Anisocoria and Poor Pupil Reactivity by Quantitative Pupillometry in Patients With Intracranial Pathology.

OBJECTIVES: To describe the prevalence and associated risk factors of new onset anisocoria (new pupil size difference of at least 1 mm) and its subtypes: new onset anisocoria accompanied by abnormal and normal pupil reactivities in patients with acute neurologic injuries. DESIGN: We tested the association of patients who experienced new onset anisocoria subtypes with degree of midline shift using linear regression. We further explored differences between quantitative pupil characteristics associated with first-time new onset anisocoria and nonnew onset anisocoria at preceding observations using mixed effects logistic regression, adjusting for possible confounders. SETTING: All quantitative pupil observations were collected at two neuro-ICUs by nursing staff as standard of care. PATIENTS: We conducted a retrospective two-center study of adult patients with intracranial pathology in the ICU with at least a 24-hour stay and three or more quantitative pupil measurements between 2016 and 2018. MEASUREMENTS AND MAIN RESULTS: We studied 221 patients (mean age 58, 41% women). Sixty-three percent experienced new onset anisocoria. New onset anisocoria accompanied by objective evidence of abnormal pupil reactivity occurring at any point during hospitalization was significantly associated with maximum midline shift (ß = 2.27 per mm; p = 0.01). The occurrence of new onset anisocoria accompanied by objective evidence of normal pupil reactivity was inversely associated with death (odds ratio, 0.34; 95% CI, 0.16-0.71; p = 0.01) in adjusted analyses. Subclinical continuous pupil size difference distinguished first-time new onset anisocoria from nonnew onset anisocoria in up to four preceding pupil observations (or up to 8 hr prior). Minimum pupil reactivity between eyes also distinguished new onset anisocoria accompanied by objective evidence of abnormal pupil reactivity from new onset anisocoria accompanied by objective evidence of normal pupil reactivity prior to first-time new onset anisocoria occurrence. CONCLUSIONS: New onset anisocoria occurs in over 60% of patients with neurologic emergencies. Pupil reactivity may be an important distinguishing characteristic of clinically relevant new onset anisocoria phenotypes. New onset anisocoria accompanied by objective evidence of abnormal pupil reactivity was associated with midline shift, and new onset anisocoria accompanied by objective evidence of normal pupil reactivity had an inverse relationship with death. Distinct quantitative pupil characteristics precede new onset anisocoria occurrence and may allow for earlier prediction of neurologic decline. Further work is needed to determine whether quantitative pupillometry sensitively/specifically predicts clinically relevant anisocoria, enabling possible earlier treatments.

Natural Language Processing of Radiology Reports to Detect Complications of Ischemic Stroke.

BACKGROUND: Abstraction of critical data from unstructured radiologic reports using natural language processing (NLP) is a powerful tool to automate the detection of important clinical features and enhance research efforts. We present a set of NLP approaches to identify critical findings in patients with acute ischemic stroke from radiology reports of computed tomography (CT) and magnetic resonance imaging (MRI). METHODS: We trained machine learning classifiers to identify categorical outcomes of edema, midline shift (MLS), hemorrhagic transformation, and parenchymal hematoma, as well as rule-based systems (RBS) to identify intraventricular hemorrhage (IVH) and continuous MLS measurements within CT/MRI reports. Using a derivation cohort of 2289 reports from 550 individuals with acute middle cerebral artery territory ischemic strokes, we externally validated our models on reports from a separate institution as well as from patients with ischemic strokes in any vascular territory. RESULTS: In all data sets, a deep neural network with pretrained biomedical word embeddings (BioClinicalBERT) achieved the highest discrimination performance for binary prediction of edema (area under precision recall curve [AUPRC] > 0.94), MLS (AUPRC > 0.98), hemorrhagic conversion (AUPRC > 0.89), and parenchymal hematoma (AUPRC > 0.76). BioClinicalBERT outperformed lasso regression (p < 0.001) for all outcomes except parenchymal hematoma (p = 0.755). Tailored RBS for IVH and continuous MLS outperformed BioClinicalBERT (p < 0.001) and linear regression, respectively (p < 0.001). CONCLUSIONS: Our study demonstrates robust performance and external validity of a core NLP tool kit for identifying both categorical and continuous outcomes of ischemic stroke from unstructured radiographic text data. Medically tailored NLP methods have multiple important big data applications, including scalable electronic phenotyping, augmentation of clinical risk prediction models, and facilitation of automatic alert systems in the hospital setting.

Development and validation of an interpretable deep learning framework for Alzheimer's disease classification.

Alzheimer's disease is the primary cause of dementia worldwide, with an increasing morbidity burden that may outstrip diagnosis and management capacity as the population ages. Current methods integrate patient history, neuropsychological testing and MRI to identify likely cases, yet effective practices remain variably applied and lacking in sensitivity and specificity. Here we report an interpretable deep learning strategy that delineates unique Alzheimer's disease signatures from multimodal inputs of MRI, age, gender, and Mini-Mental State Examination score. Our framework linked a fully convolutional network, which constructs high resolution maps of disease probability from local brain structure to a multilayer perceptron and generates precise, intuitive visualization of individual Alzheimer's disease risk en route to accurate diagnosis. The model was trained using clinically diagnosed Alzheimer's disease and cognitively normal subjects from the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset (n = 417) and validated on three independent cohorts: the Australian Imaging, Biomarker and Lifestyle Flagship Study of Ageing (AIBL) (n = 382), the Framingham Heart Study (n = 102), and the National Alzheimer's Coordinating Center (NACC) (n = 582). Performance of the model that used the multimodal inputs was consistent across datasets, with mean area under curve values of 0.996, 0.974, 0.876 and 0.954 for the ADNI study, AIBL, Framingham Heart Study and NACC datasets, respectively. Moreover, our approach exceeded the diagnostic performance of a multi-institutional team of practicing neurologists (n = 11), and high-risk cerebral regions predicted by the model closely tracked post-mortem histopathological findings. This framework provides a clinically adaptable strategy for using routinely available imaging techniques such as MRI to generate nuanced neuroimaging signatures for Alzheimer's disease diagnosis, as well as a generalizable approach for linking deep learning to pathophysiological processes in human disease.

Differential Assessment of Internal Jugular Vein Stenosis in Patients Undergoing CT and MRI with Contrast.

OBJECTIVE: Internal Jugular Vein Stenosis (IJVS) is hypothesized to play a role in the pathogenesis of diverse neurological diseases. We sought to evaluate differences in IJVS assessment between CT and MRI in a retrospective patient cohort. METHODS: We included consecutive patients who had both MRI of the brain and CT of the head and neck with contrast from 1 June 2021 to 30 June 2022 within the same admission. The degree of IJVS was categorized into five grades (0-IV). RESULTS: A total of 35 patients with a total of 70 internal jugular (IJ) veins were included in our analysis. There was fair intermodality agreement in stenosis grades (κ = 0.220, 95% C.I. = [0.029, 0.410]), though categorical stenosis grades were significantly discordant between imaging modalities, with higher grades more frequent in MRI (χ2 = 27.378, p = 0.002). On CT-based imaging, Grade III or IV stenoses were noted in 17/70 (24.2%) IJs, whereas on MRI-based imaging, Grade III or IV stenoses were found in 40/70 (57.1%) IJs. Among veins with Grade I-IV IJVS, MRI stenosis estimates were significantly higher than CT stenosis estimates (77.0%, 95% C.I. [35.9-55.2%] vs. 45.6%, 95% C.I. [35.9-55.2%], p < 0.001). CONCLUSION: MRI with contrast overestimates the degree of IJVS compared to CT with contrast. Consideration of this discrepancy should be considered in diagnosis and treatment planning in patients with potential IJVS-related symptoms.

Contrast Staining in Noninfarcted Tissue after Endovascular Treatment of Acute Ischemic Stroke.

BACKGROUND AND PURPOSE: Contrast staining is a common finding after endovascular treatment of acute ischemic stroke. It typically occurs in infarcted tissue and is considered an indicator of irreversible brain damage. Contrast staining in noninfarcted tissue has not been systematically investigated. We sought to assess the incidence, risk factors, and clinical significance of contrast staining in noninfarcted tissue after endovascular treatment. MATERIALS AND METHODS: We conducted a retrospective review of consecutive patients who underwent endovascular treatment for anterior circulation large-vessel occlusion acute ischemic stroke. Contrast staining, defined as new hyperdensity on CT after endovascular treatment, was categorized as either contrast staining in infarcted tissue if the stained region demonstrated restricted diffusion on follow-up MR imaging or contrast staining in noninfarcted tissue if the stained region demonstrated no restricted diffusion. Baseline differences between patients with and without contrast staining in noninfarcted tissue were compared. Logistic regression was used to identify independent associations for contrast staining in noninfarcted tissue after endovascular treatment. RESULTS: Among 194 patients who underwent endovascular treatment for large-vessel occlusion acute ischemic stroke and met the inclusion criteria, contrast staining in infarcted tissue was noted in 52/194 (26.8%) patients; contrast staining in noninfarcted tissue, in 26 (13.4%) patients. Both contrast staining in infarcted tissue and contrast staining in noninfarcted tissue were noted in 5.6% (11/194). Patients with contrast staining in noninfarcted tissue were found to have a higher likelihood of having an ASPECTS of 8-10, to be associated with contrast staining in infarcted tissue, and to achieve successful reperfusion compared with those without contrast staining in noninfarcted tissue. In contrast staining in noninfarcted tissue regions, the average attenuation was 40 HU, significantly lower than the contrast staining in infarcted tissue regions (53 HU). None of the patients with contrast staining in noninfarcted tissue had clinical worsening during their hospital stay. The median discharge mRS was significantly lower in patients with contrast staining in noninfarcted tissue than in those without (3 versus 4; P = .018). No independent predictors of contrast staining in noninfarcted tissue were found. CONCLUSIONS: Contrast staining can be seen outside the infarcted tissue after endovascular treatment of acute ischemic stroke, likely attributable to the reversible disruption of the BBB in ischemic but not infarcted tissue. While generally benign, understanding its characteristics is important because it may mimic pathologic conditions such as infarcted tissue and cerebral edema.

Machine Learning in Clinical Trials: A Primer with Applications to Neurology.

We reviewed foundational concepts in artificial intelligence (AI) and machine learning (ML) and discussed ways in which these methodologies may be employed to enhance progress in clinical trials and research, with particular attention to applications in the design, conduct, and interpretation of clinical trials for neurologic diseases. We discussed ways in which ML may help to accelerate the pace of subject recruitment, provide realistic simulation of medical interventions, and enhance remote trial administration via novel digital biomarkers and therapeutics. Lastly, we provide a brief overview of the technical, administrative, and regulatory challenges that must be addressed as ML achieves greater integration into clinical trial workflows.

Multimodal deep learning for Alzheimer's disease dementia assessment.

Worldwide, there are nearly 10 million new cases of dementia annually, of which Alzheimer's disease (AD) is the most common. New measures are needed to improve the diagnosis of individuals with cognitive impairment due to various etiologies. Here, we report a deep learning framework that accomplishes multiple diagnostic steps in successive fashion to identify persons with normal cognition (NC), mild cognitive impairment (MCI), AD, and non-AD dementias (nADD). We demonstrate a range of models capable of accepting flexible combinations of routinely collected clinical information, including demographics, medical history, neuropsychological testing, neuroimaging, and functional assessments. We then show that these frameworks compare favorably with the diagnostic accuracy of practicing neurologists and neuroradiologists. Lastly, we apply interpretability methods in computer vision to show that disease-specific patterns detected by our models track distinct patterns of degenerative changes throughout the brain and correspond closely with the presence of neuropathological lesions on autopsy. Our work demonstrates methodologies for validating computational predictions with established standards of medical diagnosis.

Utilization of mental health services in pediatric patients surviving penetrating trauma resulting from interpersonal violence.

BACKGROUND: Violent trauma has lasting psychological impacts. Our institution's Community Violence Response Team (CVRT) offers mental health services to trauma victims. We characterized implementation and determined factors associated with utilization by pediatric survivors of interpersonal violence-related penetrating trauma. METHODS: Analysis included survivors (0-21 years) of violent penetrating injury at our institution (2011-2017). Injury and demographic data were collected. Nonparametric regression models determined factors associated with utilization. RESULTS: There was initial rapid uptake of CVRT (2011-2013) after which it plateaued, serving >80% of eligible patients (2017). White race and higher injury severity were associated with receipt and duration of services. In post-hoc analysis, race was found to be associated with continued treatment but not with initial consultation. CONCLUSION: Successful implementation required three years, aiding >80% of patients. CVRT is a blueprint to strengthen existing violence intervention programs. Efforts should be made to ensure that barriers to providing care, including those related to race, are overcome.

Delayed lung expansion after decortication in a case of trapped lung resulting from catamenial haemothorax.

Herein, we report the case of a 35-year-old female with a trapped right lung secondary to catamenial haemothorax. Following surgical decortication, re-expansion of the lung was not observed until postoperative day 81. This delay represents a heretofore unencountered complication that should be considered in the surgical management of catamenial haemothorax due to thoracic endometriosis syndrome.

Peripheral Blood Lymphocytes and Platelets Are Prognostic in Surgical pT1 Non-Small Cell Lung Cancer.

BACKGROUND: There is a paucity of prognostic factors for patients with stage I non-small cell lung cancer (NSCLC) undergoing operations. We investigated the prognostic role of preoperative complete blood count values in patients with stage I NSCLC patients undergoing operations. METHODS: A retrospective medical record review was performed of patients who underwent operations for stage I NSCLC between 2000 and 2015. Patients who died within 30 days of the operations were excluded. The primary end point was recurrence. Preoperative complete blood count values were analyzed, and a median value was used as the cutoff. Statistical analysis used χ2 and t tests along with univariate and multivariate analyses by Cox regression modeling. RESULTS: The study included 103 patients. A high lymphocyte count was significantly associated with recurrence (5-year recurrence-free survival [RFS] of 69.8% for high vs 95.7% for low, P = .003), as well as high platelet (5-year RFS of 72.0% for high vs 91.8% for low, P = .02). Independent prognostic factors on multivariate analysis were high lymphocyte (hazard ratio [HR], 7.27; P = .005) and platelet counts (HR, 7.49; P = .003) as well as tumor (HR, 5.40; P = .008) and treatment characteristics (HR, 4.59; P = .01). Among patients with pT1 lesions, high lymphocyte (HR, 8.41; P = .03) and high platelet counts (HR, 19.78; P = .004) remained independent prognostic factors. Neither NLR nor PLR were significantly associated with recurrence. CONCLUSIONS: In patients with pathologic stage I NSCLC undergoing surgical resection, the preoperative blood count from peripheral blood may provide prognostic value. Of significance, in patients with pT1 N0 NSCLC, high lymphocyte count and high platelet count were associated with higher recurrence.

Machine learning and natural language processing methods to identify ischemic stroke, acuity and location from radiology reports.

Accurate, automated extraction of clinical stroke information from unstructured text has several important applications. ICD-9/10 codes can misclassify ischemic stroke events and do not distinguish acuity or location. Expeditious, accurate data extraction could provide considerable improvement in identifying stroke in large datasets, triaging critical clinical reports, and quality improvement efforts. In this study, we developed and report a comprehensive framework studying the performance of simple and complex stroke-specific Natural Language Processing (NLP) and Machine Learning (ML) methods to determine presence, location, and acuity of ischemic stroke from radiographic text. We collected 60,564 Computed Tomography and Magnetic Resonance Imaging Radiology reports from 17,864 patients from two large academic medical centers. We used standard techniques to featurize unstructured text and developed neurovascular specific word GloVe embeddings. We trained various binary classification algorithms to identify stroke presence, location, and acuity using 75% of 1,359 expert-labeled reports. We validated our methods internally on the remaining 25% of reports and externally on 500 radiology reports from an entirely separate academic institution. In our internal population, GloVe word embeddings paired with deep learning (Recurrent Neural Networks) had the best discrimination of all methods for our three tasks (AUCs of 0.96, 0.98, 0.93 respectively). Simpler NLP approaches (Bag of Words) performed best with interpretable algorithms (Logistic Regression) for identifying ischemic stroke (AUC of 0.95), MCA location (AUC 0.96), and acuity (AUC of 0.90). Similarly, GloVe and Recurrent Neural Networks (AUC 0.92, 0.89, 0.93) generalized better in our external test set than BOW and Logistic Regression for stroke presence, location and acuity, respectively (AUC 0.89, 0.86, 0.80). Our study demonstrates a comprehensive assessment of NLP techniques for unstructured radiographic text. Our findings are suggestive that NLP/ML methods can be used to discriminate stroke features from large data cohorts for both clinical and research-related investigations.