Biomedical Engineering and Bioengineering Commons^™

Efficient Scopeformer: Toward Scalable And Rich Feature Extraction For Intracranial Hemorrhage Detection, Yassine Barhoumi, Nidhal Carla Bouaynaya, Ghulam Rasool

Henry M. Rowan College of Engineering Faculty Scholarship

The quality and richness of feature maps extracted by convolution neural networks (CNNs) and vision Transformers (ViTs) directly relate to the robust model performance. In medical computer vision, these information-rich features are crucial for detecting rare cases within large datasets. This work presents the “Scopeformer,” a novel multi-CNN-ViT model for intracranial hemorrhage classification in computed tomography (CT) images. The Scopeformer architecture is scalable and modular, which allows utilizing various CNN architectures as the backbone with diversified output features and pre-training strategies. We propose effective feature projection methods to reduce redundancies among CNN-generated features and to control the input size of …

Evalattai: A Holistic Approach To Evaluating Attribution Maps In Robust And Non-Robust Models, Ian E. Nielsen, Ravi Ramachandran, Nidhal Carla Bouaynaya, Hassan M. Fathallah-Shaykh, Ghulam Rasool

Henry M. Rowan College of Engineering Faculty Scholarship

Eyes are one of the main critical organs of the body that provide our brain with the most information about the surrounding environment. Disturbance in the activity of this informational organ, resulting from different ocular diseases, could affect the quality of life, so finding appropriate methods for treating ocular disease has attracted lots of attention. This is especially due to the ineffectiveness of the conventional therapeutic method to deliver drugs into the interior parts of the eye, and the also presence of barriers such as tear film, blood-ocular, and blood-retina barriers. Recently, some novel techniques, such as different types of …

Efficient Scopeformer: Towards Scalable And Rich Feature Extraction For Intracranial Hemorrhage Detection Using Hybrid Convolution And Vision Transformer Networks, Yassine Barhoumi

Theses and Dissertations

The field of medical imaging has seen significant advancements through the use of artificial intelligence (AI) techniques. The success of deep learning models in this area has led to the need for further research. This study aims to explore the use of various deep learning algorithms and emerging modeling techniques to improve training paradigms in medical imaging. Convolutional neural networks (CNNs) are the go-to architecture for computer vision problems, but they have limitations in mapping long-term dependencies within images. To address these limitations, the study explores the use of techniques such as global average pooling and self-attention mechanisms. Additionally, the …

Image Processing Algorithms For Detection Of Anomalies In Orthopedic Surgery Implants, Alexander William Wiese

Theses and Dissertations

Orthopedic implant procedures for hip implants are performed on 300,000 patients annually in the United States, with 22.3 million procedures worldwide. While most such operations are successfully performed to relieve pain and restore joint function for the duration of the patient's life, advances in medicine have enabled patients to outlive the life of their implant, increasing the likelihood of implant failure. There is significant advantage to the patient, the surgeon, and the medical community in early detection of implant failures.The research work presented in this thesis demonstrates a non-invasive digital image processing technique for the automated detection of specific arthroplasty …

Inverted Cone Convolutional Neural Network For Deboning Mris, Oliver John Palumbo

Theses and Dissertations

Data plenitude is the power but also the bottleneck for data-driven approaches, including neural networks. In particular, Convolutional Neural Networks (CNNs) require an abundant database of training images to achieve a desired high accuracy. Current techniques employed for boosting small datasets are data augmentation and synthetic data generation, which suffer from computational complexity and imprecision compared to original datasets. In this thesis, we intercalate prior knowledge based on the temporal relation between the images in the third dimension. Specifically, we compute the gradient of subsequent images in the dataset to remove extraneous information and highlight subtle variations between the images. …

Diagnosing Growth In Low-Grade Gliomas With And Without Longitudinal Volume Measurements: A Retrospective Observational Study., Hassan M Fathallah-Shaykh, Andrew Deatkine, Elizabeth Coffee, Elias Khayat, Asim K Bag, Xiaosi Han, Paula Province Warren, Markus Bredel, John Fiveash, James Markert, Nidhal Carla Bouaynaya, Louis B Nabors

Henry M. Rowan College of Engineering Faculty Scholarship

BACKGROUND: Low-grade gliomas cause significant neurological morbidity by brain invasion. There is no universally accepted objective technique available for detection of enlargement of low-grade gliomas in the clinical setting; subjective evaluation by clinicians using visual comparison of longitudinal radiological studies is the gold standard. The aim of this study is to determine whether a computer-assisted diagnosis (CAD) method helps physicians detect earlier growth of low-grade gliomas.

METHODS AND FINDINGS: We reviewed 165 patients diagnosed with grade 2 gliomas, seen at the University of Alabama at Birmingham clinics from 1 July 2017 to 14 May 2018. MRI scans were collected during …

Inception Modules Enhance Brain Tumor Segmentation., Daniel E Cahall, Ghulam Rasool, Nidhal Carla Bouaynaya, Hassan M Fathallah-Shaykh

Henry M. Rowan College of Engineering Faculty Scholarship

Magnetic resonance images of brain tumors are routinely used in neuro-oncology clinics for diagnosis, treatment planning, and post-treatment tumor surveillance. Currently, physicians spend considerable time manually delineating different structures of the brain. Spatial and structural variations, as well as intensity inhomogeneity across images, make the problem of computer-assisted segmentation very challenging. We propose a new image segmentation framework for tumor delineation that benefits from two state-of-the-art machine learning architectures in computer vision, i.e., Inception modules and U-Net image segmentation architecture. Furthermore, our framework includes two learning regimes, i.e., learning to segment intra-tumoral structures (necrotic and non-enhancing tumor core, peritumoral edema, …

Automatic Analysis Of Surface Electromyography Of Reflexes For Central Nervous System Inhibition During Pregnancy, Neda Parvin

Theses and Dissertations

Eclampsia is a life-threatening neurological complication (seizures or coma) of pregnancy, and represents progression of pre-eclampsia (high blood pressure and protein in the urine isolated to pregnancy). Even though the exact mechanism of pre-eclampsia and its neurological manifestations have yet to be definitively established, it is known that there is a loss of inhibitory impulses between the cerebral cortex and the spinal cord resulting in exaggerated reflexes. In this work, we establish normative measures for deep tendon reflexes (DTR) during pregnancy. We quantified the surface electromyogram (EMG) at the knee and ankle of DTRs in 279 subjects. The signals were …