Characterizing Tissue Remodeling and Mechanical Heterogeneity in Cerebral Aneurysms

Accurately assessing the complex tissue mechanics of cerebral aneurysms (CAs) is critical for elucidating how CAs grow and whether that growth will lead to rupture. The factors that have been implicated in CA progression – blood flow dynamics, immune infiltration, and extracellular matrix remodeling – all occur heterogeneously throughout the CA. Thus, it stands to reason that the mechanical properties of CAs are also spatially heterogeneous. Here, we present a new method for characterizing the mechanical heterogeneity of human CAs using generalized anisotropic inverse mechanics, which uses biaxial stretching experiments and inverse analyses to determine the local Kelvin moduli and principal alignments within the tissue. Using this approach, we find that there is significant mechanical heterogeneity within a single acquired human CA. These results were confirmed using second harmonic generation imaging of the CA’s fiber architecture and a correlation was observed. This approach provides a single-step method for determining the complex heterogeneous mechanics of CAs, which has important implications for future identification of metrics that can improve accuracy in prediction risk of rupture.

The validity of heart failure diagnoses at hospital-discharge and ambulatory evaluation visits: insights from two Norwegian local hospitals

Background The validity of heart failure (HF) diagnoses made in hospitals has been debated and low positive predictive values (PPV) may represent a bias in epidemiological research. Purpose To validate primary and secondary HF diagnoses at discharge or during ambulatory evaluation in general hospitals aiming to obtain confirmed HF diagnoses to develop a HF-prediction risk score. Methods We extracted data on all patients with a HF diagnosis by ICD-10 codes (I50 HF, I42 cardiomyopathy and I11 hypertension with HF) in any position from the hospitals' electronic medical records from Oct. 2006 to Dec. 2018. One experienced cardiologist scrutinized all journals for events being either a valid HF event, unlikely, or uncertain due to lacking information, according to the 2016 ESC HF guidelines. In cases where first event was unlikely or uncertain subsequent events were judged for valid HF. Results A total of 3411 patients with at least one HF diagnosis were assessed (mean age 79.7±10.6 yrs, 49.1% men); 3089 after in-hospital stays and 322 after ambulatory consultations. Overall, 2174 were deemed as valid HF diagnosis with a PPV of 63.7%; PPV was higher when HF diagnosis was based on in-hospital diagnoses and when HF was the primary diagnosis (Table). Conclusions Only 64% of all HF diagnoses were likely HF according to present guidelines, with higher precision for in-hospital diagnoses and HF in the primary position. This underscores the importance to use validated HF-diagnoses for HF prediction risk score development. FUNDunding Acknowledgement Type of funding sources: Other. Main funding source(s): Boehringer Ingelheim Norway KS

Using Forecasting Methods to Incorporate Social, Economic, and Political Considerations Into Marine Protected Area Planning

As the global environmental crisis grows in scale and complexity, conservation professionals and policymakers are increasingly called upon to make decisions despite high levels of uncertainty, limited resources, and insufficient data. Global efforts to protect biodiversity in areas beyond national jurisdiction require substantial international cooperation and negotiation, both of which are characterized by unpredictability and high levels of uncertainty. Here we build on recent studies to adapt forecasting techniques from the fields of hazard prediction, risk assessment, and intelligence analysis to forecast the likelihood of marine protected area (MPA) designation in the Southern Ocean. We used two questionnaires, feedback, and a discussion round in a Delphi-style format expert elicitation to obtain forecasts, and collected data on specific biophysical, socioeconomic, geopolitical, and scientific factors to assess how they shape and influence these forecasts. We found that areas further north along the Western Antarctic Peninsula were considered to be less likely to be designated than areas further south, and that geopolitical factors, such as global politics or events, and socioeconomic factors, such as the presence of fisheries, were the key determinants of whether an area was predicted to be more or less likely to be designated as an MPA. Forecasting techniques can be used to inform protected area design, negotiation, and implementation in highly politicized situations where data is lacking by aiding with spatial prioritization, targeting scarce resources, and predicting the success of various spatial arrangements, interventions, or courses of action.

An Explainable Machine Learning Model for Early Prediction of Sepsis Using ICU Data

Early identification of individuals with sepsis is very useful in assisting clinical triage and decision-making, resulting in early intervention and improved outcomes. This study aims to develop an explainable machine learning model with the clinical interpretability to predict sepsis onset before 6 hours and validate with improved prediction risk power for every time interval since admission to the ICU. The retrospective observational cohort study is carried out using PhysioNet Challenge 2019 ICU data from three distinct hospital systems, viz. A, B, and C. Data from A and B were shared publicly for training and validation while sequestered data from all three cohorts were used for scoring. However, this study is limited only to publicly available training data. Training data contains 15,52,210 patient records of 40,336 ICU patients with up to 40 clinical variables (sourced for each hour of their ICU stay) divided into two datasets, based on hospital systems A and B. The clinical feature exploration and interpretation for early prediction of sepsis is achieved using the proposed framework, viz. the explainable Machine Learning model for Early Prediction of Sepsis (xMLEPS). A total of 85 features comprising the given 40 clinical variables augmented with 10 derived physiological features and 35 time-lag difference features are fed to xMLEPS for the said prediction task of sepsis onset. A ten-fold cross-validation scheme is employed wherein an optimal prediction risk threshold is searched for each of the 10 LightGBM models. These optimum threshold values are later used by the corresponding models to refine the predictive power in terms of utility score for the prediction of labels in each fold. The entire framework is designed via Bayesian optimization and trained with the resultant feature set of 85 features, yielding an average normalized utility score of 0.4214 and area under receiver operating characteristic curve of 0.8591 on publicly available training data. This study establish a practical and explainable sepsis onset prediction model for ICU data using applied ML approach, mainly gradient boosting. The study highlights the clinical significance of physiological inter-relations among the given and proposed clinical signs via feature importance and SHapley Additive exPlanations (SHAP) plots for visualized interpretation.

Disease and Adverse Drug Reaction Prediction using Machine Learning

The aim of the project is to use machine learning techniques for disease prediction, risk prediction and prediction of adverse drug reactions. The project is divided into two modules, an android app and a web app. The android app is to predict possible diseases based on the symptoms the person is showing. Along with that the reviews of common drugs from online healthcare forums such as medications.com are extracted and tf-idf is used to find out the possible adverse drug reactions the drugs may have. The web app does disease risk prediction based on phenotypic details and lab reports. As an addition to the project, location based medical help and health tips are also implemented.

Development and Validation of a Transfusion Risk Score for Patients Receiving Maintenance Hemodialysis

Background: In dialysis patients with anemia, avoiding red blood cell transfusions is preferable. We sought to develop and validate a novel transfusion prediction risk score for patients receiving maintenance hemodialysis. Methods: This retrospective cohort study used United States Renal Data System data to create a model development cohort (point prevalent hemodialysis patients on November 1, 2012) and a validation cohort (point prevalent hemodialysis patients on August 1, 2013). We characterized comorbidity, inflammatory conditions, hospitalizations, anemia and anemia management, iron parameters, intravenous iron use, and vitamin D use during a 6-month baseline period to predict subsequent 3-month transfusion risk. We used logistic least absolute shrinkage and selection operator regression. In an exploratory analysis, model results were used to calculate a score to predict 6- and 12-month hospitalization and mortality. Results: Variables most predictive of transfusion were prior transfusion, hemoglobin, ferritin, and number of hospital days in the baseline period. The resulting C-statistic in the validation cohort was 0.74, indicating relatively good predictive power. The score was associated with a significantly increased risk of subsequent mortality (hazard ratios 1.0, 1.22, 1.26, 1.54, 1.71 grouped from lowest to highest score), but not with hospitalization. Conclusions: We developed a transfusion prediction risk score with good performance characteristics that was associated with mortality. This score could be further developed into a clinically useful application allowing clinicians to identify hemodialysis patients most likely to benefit from a timely, proactive anemia treatment approach with the goal of avoiding red blood cell transfusions and attendant risks of adverse clinical outcomes.