Browsing by Subject "Monte Carlo Simulation"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Gaussian Process Regression and Monte Carlo Simulation to Determine VOC Biomarker Concentrations Via Chemiresistive Gas Nanosensors
    (IEEE Xplore, 2021-06) Rivera, Paula Angarita; Woollam, Mark; Siegel, Amanda P.; Agarwal, Mangilal; Mechanical and Energy Engineering, School of Engineering and Technology
    Utilizing chemiresistive gas sensors for volatile organic compound (VOC) detection has been a growing area of investigation in the last decade. VOCs have been extensively studied as potential biomarkers for biomedical applications as they are byproducts of metabolic pathways which are dysregulated by disease. Therefore, sensor arrays have been fabricated in previous studies to detect VOC biomarkers. In the process of testing these sensors, it is highly advantageous to quantify the concentration of the VOC biomarkers with high accuracy to diagnose the disease with high sensitivity and specificity. To investigate, analyze, and understand the relation between the concentrations of the VOC to the sensor resistance response, Gaussian Process (GP) models were implemented to predict the behavior of the data with respect to the resistance when the sensor is exposed to a range of concentrations of VOCs. Additionally, the relation between the concentration and resistance of the sensor was studied to predict the concentration of the VOC when a resistance is obtained. Monte Carlo Simulation Sampling from the GP model was utilized to generate data to further understand the trend. The results demonstrated that the relation between the concentration and resistance is linear. The model was tested with sampling data and its accuracy was evaluated.
  • Thumbnail Image
    Item
    Monte Carlo Simulation to Study Propagation of Light through Biological Tissues
    (2012-09-20) Prabhu Verleker, Akshay; Berbari, Edward J.; Stantz, Keith; Yoshida, Ken
    Photoacoustic Imaging is a non-invasive optical imaging modality used to image biological tissues. In this method, a pulsating laser illuminates a region of tissues to be imaged, which then generates an acoustic wave due to thermal volume expansion. This wave is then sensed using an acoustic sensor such as a piezoelectric transducer and the resultant signal is converted into an imaging using the back projection algorithm. Since different types of tissues have different photo-acoustic properties, this imaging modality can be used for imaging different types of tissues and bodily organ systems. This study aims at quantifying the process of light conversion into the acoustic signal. Light travels through tissues and gets attenuated (scattered or absorbed) or reflected depending on the optical properties of the tissues. The process of light propagation through tissues is studied using Monte Carlo simulation software which predicts the propagation of light through tissues of various shapes and with different optical properties. This simulation gives the resultant energy distribution due to light absorption and scattering on a voxel by voxel basis. The Monte Carlo code alone is not sufficient to validate the photon propagation. The success of the Monte Carlo code depends on accurate prediction of the optical properties of the tissues. It also depends on accurately depicting tissue boundaries and thus the resolution of the imaging space. Hence, a validation algorithm has been designed so as to recover the optical properties of the tissues which are imaged and to successfully validate the simulation results. The accuracy of the validation code is studied for various optical properties and boundary conditions. The results are then compared and validated with real time images obtained from the photoacoustic scanner. The various parameters for the successful validation of Monte Carlo method are studied and presented. This study is then validated using the algorithm to study the conversion of light to sound. Thus it is a significant step in the quantification of the photoacoustic effect so as to accurately predict tissue properties.