CoVar Technologies Solutions

Clarity™

Clarity™ gives you the power to explore complex data sets and advanced machine learning algorithms.

The Clarity™ data analysis toolkit is a highly flexible dashboard software tool that enables data exploration, visualization and insight generation via rapid filtering of complex, multidimensional datasets, coupled with responsive and interactive visualizations. It features a dynamic, responsive and highly customizable dashboard with interactive data visualizations that enable rapid slicing of datasets across multiple dimensions. Clarity also enables drilling down into the raw data to display assets associated with individual records in the database. These assets could include images, text, videos, radar data, medical records, etc. The ability to drill down into the data at a more granular level and view associated assets can enable much deeper insights into the data and can help guide feature engineering and machine learning algorithm evaluation.

This powerful and interactive toolkit enables a better understanding of outliers and edge cases, confounding variables, and the ability to easily assess overall machine learning algorithm performance. Simple, yet powerful, it gives you both a quick and deep visualization of your data sets. Clarity leverages the latest web technologies (HTML5, CSS, React, Node.js, D3.js) to enable rapid software development. The flexibility of Clarity enables easy implementation of custom features on an application specific basis.

Clarity™ was the foundation of the work CoVar performed for North Carolina Translational and Clinical Science Institute (NC TraCS) (integrated hub of the NIH Clinical and Translational Science Award at the University of North Carolina at Chapel Hill) to build CLARK (CLinical Annotation Research Kit), a medical record classification/prediction software package. After training the classifier on a gold standard of patients and their clinical notes, CLARK can be used to infer information about novel sets of patients, including possible diagnoses, potential outcomes, or presence or absence of nuanced symptoms. CLARK is currently being piloted in a number of studies at UNC Chapel Hill. Click here for more information about NC TraCS and its mission.

UNC