Whole-cerebrum guanidino and amide CEST mapping at 3 T by a 3D stack-of-spirals gradient echo acquisition.

PURPOSE: To develop a 3D, high-sensitivity CEST mapping technique based on the 3D stack-of-spirals (SOS) gradient echo readout, the proposed approach was compared with conventional acquisition techniques and evaluated for its efficacy in concurrently mapping of guanidino (Guan) and amide CEST in human brain at 3 T, leveraging the polynomial Lorentzian line-shape fitting (PLOF) method. METHODS: Saturation time and recovery delay were optimized to achieve maximum CEST time efficiency. The 3DSOS method was compared with segmented 3D EPI (3DEPI), turbo spin echo, and gradient- and spin-echo techniques. Image quality, temporal SNR (tSNR), and test-retest reliability were assessed. Maps of Guan and amide CEST derived from 3DSOS were demonstrated on a low-grade glioma patient. RESULTS: The optimized recovery delay/saturation time was determined to be 1.4/2 s for Guan and amide CEST. In addition to nearly doubling the slice number, the gradient echo techniques also outperformed spin echo sequences in tSNR: 3DEPI (193.8 ± 6.6), 3DSOS (173.9 ± 5.6), and GRASE (141.0 ± 2.7). 3DSOS, compared with 3DEPI, demonstrated comparable GuanCEST signal in gray matter (GM) (3DSOS: [2.14%-2.59%] vs. 3DEPI: [2.15%-2.61%]), and white matter (WM) (3DSOS: [1.49%-2.11%] vs. 3DEPI: [1.64%-2.09%]). 3DSOS also achieves significantly higher amideCEST in both GM (3DSOS: [2.29%-3.00%] vs. 3DEPI: [2.06%-2.92%]) and WM (3DSOS: [2.23%-2.66%] vs. 3DEPI: [1.95%-2.57%]). 3DSOS outperforms 3DEPI in terms of scan-rescan reliability (correlation coefficient: 3DSOS: 0.58-0.96 vs. 3DEPI: -0.02 to 0.75) and robustness to motion as well. CONCLUSION: The 3DSOS CEST technique shows promise for whole-cerebrum CEST imaging, offering uniform contrast and robustness against motion artifacts.

Alcohol for seizure induction in the epilepsy monitoring unit.

RATIONALE: Seizure induction techniques are used in the epilepsy monitoring unit (EMU) to increase diagnostic yield and reduce length of stay. There are insufficient data on the efficacy of alcohol as an induction technique. METHODS: We performed a retrospective cohort study using six years of EMU data at our institution. We compared cases who received alcohol for seizure induction to matched controls who did not. The groups were matched on the following variables: age, reason for admission, length of stay, number of antiseizure medications (ASM) at admission, whether ASMs were tapered during admission, and presence of interictal epileptiform discharges. We used both propensity score and exact matching strategies. We compared the likelihood of epileptic seizures and nonepileptic events in cases versus controls using Kaplan-Meier time-to-event analysis, as well as odds ratios for these outcomes occurring at any time during the admission. RESULTS: We analyzed 256 cases who received alcohol (median dose 2.5 standard drinks) and 256 propensity score-matched controls. Cases who received alcohol were no more likely than controls to have an epileptic seizure (X2(1) = 0.01, p = 0.93) or nonepileptic event (X2(1) = 2.1, p = 0.14) in the first 48 h after alcohol administration. For the admission overall, cases were no more likely to have an epileptic seizure (OR 0.89, 95 % CI 0.61-1.28, p = 0.58), nonepileptic event (OR 0.97, CI 0.62-1.53, p = 1.00), nor require rescue benzodiazepine (OR 0.63, CI 0.35-1.12, p = 0.15). Stratified analyses revealed no increased risk of epileptic seizure in any subgroups. Sensitivity analysis using exact matching showed that results were robust to matching strategy. CONCLUSIONS: Alcohol was not an effective induction technique in the EMU. This finding has implications for counseling patients with epilepsy about the risks of drinking alcohol in moderation in their daily lives.

Atorvastatin rescues hyperhomocysteinemia-induced cognitive deficits and neuroinflammatory gene changes.

BACKGROUND: Epidemiological data suggests statins could reduce the risk of dementia, and more specifically, Alzheimer's disease (AD). Pre-clinical data suggests statins reduce the risk of dementia through their pleiotropic effects rather than their cholesterol lowering effects. While AD is a leading cause of dementia, it is frequently found co-morbidly with cerebral small vessel disease and other vascular contributions to cognitive impairment and dementia (VCID), which are another leading cause of dementia. In this study, we determined if atorvastatin ameliorated hyperhomocysteinemia (HHcy)-induced VCID. METHODS: Wild-type (C57Bl6/J) mice were placed on a diet to induce HHcy or a control diet each with or without atorvastatin for 14 weeks. Mice underwent novel object recognition testing before tissue collection. Plasma total cholesterol and total homocysteine as well as related metabolites were measured. Using qPCR and NanoString technology, we profiled glial cell-associated gene expression changes. Finally, microglial morphology, astrocyte end feet, and microhemorrhages were analyzed using histological methods. RESULTS: Atorvastatin treatment of HHcy in mice led to no changes in total cholesterol but decreases in total homocysteine in plasma. While HHcy decreased expression of many glial genes, atorvastatin rescued these gene changes, which mostly occurred in oligodendrocytes and microglia. Microglia in HHcy mice with atorvastatin were trending towards fewer processes compared to control with atorvastatin, but there were no atorvastatin effects on astrocyte end feet. While atorvastatin treatment was trending towards increasing the area of microhemorrhages in HHcy mice in the frontal cortex, it only slightly (non-significantly) reduced the number of microhemorrhages. Finally, atorvastatin treatment in HHcy mice led to improved cognition on the novel object recognition task. CONCLUSIONS: These data suggest that atorvastatin rescued cognitive changes induced by HHcy most likely through lowering plasma total homocysteine and rescuing gene expression changes rather than impacts on vascular integrity or microglial changes.

A pharmacokinetic model of antiseizure medication load to guide care in the epilepsy monitoring unit.

OBJECTIVE: Evaluating patients with drug-resistant epilepsy often requires inducing seizures by tapering antiseizure medications (ASMs) in the epilepsy monitoring unit (EMU). The relationship between ASM taper strategy, seizure timing, and severity remains unclear. In this study, we developed and validated a pharmacokinetic model of total ASM load and tested its association with seizure occurrence and severity in the EMU. METHODS: We studied 80 patients who underwent intracranial electroencephalographic recording for epilepsy surgery planning. We developed a first order pharmacokinetic model of the ASMs administered in the EMU to generate a continuous metric of overall ASM load. We then related modeled ASM load to seizure likelihood and severity. We determined the association between the rate of ASM load reduction, the length of hospital stay, and the probability of having a severe seizure. Finally, we used modeled ASM load to predict oncoming seizures. RESULTS: Seizures occurred in the bottom 50th percentile of sampled ASM loads across the cohort (p < .0001, Wilcoxon signed-rank test), and seizures requiring rescue therapy occurred at lower ASM loads than seizures that did not require rescue therapy (logistic regression mixed effects model, odds ratio = .27, p = .01). Greater ASM decrease early in the EMU was not associated with an increased likelihood of having a severe seizure, nor with a shorter length of stay. SIGNIFICANCE: A pharmacokinetic model can accurately estimate ASM levels for patients in the EMU. Lower modeled ASM levels are associated with increased seizure likelihood and seizure severity. We show that ASM load, rather than ASM taper speed, is associated with severe seizures. ASM modeling has the potential to help optimize taper strategy to minimize severe seizures while maximizing diagnostic yield.

Long-term epilepsy outcome dynamics revealed by natural language processing of clinic notes.

OBJECTIVE: Electronic medical records allow for retrospective clinical research with large patient cohorts. However, epilepsy outcomes are often contained in free text notes that are difficult to mine. We recently developed and validated novel natural language processing (NLP) algorithms to automatically extract key epilepsy outcome measures from clinic notes. In this study, we assessed the feasibility of extracting these measures to study the natural history of epilepsy at our center. METHODS: We applied our previously validated NLP algorithms to extract seizure freedom, seizure frequency, and date of most recent seizure from outpatient visits at our epilepsy center from 2010 to 2022. We examined the dynamics of seizure outcomes over time using Markov model-based probability and Kaplan-Meier analyses. RESULTS: Performance of our algorithms on classifying seizure freedom was comparable to that of human reviewers (algorithm F1 = .88 vs. human annotator κ = .86). We extracted seizure outcome data from 55 630 clinic notes from 9510 unique patients written by 53 unique authors. Of these, 30% were classified as seizure-free since the last visit, 48% of non-seizure-free visits contained a quantifiable seizure frequency, and 47% of all visits contained the date of most recent seizure occurrence. Among patients with at least five visits, the probabilities of seizure freedom at the next visit ranged from 12% to 80% in patients having seizures or seizure-free at the prior three visits, respectively. Only 25% of patients who were seizure-free for 6 months remained seizure-free after 10 years. SIGNIFICANCE: Our findings demonstrate that epilepsy outcome measures can be extracted accurately from unstructured clinical note text using NLP. At our tertiary center, the disease course often followed a remitting and relapsing pattern. This method represents a powerful new tool for clinical research with many potential uses and extensions to other clinical questions.

CPSF3-dependent pre-mRNA processing as a druggable node in AML and Ewing's sarcoma.

The post-genomic era has seen many advances in our understanding of cancer pathways, yet resistance and tumor heterogeneity necessitate multiple approaches to target even monogenic tumors. Here, we combine phenotypic screening with chemical genetics to identify pre-messenger RNA endonuclease cleavage and polyadenylation specificity factor 3 (CPSF3) as the target of JTE-607, a small molecule with previously unknown target. We show that CPSF3 represents a synthetic lethal node in a subset of acute myeloid leukemia (AML) and Ewing's sarcoma cancer cell lines. Inhibition of CPSF3 by JTE-607 alters expression of known downstream effectors in AML and Ewing's sarcoma lines, upregulates apoptosis and causes tumor-selective stasis in mouse xenografts. Mechanistically, it prevents the release of newly synthesized pre-mRNAs, resulting in read-through transcription and the formation of DNA-RNA hybrid R-loop structures. This study implicates pre-mRNA processing, and specifically CPSF3, as a druggable target providing an avenue to therapeutic intervention in cancer.

Author Correction: CPSF3-dependent pre-mRNA processing as a druggable node in AML and Ewing's sarcoma.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Discovery of a ZIP7 inhibitor from a Notch pathway screen.

The identification of activating mutations in NOTCH1 in 50% of T cell acute lymphoblastic leukemia has generated interest in elucidating how these mutations contribute to oncogenic transformation and in targeting the pathway. A phenotypic screen identified compounds that interfere with trafficking of Notch and induce apoptosis via an endoplasmic reticulum (ER) stress mechanism. Target identification approaches revealed a role for SLC39A7 (ZIP7), a zinc transport family member, in governing Notch trafficking and signaling. Generation and sequencing of a compound-resistant cell line identified a V430E mutation in ZIP7 that confers transferable resistance to the compound NVS-ZP7-4. NVS-ZP7-4 altered zinc in the ER, and an analog of the compound photoaffinity labeled ZIP7 in cells, suggesting a direct interaction between the compound and ZIP7. NVS-ZP7-4 is the first reported chemical tool to probe the impact of modulating ER zinc levels and investigate ZIP7 as a novel druggable node in the Notch pathway.

Age and sex are associated with the plasma lipidome: findings from the GOLDN study.

BACKGROUND: Developing an understanding of the biochemistry of aging in both sexes is critical for managing disease throughout the lifespan. Lipidomic associations with age and sex have been reported, but prior studies are limited by measurements in serum rather than plasma or by participants taking lipid-lowering medications. METHODS: Our study included lipidomic data from 980 participants aged 18-87 years old from the Genetics of Lipid-Lowering Drugs and Diet Network (GOLDN). Participants were off lipid-lowering medications for at least 4 weeks, and signal intensities of 413 known lipid species were measured in plasma. We examined linear age and sex associations with signal intensity of (a) 413 lipid species; (b) 6 lipid classes (glycerolipids, glycerophospholipids, sphingolipids, sterol lipids, fatty acids, and acylcarnitines); and (c) 15 lipid subclasses; as well as with the particle sizes of three lipoproteins. RESULTS: Significant age associations were identified in 4 classes, 11 subclasses, 147 species, and particle size of one lipoprotein while significant sex differences were identified in 5 classes, 12 subclasses, 248 species, and particle sizes of two lipoproteins. For many lipid species (n = 97), age-related associations were significantly different between males and females. Age*sex interaction effects were most prevalent among phosphatidylcholines, sphingomyelins, and triglycerides. CONCLUSION: We identified several lipid species, subclasses, and classes that differ by age and sex; these lipid phenotypes may serve as useful biomarkers for lipid changes and associated cardiovascular risk with aging in the future. Future studies of age-related changes throughout the adult lifespan of both sexes are warranted. TRIAL REGISTRATION: ClinicalTrials.gov NCT00083369 ; May 21, 2004.

Dynamically Masked Audiograms With Machine Learning Audiometry.

OBJECTIVES: When one ear of an individual can hear significantly better than the other ear, evaluating the worse ear with loud probe tones may require delivering masking noise to the better ear to prevent the probe tones from inadvertently being heard by the better ear. Current masking protocols are confusing, laborious, and time consuming. Adding a standardized masking protocol to an active machine learning audiogram procedure could potentially alleviate all of these drawbacks by dynamically adapting the masking as needed for each individual. The goal of this study is to determine the accuracy and efficiency of automated machine learning masking for obtaining true hearing thresholds. DESIGN: Dynamically masked automated audiograms were collected for 29 participants between the ages of 21 and 83 (mean 43, SD 20) with a wide range of hearing abilities. Normal-hearing listeners were given unmasked and masked machine learning audiogram tests. Listeners with hearing loss were given a standard audiogram test by an audiologist, with masking stimuli added as clinically determined, followed by a masked machine learning audiogram test. The hearing thresholds estimated for each pair of techniques were compared at standard audiogram frequencies (i.e., 0.25, 0.5, 1, 2, 4, 8 kHz). RESULTS: Masked and unmasked machine learning audiogram threshold estimates matched each other well in normal-hearing listeners, with a mean absolute difference between threshold estimates of 3.4 dB. Masked machine learning audiogram thresholds also matched well the thresholds determined by a conventional masking procedure, with a mean absolute difference between threshold estimates for listeners with low asymmetry and high asymmetry between the ears, respectively, of 4.9 and 2.6 dB. Notably, out of 6200 masked machine learning audiogram tone deliveries for this study, no instances of tones detected by the nontest ear were documented. The machine learning methods were also generally faster than the manual methods, and for some listeners, substantially so. CONCLUSIONS: Dynamically masked audiograms achieve accurate true threshold estimates and reduce test time compared with current clinical masking procedures. Dynamic masking is a compelling alternative to the methods currently used to evaluate individuals with highly asymmetric hearing, yet can also be used effectively and efficiently for anyone.

Industry payments to female pelvic medicine and reconstructive surgeons: an analysis of Sunshine Act open payments from 2014-2017.

INTRODUCTION AND HYPOTHESIS: We aim to examine the financial relationship between industry and female pelvic medicine and reconstructive surgeons (FPMRS) during the first four full calendar years since the implementation of the Sunshine Act. METHODS: All board-certified FPMRS specialists were identified using the American Board of Medical Specialties directory. Program directors (PDs) were identified using an Accreditation Council for Graduate Medical Education (ACGME) database. All identified physicians were categorized by gender, specialty, and American Urological Association (AUA) region. Payment data for each individual from 2014 to 2017 were accessed using the Centers for Medicare and Medicaid Services (CMS) Open Payments website. Statistical analyses were performed to elucidate payment trends. RESULTS: Of the 1,307 FPMRS physicians identified, 25.1% (n = 328) are urology-trained and 74.9% (n = 979) are obstetrics/gynecology (OB/GYN)-trained. Of all physicians analyzed, 6.8% had no reported payments over the 4-year period. 90.1%, 86.5%, 85.3%, and 84.4% received some sort of payment in 2014 to 2017 respectively. Median total payments for all physicians decreased yearly, whereas mean payments decreased from 2014 to 2015 before increasing in all subsequent years. Median general payments were higher for men versus women, urology-trained versus OB/GYN-trained, and PDs versus non-PDs in all years analyzed. The largest contributor to overall payments was the "others" compensation category, which includes gifts, royalties, honoraria, and non-continuing medical education speaking engagements. CONCLUSIONS: Since institution of the Sunshine Act, the percentage of physicians receiving payments has decreased each year. Additionally, there has been a decrease in median total payments and an increase in yearly research payments in all years analyzed.

Disparities in seizure outcomes revealed by large language models.

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 health are associated with disparities in care access, but their impact on seizure outcomes among those with access remains unclear. Here we (1) evaluated our validated, epilepsy-specific LLM for intrinsic bias, and (2) used LLM-extracted seizure outcomes to determine if different demographic groups have different seizure outcomes. MATERIALS AND METHODS: We tested our LLM for differences and equivalences in prediction accuracy and confidence across demographic groups defined by race, ethnicity, sex, income, and health insurance, using manually annotated notes. Next, we used LLM-classified seizure freedom at each office visit to test for demographic outcome disparities, using univariable and multivariable analyses. RESULTS: We analyzed 84 675 clinic visits from 25 612 unique patients seen at our epilepsy center. We found little evidence of bias in the prediction accuracy or confidence of outcome classifications across demographic groups. Multivariable analysis indicated worse seizure outcomes for female patients (OR 1.33, P ≤ .001), those with public insurance (OR 1.53, P ≤ .001), and those from lower-income zip codes (OR ≥1.22, P ≤ .007). Black patients had worse outcomes than White patients in univariable but not multivariable analysis (OR 1.03, P = .66). CONCLUSION: We found little evidence that our LLM was intrinsically biased against any demographic group. Seizure freedom extracted by LLM revealed disparities in seizure outcomes across several demographic groups. These findings quantify the critical need to reduce disparities in the care of people with epilepsy.

Characterization of pain-related behaviors in a rat model of acute-to-chronic low back pain: single vs. multi-level disc injury.

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) to determine if multi-level disc injuries enhance/prolong pain. Methods: Twenty-three Sprague-Dawley adult female rats were used: 8 received disc puncture (DP) of one lumbar IVD (L5/6, DP-1); 8 received DP of two lumbar IVDs (L4/5 & L5/6, DP-2); 8 underwent sham surgery. Results: DP-2 rats showed local (low back) sensitivity to pressure at 6- and 12-weeks post-injury, and remote sensitivity to pressure (upper thighs) at 12- and 18-weeks and touch (hind paws) at 6, 12 and 18-weeks. DP-1 rats showed local and remote pressure sensitivity at 12-weeks only (and no tactile sensitivity), relative to Sham DP rats. Both DP groups showed reduced distance traveled during gait testing over multiple weeks, compared to pre-injury; only DP-2 rats showed reduced distance relative to Sham DP rats at 12-weeks. DP-2 rats displayed reduced positive interactions with a novel adult female rat at 3-weeks and hesitation and freezing during gait assays from 6-weeks onwards. At study end (18-weeks), radiological and histological analyses revealed reduced disc height and degeneration of punctured IVDs. Serum BDNF and TNFα levels were higher at 18-weeks in DP-2 rats, relative to Sham DP rats, and levels correlated positively with remote sensitivity in hind paws (tactile) and thighs (pressure). Discussion: Thus, multi-level disc injuries resulted in earlier, prolonged and greater discomfort locally and remotely, than single-level disc injury. BDNF and TNFα may have contributing roles.

Specific sequence determinants of miR-15/107 microRNA gene group targets.

MicroRNAs (miRNAs) target mRNAs in human cells via complex mechanisms that are still incompletely understood. Using anti-Argonaute (anti-AGO) antibody co-immunoprecipitation, followed by microarray analyses and downstream bioinformatics, 'RIP-Chip' experiments enable direct analyses of miRNA targets. RIP-Chip studies (and parallel assessments of total input mRNA) were performed in cultured H4 cells after transfection with miRNAs corresponding to the miR-15/107 gene group (miR-103, miR-107, miR-16 and miR-195), and five control miRNAs. Three biological replicates were run for each condition with a total of 54 separate human Affymetrix Human Gene 1.0 ST array replicates. Computational analyses queried for determinants of miRNA:mRNA binding. The analyses support four major findings: (i) RIP-Chip studies correlated with total input mRNA profiling provides more comprehensive information than using either RIP-Chip or total mRNA profiling alone after miRNA transfections; (ii) new data confirm that miR-107 paralogs target coding sequence (CDS) of mRNA; (iii) biochemical and computational studies indicate that the 3' portion of miRNAs plays a role in guiding miR-103/7 to the CDS of targets; and (iv) there are major sequence-specific targeting differences between miRNAs in terms of CDS versus 3'-untranslated region targeting, and stable AGO association versus mRNA knockdown. Future studies should take this important miRNA-to-miRNA variability into account.

Impact of thyroid hormone perturbations in adult mice: brain weight and blood vessel changes, gene expression variation, and neurobehavioral outcomes.

Mouse models of hyper- and hypothyroidism were used to examine the effects of thyroid hormone (TH) dyshomeostasis on the aging mammalian brain. 13-14 month-old mice were treated for 4months with either levothyroxine (hyperthyroid) or a propylthiouracil and methimazole combination (PTU/Met; hypothyroid). Hyperthyroid mice performed better on Morris Water Maze than control mice, while hypothyroid mice performed worse. Brain weight was increased in thyroxine-treated, and decreased in PTU/Met-treated animals. The brain weight change was strongly correlated with circulating and tissue T4. Quantitative measurements of microvessels were compared using digital neuropathologic methods. There was an increase in microvessel area in hyperthyroid mice. Hypothyroid mice showed a trend for elevated glial fibrillary acidic protein-immunoreactive astrocytes, indicating an increase in neuroinflammation. Gene expression alterations were associated with TH perturbation and astrocyte-expressed transcripts were particularly affected. For example, expression of Gli2 and Gli3, mediators in the Sonic Hedgehog signaling pathway, were strongly impacted by both treatments. We conclude that TH perturbations produce robust neurobehavioral, pathological, and brain gene expression changes in aging mouse models.

Generalization of finetuned transformer language models to new clinical contexts.

Objective: We have previously developed a natural language processing pipeline using clinical notes written by epilepsy specialists to extract seizure freedom, seizure frequency text, and date of last seizure text for patients with epilepsy. It is important to understand how our methods generalize to new care contexts. Materials and methods: We evaluated our pipeline on unseen notes from nonepilepsy-specialist neurologists and non-neurologists without any additional algorithm training. We tested the pipeline out-of-institution using epilepsy specialist notes from an outside medical center with only minor preprocessing adaptations. We examined reasons for discrepancies in performance in new contexts by measuring physical and semantic similarities between documents. Results: Our ability to classify patient seizure freedom decreased by at least 0.12 agreement when moving from epilepsy specialists to nonspecialists or other institutions. On notes from our institution, textual overlap between the extracted outcomes and the gold standard annotations attained from manual chart review decreased by at least 0.11 F1 when an answer existed but did not change when no answer existed; here our models generalized on notes from the outside institution, losing at most 0.02 agreement. We analyzed textual differences and found that syntactic and semantic differences in both clinically relevant sentences and surrounding contexts significantly influenced model performance. Discussion and conclusion: Model generalization performance decreased on notes from nonspecialists; out-of-institution generalization on epilepsy specialist notes required small changes to preprocessing but was especially good for seizure frequency text and date of last seizure text, opening opportunities for multicenter collaborations using these outcomes.

Disparities in seizure outcomes revealed by large language models.

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 with disparities in access to care, but their impact on seizure outcomes among those with access to specialty care remains unclear. Here we (1) evaluated our validated, epilepsy-specific LLM for intrinsic bias, and (2) used LLM-extracted seizure outcomes to test the hypothesis that different demographic groups have different seizure outcomes. Methods: First, we tested our LLM for intrinsic bias in the form of differential performance in demographic groups by race, ethnicity, sex, income, and health insurance in manually annotated notes. Next, we used LLM-classified seizure freedom at each office visit to test for outcome disparities in the same demographic groups, using univariable and multivariable analyses. Results: We analyzed 84,675 clinic visits from 25,612 patients seen at our epilepsy center 2005-2022. We found no differences in the accuracy, or positive or negative class balance of outcome classifications across demographic groups. Multivariable analysis indicated worse seizure outcomes for female patients (OR 1.33, p = 3×10-8), those with public insurance (OR 1.53, p = 2×10-13), and those from lower-income zip codes (OR ≥ 1.22, p ≤ 6.6×10-3). Black patients had worse outcomes than White patients in univariable but not multivariable analysis (OR 1.03, p = 0.66). Significance: We found no evidence that our LLM was intrinsically biased against any demographic group. Seizure freedom extracted by LLM revealed disparities in seizure outcomes across several demographic groups. These findings highlight the critical need to reduce disparities in the care of people with epilepsy.

Apolipoprotein E ε4/4 genotype limits response to dietary induction of hyperhomocysteinemia and resulting inflammatory signaling.

Vascular contributions to cognitive impairment and dementia (VCID) are the second leading cause of dementia behind Alzheimer's disease. Apolipoprotein E (ApoE) is a lipid transporting lipoprotein found within the brain and periphery. The APOE ε4 allele is the strongest genetic risk factor for late onset Alzheimer's disease and is a risk factor for VCID. Our lab has previously utilized a dietary model of hyperhomocysteinemia (HHcy) to induce VCID pathology and cognitive deficits in mice. This diet induces perivascular inflammation through cumulative oxidative damage leading to glial mediated inflammation and blood brain barrier breakdown. Here, we examine the impact of ApoE ε4 compared to ε3 alleles on the progression of VCID pathology and inflammation in our dietary model of HHcy. We report a significant resistance to HHcy induction in ε4 mice, accompanied by a number of related differences related to homocysteine (Hcy) metabolism and methylation cycle, or 1-C, metabolites. There were also significant differences in inflammatory profiles between ε3 and ε4 mice, as well as significant reduction in Serpina3n, a serine protease inhibitor associated with ApoE ε4, expression in ε4 HHcy mice relative to ε4 controls. Finally, we find evidence of pervasive sex differences within both genotypes in response to HHcy induction.

Extracting seizure frequency from epilepsy clinic notes: a machine reading approach to natural language processing.

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 this clinical information from unstructured text in clinical notes. If successful, this could improve clinical decision-making in epilepsy patients and allow for rapid, large-scale retrospective research. MATERIALS AND METHODS: We developed a finetuning pipeline for pretrained neural models to classify patients as being seizure-free and to extract text containing their seizure frequency and date of last seizure from clinical notes. We annotated 1000 notes for use as training and testing data and determined how well 3 pretrained neural models, BERT, RoBERTa, and Bio_ClinicalBERT, could identify and extract the desired information after finetuning. RESULTS: The finetuned models (BERTFT, Bio_ClinicalBERTFT, and RoBERTaFT) achieved near-human performance when classifying patients as seizure free, with BERTFT and Bio_ClinicalBERTFT achieving accuracy scores over 80%. All 3 models also achieved human performance when extracting seizure frequency and date of last seizure, with overall F1 scores over 0.80. The best combination of models was Bio_ClinicalBERTFT for classification, and RoBERTaFT for text extraction. Most of the gains in performance due to finetuning required roughly 70 annotated notes. DISCUSSION AND CONCLUSION: Our novel machine reading approach to extracting important clinical outcomes performed at or near human performance on several tasks. This approach opens new possibilities to support clinical practice and conduct large-scale retrospective clinical research. Future studies can use our finetuning pipeline with minimal training annotations to answer new clinical questions.

A Highly Predictive MicroRNA Panel for Determining Delayed Cerebral Vasospasm Risk Following Aneurysmal Subarachnoid Hemorrhage.

Approximately one-third of aneurysmal subarachnoid hemorrhage (aSAH) patients develop delayed cerebral vasospasm (DCV) 3-10 days after aneurysm rupture resulting in additional, permanent neurologic disability. Currently, no validated biomarker is available to determine the risk of DCV in aSAH patients. MicroRNAs (miRNAs) have been implicated in virtually all human diseases, including aSAH, and are found in extracellular biofluids including plasma and cerebrospinal fluid (CSF). We used a custom designed TaqMan Low Density Array miRNA panel to examine the levels of 47 selected brain and vasculature injury related miRNAs in CSF and plasma specimens collected from 31 patients with or without DCV at 3 and 7 days after aSAH, as well as from eight healthy controls. The analysis of the first 18-patient cohort revealed a striking differential expression pattern of the selected miRNAs in CSF and plasma of aSAH patients with DCV from those without DCV. Importantly, this differential expression was observed at the early time point (3 days after aSAH), before DCV event occurs. Seven miRNAs were identified as reliable DCV risk predictors along with a prediction model constructed based on an array of additional 19 miRNAs on the panel. These chosen miRNAs were then used to predict the risk of DCV in a separate, testing cohort of 15 patients. The accuracy of DCV risk prediction in the testing cohort reached 87%. The study demonstrates that our novel designed miRNA panel is an effective predictor of DCV risk and has strong applications in clinical management of aSAH patients.