A5000644086- Topic Modeling
- Epilepsy research and treatment
- Machine Learning in Healthcare
- Artificial Intelligence in Healthcare and Education
- Pharmacological Effects and Toxicity Studies
- Natural Language Processing Techniques
- Biomedical Text Mining and Ontologies
- Genetic Neurodegenerative Diseases
- Innovations in Medical Education
- Neonatal and fetal brain pathology
- Noise Effects and Management
- Medical Education and Admissions
- Reproductive Biology and Fertility
- Health Systems, Economic Evaluations, Quality of Life
- Pain Mechanisms and Treatments
- Neuroscience and Neuropharmacology Research
- Neurological disorders and treatments
- Cerebrovascular and genetic disorders
- Hearing, Cochlea, Tinnitus, Genetics
- Hearing Loss and Rehabilitation
- Handwritten Text Recognition Techniques
- EEG and Brain-Computer Interfaces
- Neurology and Historical Studies
- Musculoskeletal pain and rehabilitation
- Spine and Intervertebral Disc Pathology
University of Pennsylvania
2022-2025
Temple University
2024
California University of Pennsylvania
2024
Washington University in St. Louis
2020
Las Vegas Institute
2019-2020
Abstract Objective Seizure frequency and seizure freedom are among the most important outcome measures for patients with epilepsy. In this study, we aimed to automatically extract clinical information from unstructured text in notes. If successful, could improve decision-making epilepsy allow rapid, large-scale retrospective research. Materials Methods We developed a finetuning pipeline pretrained neural models classify as being seizure-free containing their date of last annotated 1000 notes...
Abstract Objective Large-language models (LLMs) can potentially revolutionize health care delivery and research, but risk propagating existing biases or introducing new ones. In epilepsy, social determinants of are associated with disparities in access, their impact on seizure outcomes among those access remains unclear. Here we (1) evaluated our validated, epilepsy-specific LLM for intrinsic bias, (2) used LLM-extracted to determine if different demographic groups have outcomes. Materials...
Abstract Objective Electronic medical records allow for retrospective clinical research with large patient cohorts. However, epilepsy outcomes are often contained in free text notes that difficult to mine. We recently developed and validated novel natural language processing (NLP) algorithms automatically extract key outcome measures from clinic notes. In this study, we assessed the feasibility of extracting these study history at our center. Methods applied previously NLP seizure freedom,...
Introduction Low back pain is the most common type of chronic pain. We examined pain-related behaviors across 18 weeks in rats that received injury to one or two lumbar intervertebral discs (IVD) determine if multi-level disc injuries enhance/prolong Methods Twenty-three Sprague-Dawley adult female were used: 8 puncture (DP) IVD (L5/6, DP-1); DP IVDs (L4/5 & L5/6, DP-2); underwent sham surgery. Results DP-2 showed local (low back) sensitivity pressure at 6- and 12-weeks post-injury,...
Evaluating patients with drug-resistant epilepsy often requires inducing seizures by tapering antiseizure medications (ASMs) in the monitoring unit (EMU). The relationship between ASM taper strategy, seizure timing, and severity remains unclear. In this study, we developed validated a pharmacokinetic model of total load tested its association occurrence EMU.We studied 80 who underwent intracranial electroencephalographic recording for surgery planning. We first order ASMs administered EMU to...
A wealth of important clinical information lies untouched in the Electronic Health Record, often form unstructured textual documents. For patients with Epilepsy, such includes outcome measures like Seizure Frequency and Dates Last Seizure, key parameters that guide all therapy for these patients. Transformer models have been able to extract from note text as sentences human-like accuracy; however, are not yet usable a quantitative analysis large-scale studies. In this study, we developed...
Abstract Objective We have previously developed a natural language processing pipeline using clinical notes written by epilepsy specialists to extract seizure freedom, frequency text, and date of last text for patients with epilepsy. It is important understand how our methods generalize new care contexts. Materials evaluated on unseen from nonepilepsy-specialist neurologists non-neurologists without any additional algorithm training. tested the out-of-institution specialist an outside...
Abstract Objective Large-language models (LLMs) in healthcare have the potential to propagate existing biases or introduce new ones. For people with epilepsy, social determinants of health are associated disparities access care, but their impact on seizure outcomes among those specialty care remains unclear. Here we (1) evaluated our validated, epilepsy-specific LLM for intrinsic bias, and (2) used LLM-extracted test hypothesis that different demographic groups outcomes. Methods First,...
Objectives: When one ear of an individual can hear significantly better than the other ear, evaluating worse with loud probe tones may require delivering masking noise to prevent from inadvertently being heard by ear. Current protocols are confusing, laborious, and time consuming. Adding a standardized protocol active machine learning audiogram procedure could potentially alleviate all these drawbacks dynamically adapting as needed for each individual. The goal this study is determine...
Abstract Purpose Pre-trained encoder transformer models have extracted information from unstructured clinic note text but require manual annotation for supervised fine-tuning. Large, Generative Pre- trained Transformers (GPTs) may streamline this process. In study, we explore GPTs in zero- and few-shot learning scenarios to analyze clinical health records. Materials Methods We prompt-engineered LLAMA2 13B optimize performance extracting seizure freedom epilepsy notes compared it against...