neural response to sine wave stimuli. In Chapter 3 (Fox et al., 2010), I extend my analysis of haltere function to complex stimuli in the form of band-limited Gaussian noise. I use covariance analysis and singular value decomposition to find that most haltere neurons detect a single feature and its derivative. I develop a model of feature detection that allows us to accurately predict haltere neuronal responses to arbitrary stimuli, and I use this model to predict responses to sinusoidal motions. Each cell in the population from which I recorded has a particular combination of features that endows it with a preferred stimulus phase at which it is most likely to fire. Given this population of cells with distinct phase sensitivities, the haltere system could use a simple place code to encode the shifting phases of strain that occur with varying body rotation speeds during flight. In Chapter 4, I use a three-dimensional finite element model of the haltere to measure local patterns of strain at the base in the approximate location of the campaniform sensilla. I captured high-speed video sequences of crane flies in free flight and modeled these motions with the finite element simulation. I used the spike-prediction model described in Chapter 3 to predict the spike trains resulting from constant velocity yaw rotations and natural yaw rotations at three different points on the haltere base. In doing so, I found that halteres experience and respond to a complex suite of forces in addition to the Coriolis force, and that a place code encoding scheme could be used to detect the body's rotation given the spike trains predicted. In Chapter 5, I present conclusions of this research and some open questions regarding mechanoreception in the control of insect flight.