Repository logo

Browsing by Author "Lear, Kevin, advisor"

Now showing 1 - 3 of 3
  • Thumbnail Image
    ItemOpen Access
    Air pollutant source estimation from sensor networks
    (Colorado State University. Libraries, 2024) Thakur, Tanmay, author; Lear, Kevin, advisor; Pezeshki, Ali, committee member; Carter, Ellison, committee member
    A computationally efficient model for the estimation of unknown source parameters using the Gaussian plume model, linear least square optimization, and gradient descent is presented in this work. This thesis discusses results for simulations of a two-dimensional field using advection-diffusion equations underlining the benefits of plume solutions when compared to other methods. The Gaussian plume spread for pollutant concentrations has been studied in this work and modeled in Matlab to estimate the pollutant concentration at various wireless sensor locations. To set up the model simulations, we created a field in Matlab with several pollutant-measuring sensors and one or two pollutant-emitting sources. The forward model estimated the concentration measured at the sensors when the sources emit the pollutants. These pollutants were programmed in Matlab to follow Gaussian plume equations while spreading. The initial work estimated the concentration of the pollutants with varying sensor noise, wind speed, and wind angles. The varying noise affects the sensors' readings whereas the wind speed and wind angle affect the plume shape. The forward results are then applied to solving the inverse problem to determine the possible sources and pollutant emission rates in the presence of additive white Gaussian noise (AWGN). A vector of possible sources within a region of interest is minimized using L2 minimization and gradient descent methods. Initially, the input to the inverse model is random a guess for the source location coordinates. Then, initial values for the source emission rates are calculated using the linear least squares method since the sensor readings are proportional to the source emission rates. The accuracy of this model is calculated by comparing the predicted source locations with the true locations of the sources. The cost function reaches a minimum value when the predicted sensor concentrations are close to the true concentration values. The model continues to minimize the cost function until it remains fairly constant. The inverse model is initially developed for a single source and later developed for two sources. Different configurations for the number of sources and locations of the sensors are considered in the inverse model to evaluate the accuracy. After verifying the inverse algorithm with synthetic data, we then used the algorithm to estimate the source of pollution with real air pollution sensor data collected by Purple Air sensors. For this problem, we extracted data from Purpleair.com from 4 sensors around the Woolsey forest fire area in California in 2018 and used its data as input to the inverse model. The predictions suggested the source was located close to the true high-intensity forest fire in that area. Later, we apply a neural network method to estimate the source parameters and compare estimates of the neural network with the results from the inverse problem using the physical model for the synthetic data. The neural vii model uses sequential neural network techniques for training, testing, and predicting the source parameters. The model was trained with sensor concentration readings, source locations, wind speeds, wind angles, and corresponding source emission rates. The model was tested using the testing data set to compare the predictions with the true source locations and emission rates. The training and testing data were subjected to feature engineering practices to improve the model's accuracy. To improve the accuracy of the model different configurations of activation functions, batch size, and epoch size were used. The neural network model was able to obtain an accuracy above 90% in predicting the source emission rates and source locations. This accuracy varied depending upon the type of configuration used such as single source, multiple sources, number of sensors, noise levels, wind speed, and wind angle used. In the presence of sensor noise, the neural network model was more accurate than the physical inverse model in predicting the source location based on a comparison of R2 scores for fitting the predicted source location to the true source location. Further work on this model's accuracy will help the development of a real-time air quality wireless sensor network application with automatic pollutant source detection.
  • Thumbnail Image
    ItemOpen Access
    Modeling effects of microvilli on somatic signal propagation
    (Colorado State University. Libraries, 2018) Aldohbeyb, Ahmed A., author; Lear, Kevin, advisor; Vigh, Jozsef, committee member; Bailey, Ryan, committee member
    The electrical behavior of small compartments in neurons such as dendritic spines, synaptic terminals, and microvilli has been of interest for decades. Most of these fine structures are found in the dendrite, where most excitatory inputs are received, or in the axon where the action potential is generated and propagates. However, a recent study has shown expression of sodium voltage-gated channels (VGCs) in the soma of intrinsically photosensitive retinal ganglion cells (ipRGCs). Confocal imaging locates these sodium VGCs outside the main soma membrane, which implies that the VGCs occur in structures that protrude from the soma but are too small to be resolved with conventional optical microscopy. An investigator has hypothesized the voltage-gated sodium channels are positioned in microvilli. The microvilli hypothesis raises the question of the role of voltage-gated sodium channels on microvilli and more specifically what effect they would have on propagation of signals in the soma. The nanoscale dimensions of the microvilli, which are much smaller than patch-clamp probes, prevent conventional electrical studies that isolate individual compartments. In the absence of direct, high-spatial resolution measurements, computational models are valuable tools for developing a better understanding of the electrical behavior of the neuronal compartments. Well known models such as Hodgkin–Huxley models and cable theory have been the foundation of many advances in neuroscience. In this work, initial insights about the role of somatic microvilli are being generated using an equivalent circuit model based on the cable equation. For the circuit model, microvilli stubs containing resistor-capacitor networks and sodium channels are treated as branches off the main soma membrane. Circuit models of the soma membrane without microvilli serve as controls. The circuit models were simulated using Simulink. The results show that voltage-gated sodium channels placed on the main soma membrane or on the microvilli increase the amplitude of somatic signals as they propagate to the axon initial segment. Moreover, restriction of the VGCs to the somatic microvilli reduces the probability of misfires originating from spontaneous ion channel opening while still enhancing above threshold depolarizations propagating in the main soma membrane. For comparison, simulations of somatic signal propagation were also performed using the NEURON software as it readily incorporated the Hodgkin and Huxley model, including both sodium and potassium voltage-gated channels. The dendritic input signal was generated using the current clamp technique. The results show that the presence of VGCs on the main soma membrane lower the threshold for triggering the AIS to generate action potential. However, restricting sodium VGCs to the microvilli only did not initiate an action potential at the AIS. The ability of the microvilli Na+ VGCs to serve as excitatory inputs directly to the soma in the absence of the dendritic input was also investigated using NEURON. Using a current clamp, current was injected at the tip of the microvilli and the signal was recorded at the AIS. The results show that the signal at the AIS increases linearly with the injected current. However, the amplitude of the AIS potential was lower than the microvilli signal due to the high microvilli neck resistance. The results support the view that the microvilli act as electrical compartments that attenuate the microvilli VGCs' signals.
  • Thumbnail Image
    ItemOpen Access
    Single cancer cell detection with optofluidic intracavity spectroscopy
    (Colorado State University. Libraries, 2012) Wang, Weina, author; Lear, Kevin, advisor; Chandrasekar, V., committee member; Krapf, Diego, committee member; Reardon, Kenneth, committee member
    The detection of cancer cells is the basis for cancer diagnostics, cancer screening and cancer treatment monitoring. Non-destructive and non-chemical optical methods may help reduce the complexity and cost of related test, making them more available to the public. The label-free technique of optofluidic intracavity spectroscopy (OFIS) uses light transmitted through a cellular body in a microfluidic optical resonator to distinguish different types of cells by their spectral signatures. The OFIS chips are fabricated in the CSU semiconductor clean room and the fabrication process was reported by a previous Ph.D student, Hua Shao. She also did some initial exploration on combining dielectrophoresis (DEP) with the OFIS technique. Since then, some revisions to the fabrication technique have been made to improve the alignment, bonding and sealing of this microfluidic chip. In addition, new DEP electrode designs have been designed and fabricated to further improve the trapping performance of the traps and facilitate automated cell trapping and analysis. Viability tests were carried out to investigate the effect of heating (induced by DEP electrodes) on cells in chips built with borosilicate and sapphire substrates. These experiments used specially designed DEP electrodes that help more accurately control the DEP exposure time and strength. The survival rate of cells out of DEP enabled OFIS system is greatly affected by the substrate type and DEP exposure dose. The OFIS technique has differentiated red and white human blood cells, as well as canine lymphoma and lymphocytes based on their distinctive transmission spectra. Using OFIS chips fabricated with the modified process, OFIS spectra of settled cells from canine hemangiosarcoma (HSA) cell lines and monocytes in peripheral blood mononuclear cells (PBMCs) were collected and analyzed. To quantify the strength of transverse modes in their spectra, a single characteristic parameter was determined for each cell by forming a linear combination of the mean and standard deviation of the transmission spectra over one free spectral range excluding the residual longitudinal peaks of the bare Fabry-PĂ©rot (F-P) cavities filled with cell suspending medium only. The difference in the characteristic parameters of HSA and monocyte samples was highly statistically significant with a p-value as low as 10-6. A receiver operating characteristic (ROC) curve constructed from t-distributions fit to the HSA and monocytes spectra indicates that the cell classification based on their characteristic parameters can achieve 95% sensitivity and 98% specificity simultaneously. Furthermore, some features observed in the spectra of HSA cells motivated a new optical model of the cell loaded F-P cavity. The OFIS spectra of individual cells from canine HSA and canine lymphoma cancer cell lines exhibit relatively uniformly spaced multiple transverse modes repeated in each free spectral range of a microfluidic F-P cavity while similar spectra of healthy canine monocytes and lymphocytes only have up to 2 or no transverse mode peaks. Modeling of the cells as thin lenses allows paraxial Gaussian beam resonator analysis that produces spectral features that quantitatively match the frequencies of transverse modes and qualitatively agree with the trends in maximum transmission of the modes when aperture losses are included. The extracted experimental focal lengths are significantly larger for cancerous cells than for noncancerous cells and can be used as a potential cell malignancy indicator. Furthermore, a thick lens model was developed, allowing manipulation of more parameters related to cell morphology and its location in the cavity. This model was used to interpret experimental results acquired from settled and suspended cells.