Open Access. Powered by Scholars. Published by Universities.®

Bioimaging and Biomedical Optics Commons™

Open Access. Powered by Scholars. Published by Universities.®

Air Force Institute of Technology

Discipline
Keyword
Publication Year
Publication
Publication Type

Articles 1 - 12 of 12

Full-Text Articles in Bioimaging and Biomedical Optics

An Improved Saliency Map With Trustworthiness For Localizing Abnormalities In Medical Imaging, Nolan C. Skelly Dec 2023

An Improved Saliency Map With Trustworthiness For Localizing Abnormalities In Medical Imaging, Nolan C. Skelly

Theses and Dissertations

Saliency maps are a widely used methodology to make deep learning models more interpretable. They provide post-hoc explanations through identification of the most pertinent areas of an input medical image. These techniques have been assessed based on 1) localization utility, 2) sensitivity to model weight randomization, 3) repeatability, and 4) reproducibility. Ten saliency map techniques will be tested, Grad, Smooth Grad, Integrated Gradients, Smooth Integrated Gradients, XRAI, Grad-CAM, Guided Backpropagation, Guided GradCAM, GradCAM++, and ScoreCAM. The neural networks used to predict and read medical information require a reliable solution to provide medical practitioners intelligible results. Using the information of two, …

Go to article

Monolithic Optofluidic Ring Resonator Lasers Created By Femtosecond Laser Nanofabrication, Hengky Chandrahalim, Qiushu Chen, Ali A. Said, Mark Dugan, Xudong Fan Jan 2015

Monolithic Optofluidic Ring Resonator Lasers Created By Femtosecond Laser Nanofabrication, Hengky Chandrahalim, Qiushu Chen, Ali A. Said, Mark Dugan, Xudong Fan

Faculty Publications

We designed, fabricated, and characterized a monolithically integrated optofluidic ring resonator laser that is mechanically, thermally, and chemically robust. The entire device, including the ring resonator channel and sample delivery microfluidics, was created in a block of fused-silica glass using a 3-dimensional femtosecond laser writing process. The gain medium, composed of Rhodamine 6G (R6G) dissolved in quinoline, was flowed through the ring resonator. Lasing was achieved at a pump threshold of approximately 15 μJ/mm2. Detailed analysis shows that the Q-factor of the optofluidic ring resonator is 3.3 × 104, which is limited by both solvent …

Go to article

Quantitative Object Reconstruction Using Abel Transform Tomography And Mixed Variable Optimization, Kevin R. O'Reilly Mar 2006

Quantitative Object Reconstruction Using Abel Transform Tomography And Mixed Variable Optimization, Kevin R. O'Reilly

Theses and Dissertations

Researchers at the Los Alamos National Laboratory (LANL) are interested in quantitatively reconstructing an object using Abel transform x-ray tomography. Specifically, they obtain a radiograph by x-raying an object and attempt to quantitatively determine the number and types of materials and the thicknesses of each material layer. Their current methodologies either fail to provide a quantitative description of the object or are generally too slow to be useful in practice. As an alternative, the problem is modeled here as a mixed variable programming (MVP) problem, in which some variables are nonnumeric and for which no derivative information is available. The …

Go to article

Evaluation Of Segmentation For Bone Structures In 3d Rendering Of Ultrasound Residual Limb Images, Min C. Baker Dec 1996

Evaluation Of Segmentation For Bone Structures In 3d Rendering Of Ultrasound Residual Limb Images, Min C. Baker

Theses and Dissertations

Prosthetists today widely practice manual socket fitting, which produces subjective, inconsistent results. To address this problem, the Computerized Anthropometry Research and Design (CARD) Laboratory is developing a computer-aided socket design system that acquires ultrasound datasets of an amputee's residual limb, creates a 3D model, and helps identify load- bearing and pressure-relief areas. This research project focuses on providing 3D visualization of a residual limb to support the CARD Laboratory's efforts. Creating the 3D model of the skin and two bone contours requires two major steps: segmentation to identify the objects of interest and a surface tracking algorithm to generate the …

Go to article

Clustered Microcalcification Detection Using Optimized Difference Of Gaussians, Edward M. Ochoa Dec 1996

Clustered Microcalcification Detection Using Optimized Difference Of Gaussians, Edward M. Ochoa

Theses and Dissertations

The objective of this thesis is to design an automated microcalcification detection system to be used as an aid in radiologic mammogram interpretation. This research proposes the following methodology for clustered microcalcification detection. First, preprocess the digitized film mammogram to reduce digitization noise. Second, spatially filter the image with a difference of Gaussians (DoG) kernel. To detect potential microcalcifications, segment the filtered image using global and local thresholding. Next, cluster and index these detections into regions of interest (ROIs). Identify ROIs on the digitized image (or hardcopy printout) for final radiologic diagnosis.

Go to article

Using Kriging To Interpolate Spatially Distributed Volumetric Medical Data, Stephen M. Matechik Dec 1996

Using Kriging To Interpolate Spatially Distributed Volumetric Medical Data, Stephen M. Matechik

Theses and Dissertations

Routine cases in diagnostic radiology require the interpolation of volumetric medical imaging data sets. Inaccurate renditions of interpolated volumes can lead to the misdiagnosis of a patient's condition. It is therefore essential that interpolated modality space estimates accurately portray patient space. Kriging is investigated in this research to interpolate medical imaging volumes. Kriging requires data to be spatially distributed. Therefore, magnetic resonance imaging (MRI) data is shown to exhibit spatially regionalized characteristics such that it can be modeled using regionalized variables and subsequently be interpolated using kriging. A comprehensive, automated, three-dimensional structural analysis of the MRI data is accomplished to …

Go to article

Pulse Coupled Neural Networks For The Segmentation Of Magnetic Resonance Brain Images, Shane L. Abrahamson Dec 1996

Pulse Coupled Neural Networks For The Segmentation Of Magnetic Resonance Brain Images, Shane L. Abrahamson

Theses and Dissertations

This research develops an automated method for segmenting Magnetic Resonance (MR) brain images based on Pulse Coupled Neural Networks (PCNN). MR brain image segmentation has proven difficult, primarily due to scanning artifacts such as interscan and intrascan intensity inhomogeneities. The method developed and presented here uses a PCNN to both filter and segment MR brain images. The technique begins by preprocessing images with a PCNN filter to reduce scanning artifacts. Images are then contrast enhanced via histogram equalization. Finally, a PCNN is used to segment the images to arrive at the final result. Modifications to the original PCNN model are …

Go to article

Detection Of Clustered Microcalcifications Using Wavelets, Donald A. Mccandless Dec 1995

Detection Of Clustered Microcalcifications Using Wavelets, Donald A. Mccandless

Theses and Dissertations

An automated method for detecting microcalcification clusters is presented. The algorithm begins with a digitized mammogram and outputs the center coordinates of regions of interest (ROIs) that contain microcalcification clusters. The method presented uses a 12-tap Least Asymmetric Daubechies (LAD12) wavelet in a tree structured filter bank to increase the signal to noise level of microcalcifications. The signal to noise level gain achieved by the filtering allows subsequent thresholding to eliminate on average 90% of the image from further consideration without eliminating actual microcalcifications 95% of the time. A novel approach to isolating individual calcifications from background tissue through non-stationary …

Go to article

Computer Aided Detection Of Microcalcifications Utilizing Texture Analysis, Ronald C. Dauk Dec 1995

Computer Aided Detection Of Microcalcifications Utilizing Texture Analysis, Ronald C. Dauk

Theses and Dissertations

A comparative study of texture measures for the classification of breast tissue is presented. The texture features investigated include Angular Second Moments, Power Spectrum Analysis and a novel feature, Laws Energy Ratios. The texture study was accomplished as part of the development of a Model Based Vision (MBV) system for the automatic detection of microcalcifications. An overview of the Microcalcification Detection System is presented, which applies image differencing techniques, feature selection methods, and neural networks for locating microcalcification clusters in mammograms. The Power Spectrum Analysis feature set had the best overall performance with an 83% Probability of Detection and an …

Go to article

Computer-Aided Diagnosis Of Mammographic Masses, William E. Polakowski Dec 1995

Computer-Aided Diagnosis Of Mammographic Masses, William E. Polakowski

Theses and Dissertations

A new Model-Based Vision algorithm was developed to find possibly cancerous regions of interest (ROIs) in digitized mammograms and to correctly identify the malignant masses. This work has shown a sensitivity of 92 percent for locating malignant ROIs. The database contained 272 images (12 bit, 1OO microns) with 36 malignant and 53 benign mass images. Of the 53 biopsied benign cases, 74 percent were correctly classified. The Focus of Attention (segmentation) Module algorithm used a physiologically motivated Difference of Gaussians (DoG) filter to highlight mass-like regions in the mammogram. The Index Module labeled the regions by their hypothesized class: large …

Go to article

Three Dimensional Localization Of Lesions From Digitized Mammograms, Amy L. Magnus Dec 1995

Three Dimensional Localization Of Lesions From Digitized Mammograms, Amy L. Magnus

Theses and Dissertations

This thesis describes new algorithms to localize regions-of-interests (ROIs) three dimensionally from a pair of digitized mammograms. This work is intended to add a layer of sophistication to computer aided diagnosis by putting to use the fixed and measurable parameters of mammographic imaging. The fixed parameters are the source to film orientation and the podium-image distance. Measurable parameters are the rotation angle of the x-ray tube from the horizontal and the compression depth (the distance from compression paddle to contact podium distance) at the time of imaging. As an additional benefit, three dimensional localization algorithms alleviate the confusion radiologists may …

Go to article

Evaluation Of Scalar Value Estimation Techniques For 3d Medical Imaging, Rob W. Parrott Dec 1991

Evaluation Of Scalar Value Estimation Techniques For 3d Medical Imaging, Rob W. Parrott

Theses and Dissertations

Scalar value estimation in 3D medical imaging increases data resolution for enhanced renditions and corrects inaccurate surface formations. Accurate estimations are vital because clinical assessment is often aided by examination of 3D medical images. This thesis explores different estimation techniques and introduces the geostatistical estimation technique called kriging to the field of 3D imaging. Kriging theory claims to be the optimal estimator-- better than the standard deterministic methods commonly used. The techniques investigated are linear interpolation, trilinear interpolation, tricubic interpolation, and kriging. This research investigates scalar value estimation in the volume pre-processing operation of slice interpolation and in a surface …

Go to article