A5055282528- Medical Imaging and Analysis
- Spine and Intervertebral Disc Pathology
- Spinal Fractures and Fixation Techniques
- Artificial Intelligence in Healthcare and Education
- Traumatic Brain Injury and Neurovascular Disturbances
- Electronic Health Records Systems
- Pelvic and Acetabular Injuries
- Radiomics and Machine Learning in Medical Imaging
- Cerebrospinal fluid and hydrocephalus
- Spinal Dysraphism and Malformations
- Musculoskeletal pain and rehabilitation
- Traumatic Brain Injury Research
- Scoliosis diagnosis and treatment
- Spinal Cord Injury Research
- Clinical Reasoning and Diagnostic Skills
University of Miami
2024
University of California, Los Angeles
2021-2023
Abstract BACKGROUND The referral process for consultation with a spine surgeon remains inefficient, given substantial proportion of referrals to surgeons are nonoperative. OBJECTIVE To develop machine-learning-based algorithm which accurately identifies patients as candidates surgeon, using only magnetic resonance imaging (MRI). METHODS We trained deep U-Net machine learning model delineate spinal canals on axial slices 100 normal lumbar MRI scans were previously delineated by expert...
<sec> <title>BACKGROUND</title> Artificial intelligence (AI) and machine learning (ML) models are frequently developed in medical research to optimize patient care, yet they remain rarely utilized clinical practice. </sec> <title>OBJECTIVE</title> The present study aims understand the disconnect between model development implementation by surveying physicians of all specialties across United States. <title>METHODS</title> A HIPAA-compliant survey was emailed residency coordinators at...
INTRODUCTION: Few interventions have been shown to substantially modify neurologic outcome following acute traumatic spinal cord injury (SCI). Hyperosmolar therapy is an important adjunct in the treatment of brain (TBI); despite presumed overlap pathophysiology SCI and TBI, there no literature supporting use hyperosmolar for patients. METHODS: We performed a retrospective cohort study patients presenting our institution between 2005 2020. identified all admitted ICU 2020 with new diagnosis...
<title>Abstract</title> Many patients who present to their primary care physician for neck pain undergo magnetic resonance imaging (MRI) as part of diagnostic workup. The is then tasked with deciding if the findings MRI and workup warrant referral a spine surgery, an intricate task complicated by high rates background findings. This results in number non-surgical being referred surgery. Although there are multitude reasons still see subspecialist, deep learning has potential help inform...
Paraspinal muscle degeneration, defined by changes to cross-sectional area and fatty infiltration of the muscle, has been linked presence low back pain, sagittal imbalance, overall functional limitations. As a result, there is significant clinical value in efficiently evaluating degeneration. Segmentation paraspinal muscles difficult due considerable inter- intra- patient variability ambiguous boundaries between muscles. Identification adipose adds this challenge streaks within around In...
Machine learning algorithms tend to perform better within the setting wherein they are trained, a phenomenon known as domain effect. Deep learning-based medical image segmentation often trained using data acquired from specific scanners; however, these expected accurately segment anatomy in images scanners different ones used obtain training for such algorithms. In this work, we present evidence of scanner and magnet strength effect deep-U-Net spinal canals on axial MR images. The network...
The automated interpretation of spinal imaging using machine learning has emerged as a promising method for standardizing the assessment and diagnosis numerous column pathologies. While magnetic resonance images (MRIs) lumbar spine have been extensively studied in this context, cervical remains vastly understudied. Our objective was to develop automatically delineating cord neural foramina on axial MRIs learning. In study, we train state-of-the-art algorithm, namely multiresolution ensemble...