Citations

Abstract Aims An artificial intelligence-augmented electrocardiogram (AI-ECG) algorithm can identify left ventricular systolic dysfunction (LVSD). We sought to determine whether this AI-ECG algorithm could stratify mortality risk in cardiac intensive care unit (CICU) patients, independent of the presence of LVSD by transthoracic echocardiography (TTE). Methods and results We included 11 266 unique Mayo Clinic CICU patients admitted from 2007 to 2018 who underwent AI-ECG after CICU admission. Left ventricular ejection fraction (LVEF) data were extracted for patients with a TTE during hospitalization. Hospital mortality was analysed using multivariable logistic regression. Mean age was 68 ± 15 years, including 37% females. Higher AI-ECG probability of LVSD remained associated with higher hospital mortality [adjusted odds ratio (OR) 1.05 per 0.1 higher, 95% confidence interval (CI) 1.02–1.08, P = 0.003] after adjustment for LVEF, which itself was inversely related with the risk of hospital mortality (adjusted OR 0.96 per 5% higher, 95% CI 0.93–0.99, P = 0.02). Patients with available LVEF data (n = 8242) were divided based on the presence of predicted (by AI-ECG) vs. observed (by TTE) LVSD (defined as LVEF ≤ 35%), using TTE as the gold standard. A stepwise increase in hospital mortality was observed for patients with a true negative, false positive, false negative, and true positive AI-ECG. Conclusion The AI-ECG prediction of LVSD is associated with hospital mortality in CICU patients, affording risk stratification in addition to that provided by echocardiographic LVEF. Our results emphasize the prognostic value of electrocardiographic patterns reflecting underlying myocardial disease that are recognized by the AI-ECG.

Study Design. Retrospective analysis of a prospective registry. Objective. We utilized the Quality Outcomes Database (QOD) registry to investigate the “July Effect” at QOD spondylolisthesis module sites with residency trainees. Summary of Background Data. There is a paucity of investigation on the long-term outcomes following surgeries involving new trainees utilizing high-quality, prospectively collected data. Methods. This was an analysis of 608 patients who underwent single-segment surgery for grade 1 degenerative lumbar spondylolisthesis at 12 high-enrolling sites. Surgeries were classified as occurring in July or not in July (non-July). Outcomes collected included estimated blood loss, length of stay, operative time, discharge disposition, complications, reoperation and readmission rates, and patient-reported outcomes (Oswestry Disability Index [ODI], Numeric Rating Scale [NRS] Back Pain, NRS Leg Pain, EuroQol-5D [EQ-5D] and the North American Spine Society [NASS] Satisfaction Questionnaire). Propensity score-matched analyses were utilized to compare postoperative outcomes and complication rates between the July and non-July groups. Results. Three hundred seventy-one surgeries occurred at centers with a residency training program with 21 (5.7%) taking place in July. In propensity score-matched analyses, July surgeries were associated with longer operative times ( average treatment effect = 22.4 minutes longer, 95% confidence interval 0.9–449.0, P = 0.041). Otherwise, July surgeries were not associated with significantly different outcomes for the remaining perioperative parameters (estimated blood loss, length of stay, discharge disposition, postoperative complications), overall reoperation rates, 3-month readmission rates, and 24-month ODI, NRS back pain, NRS leg pain, EQ-5D, and NASS satisfaction score (P > 0.05, all comparisons). Conclusion. Although July surgeries were associated with longer operative times, there were no associations with other clinical outcomes compared to non-July surgeries following lumbar spondylolisthesis surgery. These findings may be due to the increased attending supervision and intraoperative education during the beginning of the academic year. There is no evidence that the influx of new trainees in July significantly affects long-term patient-centered outcomes. Level of Evidence: 3

Abstract Microglia are innate immune cells of the central nervous system that sense extracellular cues. Brain injuries, inflammation, and pathology evoke dynamic structural responses in microglia, altering their morphology and motility. The dynamic motility of microglia is hypothesized to be a critical first step in sensing local alterations and engaging in pattern‐specific responses. Alongside their pathological responses, microglia also sense and regulate neuronal activity. In this review, we consider the extracellular molecules, receptors, and mechanisms that allow microglia to sense neuronal activity changes under both hypoactivity and hyperactivity. We also highlight emerging in vivo evidence that microglia regulate neuronal activity, ranging from physiological to pathophysiological conditions. In addition, we discuss the emerging role of calcium signaling in microglial responses to the extracellular environment. The dynamic function of microglia in monitoring and influencing neuronal activity may be critical for brain homeostasis and circuit modification in health and disease.

Study Design. Retrospective analysis of prospectively collected registry data. Objective. The aim of this study was to compare the performance of 30% reduction to established absolute point-change values for measures of disability and pain in patients undergoing elective cervical spine surgery. Summary of Background Data. Recent studies recommend using a proportional change from baseline instead of an absolute point-change value to define minimum clinically important difference (MCID). Methods. Analyses included 13,179 patients who underwent cervical spine surgery for degenerative disease between April 2013 and February 2018. Participants completed a baseline and 12-month follow-up assessment that included questionnaires to assess disability (Neck Disability Index [NDI]), neck and arm pain (Numeric Rating Scale [NRS-NP/AP], and satisfaction [NASS scale]). Participants were classified as met or not met 30% reduction from baseline in each of the respective measures. The 30% reduction in scores at 12 months was compared to a wide range of established absolute point-change MCID values using receiver-operating characteristic curves, area under the receiver-operating characteristic curve (AUROC), and logistic regression analyses. These analyses were conducted for the entire patient cohort, as well as for subgroups based on baseline severity and surgical approach. Results. Thirty percent reduction in NDI and NRS-NP/AP scores predicted satisfaction with more accuracy than absolute point-change values for the total population and ACDF and posterior fusion procedures (P < 0.05). The largest AUROC differences, in favor of 30% reduction, were found for the lowest disability (ODI 0–20%: 16.8%) and bed-bound disability (ODI 81%–100%: 16.6%) categories. For pain, there was a 1.9% to 11% and 1.6% to 9.6% AUROC difference for no/mild neck and arm pain (NRS 0–4), respectively, in favor of a 30% reduction threshold. Conclusion. A 30% reduction from baseline is a valid method for determining MCID in disability and pain for patients undergoing cervical spine surgery. Level of Evidence: 3

AbstractBackgroundAlthough semantic dementia (SMD) presents with language decline, patients often develop behavioral changes. Behavioral symptoms in SMD are attributed to both cortical (orbitofrontal cortex atrophy) and subcortical (nucleus accumbens) degeneration in imaging and neuropathological studies. The caudate nucleus has been implicated in early behavioral changes in some phenotypes of frontotemporal dementia. The objective of this study was to determine whether asymmetric ipsilateral caudate nucleus hypometabolism on [18F]‐fluorodeoxyglucose positron emission tomography (FDG‐PET) is associated with early behavioral changes in the absence of frontal hypometabolism in SMD.MethodFifteen patients with a diagnosis of SMD evaluated by the Neurodegenerative Research Group (NRG) at the Mayo Clinic, Rochester from 2006‐2016, who had FDG‐PET, were included in this study. Clinical data was extracted from the medical records including age at presentation, sex, behavioral symptoms and duration of symptoms. Automated average Z‐scores generated from CortexID (GE Healthcare, WI, USA) using 3D stereotactic surface projections were obtained for the following regions bilaterally: temporal lobes, frontal lobes and caudate nucleus. Patients were categorized as left or right hemisphere lateralized based on the higher temporal lobe Z‐scores. Caudate nucleus and frontal lobe Z‐scores were then compared between the two sides for each patient. Hypometabolism was considered present if Z‐scores were greater than 2.ResultOf our cohort, 9/15 (60%) were females with a median age of 68 at presentation and had 3.5 years symptom duration. 13/15 had at least 1 behavioral symptom at presentation. Eleven patients were considered to have left‐lateralized whereas 4 were right‐lateralized in their temporal lobe hypometabolism. Median Z‐scores were significantly different for ipsilateral and contralateral caudate nuclei at 2.5 and 1.6, respectively (p=0.02, Wilcoxon nonparametric test). Frontal lobe median Z‐scores were 1.2 (ipsilateral) and 0.59 (contralateral) with no statistical difference between two sides (p=0.13).ConclusionIn SMD patients who exhibit behavioral symptoms with no frontal hypometabolism on FDG‐PET, asymmetric caudate nuclei hypometabolism should be explored as a possible imaging marker. This implicates abnormal frontal‐striatal circuitry in the development of behavioral symptoms even in the absence of frontal hypometabolism in early behavioral changes in SMD.

AbstractBackgroundData shared for research use must be de‐identified to protect participant confidentiality. A recent study has shown that face imagery from brain imaging studies could be used to re‐identify anonymous participants via automatic face recognition, linking them with all study‐collected protected health information. Here we measure how well existing software for removing MRI face imagery protects against this privacy threat.MethodWe recruited 153 participants with existing 3D FLAIR MRI and photographed their face from 5 varying angles. We used surf_ice to reconstruct each participant’s face from their MRI and tested whether free, publicly‐available face recognition software (Microsoft Azure) could match their photos to their correct MRI. We tested two existing programs for removing face voxels in MRI: mri_deface (FreeSurfer) and pydeface (python package), and one in‐house implementation. We observed that existing methods often left participants’ eyes intact while removing most of their nose/mouth/chin. This prevented Azure from detecting the reconstructions as faces, therefore preventing its attempt of face recognition. We hypothesized that the retained eye imagery alone could be sufficient for face recognition, if a face were detected. Therefore, we also reconstructed each de‐faced MRI after applying novel pre‐processing that replaced any removed/missing face voxels with those from our population‐average template, e.g. replacing missing noses/mouths/chins with population‐average noses/mouths/chins, before testing face recognition.ResultsWithout any MRI de‐facing, Microsoft Azure correctly matched 78% of participants’ photos to their MRI. After mri_deface and pydeface, 8% and 6% were still identified, respectively. After preprocessing the de‐faced MRI to replace missing areas with those from a population‐average template prior to face reconstruction, the face imagery remaining in the de‐faced images was enough to allow re‐identification of 23% and 20% of participants, respectively. With an in‐house face removal implementation that removes the whole face from every image, 0% of faces were re‐identified from the standard reconstructions, and 5% were identified from the “missing‐parts‐replacement” reconstructions.ConclusionsEven after applying popular MRI de‐facing software, Microsoft Azure face recognition automatically re‐identified up to 23% of participants. Better de‐facing software is needed to remove more of the face more reliably, without affecting brain measurements from de‐faced images.

AbstractBackgroundTau deposition is a key biological feature of Alzheimer’s disease (AD) that is closely related to cognitive impairment. However, it remains poorly understood why certain individuals may be more susceptible to tau accumulation while others are more resistant. The recent availability of tau positron emission tomography (tau‐PET) provides an opportunity to test the hypothesis that genetic factors are associated with differential tau deposition.MethodsWe performed a genome‐wide association study (GWAS) of tau‐PET on a sample of 754 individuals over age 50 (mean age 72.4 years, 54.6% men, 87.4% cognitively unimpaired) from the population‐based Mayo Clinic Study of Aging. Following genotyping with the Illumina GSA array and standard quality control, 515,206 single nucleotide polymorphisms (SNPs) were tested for association with entorhinal cortex 18F‐flortaucipir tau‐PET burden using linear regression under an additive genetic model and covarying for age and sex.ResultA genome‐wide significant association was identified for rs75546066, an intergenic SNP on chromosome 9 between MED27 (mediator complex subunit 27) and NTNG2 (netrin G2), with the minor (A) allele associated with lower tau burden (p=2.85 x 10‐8, β=‐0.48). Linkage disequilibrium and public expression quantitative trait loci (eQTL) data suggested a functional relationship between rs75546066 and NTNG2, the latter of which is highly expressed in the brain. In addition to this novel finding, three SNPs within MAPT (microtubule‐associated protein tau) displayed nominal associations to tau‐PET burden, including rs3785883 (p=0.04, β=0.07) which was previously associated with higher cerebrospinal fluid tau in an independent cohort. No associations with tau‐PET burden were identified for the APOE (apolipoprotein E) ɛ4 or ɛ2 alleles or for genotyped SNPs previously associated with AD in large case‐control studies.ConclusionTo our knowledge, this study represents the first reported neuroimaging GWAS of tau pathology. Our findings confirm the association of MAPT with tau burden and nominate the rs75546066‐NTNG2 locus as a novel potential protective influence against tau pathology which warrants further functional characterization. Our data also supports the inference that tau accumulation may have a genetic architecture distinct from APOE and other known AD susceptibility genes, which may have implications for improved risk stratification and therapeutic targeting models.

Purpose To assess whether an asymmetry exists in the prevalence of carotid artery intraplaque hemorrhage (IPH) between right- and left-sided arteries. Materials and methods The records of all patients with atherosclerotic carotid artery disease that underwent neck magnetic resonance angiography imaging with high-resolution plaque sequences between 2017 and 2020 at our institution were retrospectively reviewed. The prevalence of stenosis and IPH was determined for all patients and compared between the left and right carotid arteries of those with unilateral anterior circulation ischemic strokes. Multiple regression analysis was performed to determine potential independent associations of IPH laterality with ischemic strokes. Results A total of 368 patients were included overall and 241 were male (65.4%). There were a total of 125 asymptomatic patients and 211 patients with unilateral anterior circulation ischemic strokes. Of patients with ischemic strokes, 55.5% had left-sided strokes compared with 44.5% who had right-sided strokes ( p = 0.03). Patients with left-sided strokes had a higher prevalence of ipsilateral IPH than those with right-sided strokes (64.1% versus 36.2%, p < 0.0001), despite similar degrees of stenosis. Both age (odds ratio (OR): 1.0; 95% confidence interval (CI): 1.0–1.1; p = 0.007) and the presence of left-sided IPH (OR: 3.2; 95% CI: 1.5–6.8; p = 0.003) were independently associated with unilateral ischemic strokes. Conclusions Left-sided plaques more frequently have IPH and may be more likely to result in ipsilateral ischemic strokes compared with right-sided plaques. The underlying mechanism of asymmetric distribution of IPH between right and left carotids remains unclear.