Quantitative epileptiform burden and electroencephalography background features predict post-traumatic epilepsy
Chen Y, Li S, Ge W, Jing J, Chen HY, Doherty D, Herman A, Kaleem S, Ding K, Osman G, Swisher CB, Smith C, Maciel CB, Alkhachroum A, Lee JW, Dhakar MB, Gilmore EJ, Sivaraju A, Hirsch LJ, Omay SB, Blumenfeld H, Sheth KN, Struck AF, Edlow BL, Westover MB, Kim JA. Quantitative epileptiform burden and electroencephalography background features predict post-traumatic epilepsy. Journal Of Neurology Neurosurgery & Psychiatry 2022, 94: 245-249. PMID: 36241423, PMCID: PMC9931627, DOI: 10.1136/jnnp-2022-329542.Peer-Reviewed Original ResearchConceptsPost-traumatic epilepsyTraumatic brain injuryCT abnormalitiesElectroencephalography featuresAdmission Glasgow Coma Scale scoreGlasgow Coma Scale scoreRetrospective case-control studyMultivariable logistic regression analysisRisk stratification modelCase-control studyLogistic regression analysisTBI admissionsSevere complicationsFuture trialsBrain injuryCT reportsSeizure diagnosisPredictive valueScale scorePatientsLogistic regressionStratification modelQuantitative electroencephalogramTBI mechanismsRegression analysisDevelopment and Validation of a Model to Identify Critical Brain Injuries Using Natural Language Processing of Text Computed Tomography Reports
Torres-Lopez VM, Rovenolt GE, Olcese AJ, Garcia GE, Chacko SM, Robinson A, Gaiser E, Acosta J, Herman AL, Kuohn LR, Leary M, Soto AL, Zhang Q, Fatima S, Falcone GJ, Payabvash MS, Sharma R, Struck AF, Sheth KN, Westover MB, Kim JA. Development and Validation of a Model to Identify Critical Brain Injuries Using Natural Language Processing of Text Computed Tomography Reports. JAMA Network Open 2022, 5: e2227109. PMID: 35972739, PMCID: PMC9382443, DOI: 10.1001/jamanetworkopen.2022.27109.Peer-Reviewed Original ResearchConceptsNatural language processingF-scoreTest data setsLanguage processingIndependent test data setsData setsBidirectional Encoder RepresentationsAcute brain injuryLarge data setsHead CTBrain injuryNLP toolsF1 scoreNER modelTransformer architectureClinical textEncoder RepresentationsNLP algorithmNLP modelsCT reportsCustom dictionaryTraining setCross-validation performancePerformance metricsAvailable new tools