Search

ObjectiveTo optimally leverage the scalability and unique features of the electronic health records (EHR) for research that would ultimately improve patient care, we need to accurately identify patients and extract clinically meaningful measures. Using multiple sclerosis (MS) as a proof of principle, we showcased how to leverage routinely collected EHR data to identify patients with a complex neurological disorder and derive an important surrogate measure of disease severity heretofore only available in research settings.MethodsIn a cross-sectional observational study, 5,495 MS patients were identified from the EHR systems of two major referral hospitals using an algorithm that includes codified and narrative information extracted using natural language processing. In the subset of patients who receive neurological care at a MS Center where disease measures have been collected, we used routinely collected EHR data to extract two aggregate indicators of MS severity of clinical relevance multiple sclerosis severity score (MSSS) and brain parenchymal fraction (BPF, a measure of whole brain volume).ResultsThe EHR algorithm that identifies MS patients has an area under the curve of 0.958, 83% sensitivity, 92% positive predictive value, and 89% negative predictive value when a 95% specificity threshold is used. The correlation between EHR-derived and true MSSS has a mean R(2) = 0.38±0.05, and that between EHR-derived and true BPF has a mean R(2) = 0.22±0.08. To illustrate its clinical relevance, derived MSSS captures the expected difference in disease severity between relapsing-remitting and progressive MS patients after adjusting for sex, age of symptom onset and disease duration (p = 1.56×10(-12)).ConclusionIncorporation of sophisticated codified and narrative EHR data accurately identifies MS patients and provides estimation of a well-accepted indicator of MS severity that is widely used in research settings but not part of the routine medical records. Similar approaches could be applied to other complex neurological disorders.

IntroductionThe purpose of this study is to compare the variability of PCT results obtained by automatic selection of the arterial input function (AIF), venous output function (VOF) and symmetry axis versus manual selection.MethodsImaging data from 30 PCT studies obtained as part of standard clinical stroke care at our institution in patients with suspected acute hemispheric ischemic stroke were retrospectively reviewed. Two observers performed the post-processing of 30 CTP datasets. Each observer processed the data twice, the first time employing manual selection of AIF, VOF and symmetry axis, and a second time using automated selection of these same parameters, with the user being allowed to adjust them whenever deemed appropriate. The volumes of infarct core and of total perfusion defect were recorded. The cerebral blood volume (CBV), cerebral blood flow (CBF), mean transit time (MTT) and blood-brain barrier permeability (BBBP) values in standardized regions of interest were recorded. Interobserver variability was quantified using the Bland and Altman's approach.ResultsAutomated post-processing yielded lower coefficients of variation for the volume of the infarct core and the volume of the total perfusion defect (15.7% and 5.8%, respectively) compared to manual post-processing (31.0% and 12.2%, respectively). Automated post-processing yielded lower coefficients of variation for PCT values (11.3% for CBV, 9.7% for CBF, and 9.5% for MTT) compared to manual post-processing (23.7% for CBV, 32.8% for CBF, and 16.7% for MTT).ConclusionAutomated post-processing of PCT data improves interobserver agreement in measurements of CBV, CBF and MTT, as well as volume of infarct core and penumbra.

ABSTRACT BACKGROUND AND PURPOSE From the literature, the prevalence of fluorodeoxyglucose (FDG) uptake in large artery atherosclerotic plaques shows great heterogeneity. We retrospectively reviewed 100 consecutive patients who underwent FDG-positron emission tomography-computed tomography (PET/CT) imaging of their whole body, to evaluate FDG uptake in the arterial wall. MATERIALS AND METHODS We retrospectively evaluated 100 whole-body PET-CT scans. The PET images coregistered with CT were reviewed for abnormal 18F-FDG uptake. The mean standard uptake value (SUV) was measured in regions of interest (ROIs). The prevalence of PET+ plaques was determined based on the qualitative PET review, used as the gold standard in a receiver-operating characteristic (ROC) curve analysis to determine an optimal threshold for the quantitative PET analysis. RESULTS The qualitative, visual assessment demonstrated FDG uptake in the arterial walls of 26 patients. A total of 85 slices exhibited FDG uptake within the arterial wall of 37 artery locations. 11, 17, and 2 patients exhibited FDG uptake within the wall of carotid arteries, of the aorta, and of the iliac arteries, respectively. Only 4 of the 26 patients had positive FDG uptake in more than one artery location. In terms of quantitative analysis, a threshold of 2.8 SUV was associated with a negative predictive value of 99.4% and a positive predictive value of 100% to predict qualitative PET+ plaques. A threshold of 1.8 SUV was associated with a negative predictive value of 100% and a positive predictive value of 99.4%. Area under the ROC curve was .839. CONCLUSION The prevalence of PET uptake in arterial walls in a consecutive population of asymptomatic patients is low and usually confined to one type of artery, and its clinical relevance in terms of vulnerability to ischemic events remains to be determined. [ABSTRACT FROM AUTHOR]

Though electronic health record (EHR) systems are a rich repository of clinical information with large potential, the use of EHR-based phenotyping algorithms is often hindered by inaccurate diagnostic records, the presence of many irrelevant features, and the requirement for a human-labeled training set. In this paper, we describe a knowledge-driven online multimodal automated phenotyping (KOMAP) system that i) generates a list of informative features by an online narrative and codified feature search engine (ONCE) and ii) enables the training of a multimodal phenotyping algorithm based on summary data. Powered by composite knowledge from multiple EHR sources, online article corpora, and a large language model, features selected by ONCE show high concordance with the state-of-the-art AI models (GPT4 and ChatGPT) and encourage large-scale phenotyping by providing a smaller but highly relevant feature set. Validation of the KOMAP system across four healthcare centers suggests that it can generate efficient phenotyping algorithms with robust performance. Compared to other methods requiring patient-level inputs and gold-standard labels, the fully online KOMAP provides a significant opportunity to enable multi-center collaboration.