Bioimaging and Biomedical Optics Commons^™

Robustness Of Ai-Driven Histopathology Under Real-World Adversarial Examples, Ruchik N. Yajnik

Theses

This robustness of histopathology classification models under adversarial and real-world perturbations resembling clinical artifacts is being investigated.

Using whole-slide images from the CAMELYON17 cohort, four representative architectures—ResNet-18, ResNet-50, HIPT-2MLP, and ViT-B/16 —are benchmarked across controlled pixel-level distortions and artifact-like transformations. Adversarial methods include iterative Fast Gradient Sign, Projected Gradient Descent, Salt-and-Pepper noise, and the Adversarial Watermark—Stain Shift (AWSS). Three defense strategies—Randomized Smoothing, Adversarial Training, and an Artifact Detector—are evaluated for their ability to preserve diagnostic accuracy and model reliability. Structured perturbations consistently degrade performance, with transformer-based models showing the greatest sensitivity. The benchmark developed here offers a reproducible framework for …

Untrained Position-Encoded Multilayer Perceptron Network For Structured Illumination Microscopy Reconstruction, Sahil Sharma, Leonidas Zimianitis, Krishnendu Samanta, Balpreet Singh Ahluwalia, Joby Joseph, Dushan N. Wadduwage

Computer Science Faculty Publications

Structured Illumination Microscopy (SIM) enables super-resolution imaging by encoding high-frequency spatial information through patterned light. While traditional Fourier-based reconstruction methods are prone to artifacts under suboptimal conditions, recent deep learning approaches often require large training datasets and lack adaptability across different imaging setups. In this work, we present Position Encoded Multi-Layer Perceptron (PEM) network that leverages implicit neural representations (INRs) and SIM forward-model-driven modeling to reconstruct super-resolved images without any training data. PEM-SIM represents each spatial coordinate as a combination of sinusoidal functions across multiple frequencies, enabling rich encoding of fine spatial detail. A forward model grounded in SIM image …

Susceptibility To High-Fidelity Misinformation: An Eye-Tracking Analysis, Yasasi Abeysinghe, Gavindya Jayawardena, Enkelejda Kasneci, Sampath Jayarathna

Computer Science Faculty Publications

With the rise of online misinformation and AI-generated text, understanding human perception of news truthfulness is critical. In this study, we examine visual attention and cognitive processing using eye-tracking measures as individuals read fake and real news articles sharing nearly identical structure and imagery, differing only in subtle textual changes. Using the public FakeNewsPerception dataset, we analyze advanced gaze measures, including scanpaths, AOI transitions, and luminance-corrected pupil measures, beyond basic gaze features, in relation to news truthfulness and perceived believability. Results show that, given the high fidelity of the fake news, readers exhibited comparable visual scanning patterns, attention allocation across …

On The Scalability Of Anisotropic Mesh Adaptation On Distributed And Shared Memory Architectures For Numerical Approximations, Kevin Mark Garner Jr.

Computer Science Theses & Dissertations

Mesh generation is a critical component in numerical approximations of Partial Differential Equations (PDEs). One such example includes Computational Fluid Dynamics (CFD), as CFD simulations in turn are crucial for applications in many industries, such as personalized healthcare and the design of aerospace vehicles. Generating high quality meshes for large-scale CFD problems presents a significant bottleneck in the CFD workflow. This dissertation proposes “fast,” parallel 3D mesh generation methodologies that are designed to leverage the concurrency offered by emerging High-Performance Computing (HPC) architectures. First, a distributed memory method is presented that integrates a sequential state-of-the-art isotropic, advancing front local reconnection-based …

Design And Development Of A Standalone Digital Holographic Microscope Employing Phase-Driven Reconstruction And Classification For Biomedical Imaging And Optical Diagnostics, Charlotte Kyeremah

Graduate Doctoral Dissertations

Access to advanced biomedical imaging technologies remains a significant challenge in resource-limited settings, especially for early disease detection and monitoring of diseases such as malaria, HIV, and other blood-borne diseases. Although point-of-care (POC) devices have gained popularity in global health, many rely on antibody-based tests, lateral flow strips, or optical readouts that often lack quantitative capabilities, sensitivity to early infections, or versatility in different diagnostic targets. In addition, these systems are typically dependent on disposable reagents or manual interpretation, which limits their effectiveness in remote areas. Digital Holographic Microscopy (DHM) presents a promising alternative as a label-free imaging method capable …

Multimodal Ai-Driven Biomarker For Early Detection Of Cancer Cachexia, Sabeen Ahmed

USF Tampa Graduate Theses and Dissertations

Cancer cachexia is a metabolic syndrome characterized by substantial skeletal muscle loss, impacting cancer patients' survival and quality of life. Despite its clinical significance, early detection remains a challenge due to the lack of standardized diagnostic criteria and the reliance on indirect markers. This work presents an AI-driven approach to enhance cachexia detection and monitoring by integrating multiple deep learning methodologies. We explore transformer architectures for time-series analysis to model sequential medical data, enabling disease prediction and progression modeling. To ensure robust and reliable decision-making in clinical settings, we explore Bayesian deep neural networks for uncertainty estimation. Additionally, we introduce …

Automation To Autonomy: Temporal Dynamics Of Trust And Visual Attention Allocation Did Not Evolve, Tetsuya Sato, Eric Chancey, Yusuke Yamani

Psychology Faculty Publications

Emerging work environments are expected to implement autonomy that performs various functions without human input. Previous works has shown that trust in automation is negatively correlated with visual attention allocation, indicating that trust is a dynamic construct. Moreover, trust in automation and trust in autonomy appears to evolve in similar ways. However, recent work has demonstrated differences between trust in automation and trust in autonomy within Kaber’s (2018) theoretical framework (Sato et al., 2023b). Yet, it is uncertain whether the development of trust and visual attention allocation differs between automation and autonomy. The present study examined the temporal dynamics of …

Adaptive Fusion Neural Networks For Sparse-Angle X-Ray 3d Reconstruction, Shaoyong Hong, Bo Yang, Yan Chen, Hao Quan, Shan Liu, Minyi Tang, Jiawei Tian

Electrical & Computer Engineering Faculty Publications

3D medical image reconstruction has significantly enhanced diagnostic accuracy, yet the reliance on densely sampled projection data remains a major limitation in clinical practice. Sparse-angle X-ray imaging, though safer and faster, poses challenges for accurate volumetric reconstruction due to limited spatial information. This study proposes a 3D reconstruction neural network based on adaptive weight fusion (AdapFusionNet) to achieve high-quality 3D medical image reconstruction from sparse-angle X-ray images. To address the issue of spatial inconsistency in multi-angle image reconstruction, an innovative adaptive fusion module was designed to score initial reconstruction results during the inference stage and perform weighted fusion, thereby improving …

Implementing Unmanned Aerial Vehicles To Collect Human Gait Data At Distance And Altitude For Identification And Re-Identification, Donn E. Bartram

Graduate Theses, Dissertations, and Problem Reports (ETD)

Gait patterns are a class of biometric information pertaining to the way a person moves and poses. Gait information is unique to each person and can be used to identify and reidentify people. Historically, this task has been achieved through the use of multiple ground-based imaging sensors. However, as Unmanned Aerial Vehicles (UAVs) advance, they present the opportunity to evolve the process of persons identification and re-identification. Collecting human gait data using UAVs at distances ranging from 20m to 500m and altitudes ranging from 0m to 120m is a challenging task. The current biometric data collection methods, primarily designed for …

Domain Adaptive Federated Learning For Multi-Institution Molecular Mutation Prediction And Bias Identification, W. Farzana, M. A. Witherow, I. Longoria, M. S. Sadique, A. Temtam, K. M. Iftekharuddin

Electrical & Computer Engineering Faculty Publications

Deep learning models have shown potential in medical image analysis tasks. However, training a generalized deep learning model requires huge amounts of patient data that is usually gathered from multiple institutions which may raise privacy concerns. Federated learning (FL) provides an alternative to sharing data across institutions. Nonetheless, FL is susceptible to a few challenges including inversion attacks on model weights, heterogenous data distributions, and bias. This study addresses heterogeneity and bias issues for multi-institution patient data by proposing domain adaptive FL modeling using several radiomics (volume, fractal, texture) features for O6-methylguanine-DNA methyltransferase (MGMT) classification across multiple institutions. The proposed …

Don't Fear The Artificial Intelligence: A Systematic Review Of Machine Learning For Prostate Cancer Detection In Pathology, Aaryn Frewing, Alexander B. Gibson, Richard Robertson, Paul Urie, Dennis Della Corte

Faculty Publications

The adoption of whole slide image (WSI) scanners in clinical practice was accelerated by US Food and Drug Administration approval in 2017, which allowed primary pathologic diagnoses to be made on scanned images. Images in the digital domain allow the application of pathology artificial intelligence (AI), including clinical decision support with algorithms performing specific diagnoses.1,2 These algorithms, if trained properly, could go beyond the ability of human observation to detect and quantify features that are not recognizable by human perception.1,3,4

Enhancement Of Deep Learning Protein Structure Prediction, Ruoming Shen

Modeling, Simulation and Visualization Student Capstone Conference

Protein modeling is a rapidly expanding field with valuable applications in the pharmaceutical industry. Accurate protein structure prediction facilitates drug design, as extensive knowledge about the atomic structure of a given protein enables scientists to target that protein in the human body. However, protein structure identification in certain types of protein images remains challenging, with medium resolution cryogenic electron microscopy (cryo-EM) protein density maps particularly difficult to analyze. Recent advancements in computational methods, namely deep learning, have improved protein modeling. To maximize its accuracy, a deep learning model requires copious amounts of up-to-date training data.

This project explores DeepSSETracer, a …

Evaluating Human Eye Features For Objective Measure Of Working Memory Capacity, Yasasi Abeysinghe, Enkelejda Kasneci (Ed.), Frederick Shic (Ed.), Mohamed Khamis (Ed.)

Computer Science Faculty Publications

Eye tracking measures can provide means to understand the underlying development of human working memory. In this study, we propose to develop machine learning algorithms to find an objective relationship between human eye movements via oculomotor plant and their working memory capacity, which determines subjective cognitive load. Here we evaluate oculomotor plant features extracted from saccadic eye movements, traditional positional gaze metrics, and advanced eye metrics such as ambient/focal coefficient , gaze transition entropy, low/high index of pupillary activity (LHIPA), and real-time index of pupillary activity (RIPA). This paper outlines the proposed approach of evaluating eye movements for obtaining an …

The Role Of Generative Adversarial Networks In Bioimage Analysis And Computational Diagnostics., Ahmed Naglah

Electronic Theses and Dissertations

Computational technologies can contribute to the modeling and simulation of the biological environments and activities towards achieving better interpretations, analysis, and understanding. With the emergence of digital pathology, we can observe an increasing demand for more innovative, effective, and efficient computational models. Under the umbrella of artificial intelligence, deep learning mimics the brain’s way in learn complex relationships through data and experiences. In the field of bioimage analysis, models usually comprise discriminative approaches such as classification and segmentation tasks. In this thesis, we study how we can use generative AI models to improve bioimage analysis tasks using Generative Adversarial Networks …

Machine Learning And Scalable Informatics Methods To Predict Disease Status From Multimodal Biomedical Data, Hossein Mohammadian Foroushani

McKelvey School of Engineering Graduate Student Theses & Dissertations

Biological understanding of complex diseases such as stroke and obesity is critical for the advancement of medicine. Further knowledge discovery can provide effective biomarkers to improve disease diagnosis and prognosis, identify driver mutations, predict individual genetic susceptibility for early prevention and effective disease management, and facilitate development of personalized drugs. Stroke is the second leading cause of death and long-term disability in the world. Thus, stroke management is a time-sensitive emergency. The initial hours after stroke onset map the trajectory of subsequent neurologic complications. Cerebral edema develops hours to days after acute ischemic stroke and may result in midline shift …

Analysis Of An Existing Method In Refinement Of Protein Structure Predictions Using Cryo-Em Images, Maytha Alshammari, Jing He, Willy Wriggers, Jiangwen Sun

College of Sciences Posters

Protein structure prediction produces atomic models from its amino acid sequence. Three-dimensional structures are important for understanding the function mechanism of proteins. Knowing the structure of a given protein is crucial in drug development design of novel enzymes. AlphaFold2 is a protein structure prediction tool with good performance in recent CASP competitions. Phenix is a tool for determination of a protein structure from a high-resolution 3D molecular image. Recent development of Phenix shows that it is capable to refine predicted models from AlphaFold2, specifically the poorly predicted regions, by incorporating information from the 3D image of the protein. The goal …

Two-Stage Transfer Learning For Facial Expression Classification In Children, Gregory Hubbard, Megan Witherow, Khan Iftekharuddin

Undergraduate Research Symposium

Studying facial expressions can provide insight into the development of social skills in children and provide support to individuals with developmental disorders. In afflicted individuals, such as children with Autism Spectrum Disorder (ASD), atypical interpretations of facial expressions are well-documented. In computer vision, many popular and state-of-the-art deep learning architectures (VGG16, EfficientNet, ResNet, etc.) are readily available with pre-trained weights for general object recognition. Transfer learning utilizes these pre-trained models to improve generalization on a new task. In this project, transfer learning is implemented to leverage the pretrained model (general object recognition) on facial expression classification. Though this method, the …

Design And Development Of Software With A Graphical User Interface To Display And Convert Multiple Microscopic Histology Images, Sayed Ahmadreza Razian, Majid Jadidi, Alexey Kamenskiy

UNO Student Research and Creative Activity Fair

Histological images are widely used to assess the microscopic anatomy of biological tissues. Recent advancements in image analysis allow the identification of structural features on histological sections that can help advance medical device development, brain and cancer research, drug discovery, vascular mechanobiology, and many other fields. Histological slide scanners create images in SVS and TIFF formats that were designed to archive image blocks and high-resolution textual information. Because these formats were primarily intended for storage, they are often not compatible with conventional image analysis software and require conversion before they can be used in research. We have developed a user-friendly …

Segmentation Of Intracranial Structures From Noncontrast Ct Images With Deep Learning, Evan Porter

Wayne State University Dissertations

Presented in this work is an investigation of the application of artificially intelligent algorithms, namely deep learning, to generate segmentations for the application in functional avoidance radiotherapy treatment planning. Specific applications of deep learning for functional avoidance include generating hippocampus segmentations from computed tomography (CT) images and generating synthetic pulmonary perfusion images from four-dimensional CT (4DCT).A single institution dataset of 390 patients treated with Gamma Knife stereotactic radiosurgery was created. From these patients, the hippocampus was manually segmented on the high-resolution MR image and used for the development of the data processing methodology and model testing. It was determined that …

Development Of Deep Learning Neural Network For Ecological And Medical Images, Shaobo Liu

Dissertations

Deep learning in computer vision and image processing has attracted attentions from various fields including ecology and medical image. Ecologists are interested in finding an effective model structure to classify different species. Tradition deep learning model use a convolutional neural network, such as LeNet, AlexNet, VGG models, residual neural network, and inception models, are first used on classifying bee wing and butterfly datasets. However, insufficient data sample and unbalanced samples in each class have caused a poor accuracy. To make improvement the test accuracy, data augmentation and transfer learning are applied. Recently developed deep learning framework based on mathematical morphology …

Using Deep Learning To Analyze Materials In Medical Images, Carson Molder

Computer Science and Computer Engineering Undergraduate Honors Theses

Modern deep learning architectures have become increasingly popular in medicine, especially for analyzing medical images. In some medical applications, deep learning image analysis models have been more accurate at predicting medical conditions than experts. Deep learning has also been effective for material analysis on photographs. We aim to leverage deep learning to perform material analysis on medical images. Because material datasets for medicine are scarce, we first introduce a texture dataset generation algorithm that automatically samples desired textures from annotated or unannotated medical images. Second, we use a novel Siamese neural network called D-CNN to predict patch similarity and build …

An Improved Clipped Sub-Histogram Equalization Technique Using Optimized Local Contrast Factor For Mammogram Image Analysis, Nurshafira Hazim Chan

Student Works (2020-2029)

Mammography has been known worldwide as the most common imaging modalities utilized for early detection of breast cancer. The mammographic images produced are in greyscale, however they often produced poor contrast images, non-uniform illumination, and the image often contain artefacts and noise. These limitations can be overcame during the pre-processing stage by improving the image enhancement process. Therefore, in this research an optimized enhancement framework is developed where the local contrast factor is manipulated to preserve details of the image. This technique aims to improve the overall image visibility without altering histogram of the original image, which will affect the …

Deep Learning-Based Breast Cancer Detection And Classification Using Histopathology Images, Ghulam Murtaza

Student Works (2020-2029)

Cancer disease is drastically increasing worldwide over the past few years. Among all types of cancers in women, breast cancer (BrC) is the main cause of abnormal deaths. For a confident diagnosis of BrC, histopathology (Hp) images are usually suggested by the doctors. BrC detection is a diagnostic test for benign (non-cancerous) and malignant (cancerous) breast tumors (BrT). Once the BrT is diagnosed, then it needs to be classified for subtypes of benign and malignant to start specific treatment. Several studies developed BrC detection and classification models using Hp images. However, the existing models required high computational resources, long training …

Three-Dimensional (3d) Reconstruction Of Computer Tomography Cardiac Images Using Visualization Toolkit (Vtk), Mohammad Zahid Zamaludin

Student Works (2020-2029)

Cardiac study proves to be important to diagnose and treat cardiovascular disease that accounts for 20% of death in the world. Three-dimensional (3D) images of the heart are important to understand its real structure, functions and physical abnormalities occurring in the tissue. The objective of this paper is to reconstruct 3D images from 2D CT Scan slices of the cardiac region of interest using Visualisation Toolkit (VTK). The Visualization Toolkit (VTK) is non-propriety software with state-of-the-art tools for 3D rendering that is easily customizable programming. 3D images are reconstructed using surface rendering technique and volume rendering technique. Volume rendering shows …

Perceptually Improved Medical Image Translations Using Conditional Generative Adversarial Networks, Anurag Vaidya

Honors Theses

Magnetic resonance imaging (MRI) can help visualize various brain regions. Typical MRI sequences consist of T1-weighted sequence (favorable for observing large brain structures), T2-weighted sequence (useful for pathology), and T2-FLAIR scan (useful for pathology with suppression of signal from water). While these different scans provide complementary information, acquiring them leads to acquisition times of ~1 hour and an average cost of $2,600, presenting significant barriers. To reduce these costs associated with brain MRIs, we present pTransGAN, a generative adversarial network capable of translating both healthy and unhealthy T1 scans into T2 scans. We show that the addition of non-adversarial …

Fmri Feature Extraction Model For Adhd Classification Using Convolutional Neural Network, Senuri De Silva, Sanuwani Udara Dayarathna, Gangani Ariyarathne, Dulani Meedeniya, Sampath Jayarathna

Computer Science Faculty Publications

Biomedical intelligence provides a predictive mechanism for the automatic diagnosis of diseases and disorders. With the advancements of computational biology, neuroimaging techniques have been used extensively in clinical data analysis. Attention deficit hyperactivity disorder (ADHD) is a psychiatric disorder, with the symptomology of inattention, impulsivity, and hyperactivity, in which early diagnosis is crucial to prevent unwelcome outcomes. This study addresses ADHD identification using functional magnetic resonance imaging (fMRI) data for the resting state brain by evaluating multiple feature extraction methods. The features of seed-based correlation (SBC), fractional amplitude of low-frequency fluctuation (fALFF), and regional homogeneity (ReHo) are comparatively applied to …

Combining Cryo-Em Density Map And Residue Contact For Protein Secondary Structure Topologies, Maytha Alshammari, Jing He

Computer Science Faculty Publications

Although atomic structures have been determined directly from cryo-EM density maps with high resolutions, current structure determination methods for medium resolution (5 to 10 Å) cryo-EM maps are limited by the availability of structure templates. Secondary structure traces are lines detected from a cryo-EM density map for α-helices and β-strands of a protein. A topology of secondary structures defines the mapping between a set of sequence segments and a set of traces of secondary structures in three-dimensional space. In order to enhance accuracy in ranking secondary structure topologies, we explored a method that combines three sources of information: a set …

Three-Dimensional (3d) Reconstruction Of Ultrasound Foetal Images Using Visualisation Toolkit (Vtk), Henry Dusim Grace Anabela

Student Works (2020-2029)

Three-dimensional (3D) reconstruction of two-dimensional (2D) ultrasound images has been used for easier analysing of anatomy related to abnormalities. 3D image reconstruction is required as it helps doctors interpret results and assess any defects quickly. Not only does it improve the accuracy of diagnosis, but it also saves a lot of time. In this thesis, 295 slices of 2D ultrasound images of a foetal are used to do 3D reconstruction using surface rendering, volume rendering, and multiplanar rendering technique. A similar software known as 3D Slicer is also used to reconstruct the ultrasound image for comparison purpose. VTK can easily …

Identification Of Microscopy Cell Images By Using Convolutional Neural Network Application, Ajis Norfatin Farisya

Student Works (2020-2029)

Breast cancer has been the major factor of cancer death and the second main cause of women’s deaths in the world. The false positive results of this cancer cell detection during the screening test leads to false treatment and emotional disturbance of the patients. Thus, breast cancer cell lines (MCF7) is used as the microscopy image samples together with the Human Bone Osteosarcoma Epithelial Cells (U2OS), and Human Hepatocyte as control to study the effectiveness of convolutional neural network (CNN) as a method of image recognition. The objectives of this study are to determine the ability of convolutional neural network …

An Improved Bone Age Assessment Using Advanced Image Processing And Deep Learning Approach, Kim Meng Liang

Student Works (2020-2029)

Pediatricians often apply bone age assessment to measure the skeletal maturity of children and to predict the future height. These discrepancies are good indicators for diagnosing growth disorders. Normally, left hand skeletal is employed in this assessment. The low quality of ossification sites of carpals deteriorates the pediatrician’s visibility in inspecting the pertinent radiographic manifestations. This in turn affects the bone age assessment. Therefore, we have to enhance the quality before assessing them. Histogram equalization is one of the contrast enhancement techniques that suit this type of enhancement. Existing histogram equalizations, however, are confronting with problems in preserving the brightness …