Blending Knowledge in Deep Recurrent Networks for Adverse Event Prediction at Hospital Discharge.

Deep learning architectures have an extremely high-capacity for modeling complex data in a wide variety of domains. However, these architectures have been limited in their ability to support complex prediction problems using insurance claims data, such as readmission at 30 days, mainly due to data sparsity issue. Consequently, classical machine learning methods, especially those that embed domain knowledge in handcrafted features, are often on par with, and sometimes outperform, deep learning approaches. In this paper, we illustrate how the potential of deep learning can be achieved by blending domain knowledge within deep learning architectures to predict adverse events at hospital discharge, including readmissions. More specifically, we introduce a learning architecture that fuses a representation of patient data computed by a self-attention based recurrent neural network, with clinically relevant features. We conduct extensive experiments on a large claims dataset and show that the blended method outperforms the standard machine learning approaches.

AI-assisted tracking of worldwide non-pharmaceutical interventions for COVID-19.

The Coronavirus disease 2019 (COVID-19) global pandemic has transformed almost every facet of human society throughout the world. Against an emerging, highly transmissible disease, governments worldwide have implemented non-pharmaceutical interventions (NPIs) to slow the spread of the virus. Examples of such interventions include community actions, such as school closures or restrictions on mass gatherings, individual actions including mask wearing and self-quarantine, and environmental actions such as cleaning public facilities. We present the Worldwide Non-pharmaceutical Interventions Tracker for COVID-19 (WNTRAC), a comprehensive dataset consisting of over 6,000 NPIs implemented worldwide since the start of the pandemic. WNTRAC covers NPIs implemented across 261 countries and territories, and classifies NPIs into a taxonomy of 16 NPI types. NPIs are automatically extracted daily from Wikipedia articles using natural language processing techniques and then manually validated to ensure accuracy and veracity. We hope that the dataset will prove valuable for policymakers, public health leaders, and researchers in modeling and analysis efforts to control the spread of COVID-19.

Evaluation of artificial intelligence systems for assisting neurologists with fast and accurate annotations of scalp electroencephalography data.

BACKGROUND: Assistive automatic seizure detection can empower human annotators to shorten patient monitoring data review times. We present a proof-of-concept for a seizure detection system that is sensitive, automated, patient-specific, and tunable to maximise sensitivity while minimizing human annotation times. The system uses custom data preparation methods, deep learning analytics and electroencephalography (EEG) data. METHODS: Scalp EEG data of 365 patients containing 171,745 s ictal and 2,185,864 s interictal samples obtained from clinical monitoring systems were analysed as part of a crowdsourced artificial intelligence (AI) challenge. Participants were tasked to develop an ictal/interictal classifier with high sensitivity and low false alarm rates. We built a challenge platform that prevented participants from downloading or directly accessing the data while allowing crowdsourced model development. FINDINGS: The automatic detection system achieved tunable sensitivities between 75.00% and 91.60% allowing a reduction in the amount of raw EEG data to be reviewed by a human annotator by factors between 142x, and 22x respectively. The algorithm enables instantaneous reviewer-managed optimization of the balance between sensitivity and the amount of raw EEG data to be reviewed. INTERPRETATION: This study demonstrates the utility of deep learning for patient-specific seizure detection in EEG data. Furthermore, deep learning in combination with a human reviewer can provide the basis for an assistive data labelling system lowering the time of manual review while maintaining human expert annotation performance. FUNDING: IBM employed all IBM Research authors. Temple University employed all Temple University authors. The Icahn School of Medicine at Mount Sinai employed Eren Ahsen. The corresponding authors Stefan Harrer and Gustavo Stolovitzky declare that they had full access to all the data in the study and that they had final responsibility for the decision to submit for publication.

Using artificial intelligence to identify anti-hypertensives as possible disease modifying agents in Parkinson's disease.

PURPOSE: Drug repurposing is an effective means of increasing treatment options for diseases, however identifying candidate molecules for the indication of interest from the thousands of approved drugs is challenging. We have performed a computational analysis of published literature to rank existing drugs according to predicted ability to reduce alpha synuclein (aSyn) oligomerization and analyzed real-world data to investigate the association between exposure to highly ranked drugs and PD. METHODS: Using IBM Watson for Drug Discoveryâ (WDD) we identified several antihypertensive drugs that may reduce aSyn oligomerization. Using IBM MarketScanâ Research Databases we constructed a cohort of individuals with incident hypertension. We conducted univariate and multivariate Cox proportional hazard analyses (HR) with exposure as a time-dependent covariate. Diuretics were used as the referent group. Age at hypertension diagnosis, sex, and several comorbidities were included in multivariate analyses. RESULTS: Multivariate results revealed inverse associations for time to PD diagnosis with exposure to the combination of the combination of angiotensin receptor II blockers (ARBs) and dihydropyridine calcium channel blockers (DHP-CCB) (HR = 0.55, p < 0.01) and angiotensin converting enzyme inhibitors (ACEi) and diuretics (HR = 0.60, p-value <0.01). Increased risk was observed with exposure to alpha-blockers alone (HR = 1.81, p < 0.001) and the combination of alpha-blockers and CCB (HR = 3.17, p < 0.05). CONCLUSIONS: We present evidence that a computational approach can efficiently identify leads for disease-modifying drugs. We have identified the combination of ARBs and DHP-CCBs as of particular interest in PD.