Electronic Health Records

Vast amounts of heterogeneous and complex data from Electronic Health Records (EHRs) are ubiquitously being recorded at the patient level in healthcare (“big data”). This represents a largely untapped source of data-driven clinical information having the potential to transform health by developing autonomous monitoring systems as well as diagnosis and decision support tools. This will leap forward the quality of care for the individual patient, and lead to reduced costs in healthcare.

We will build on research developed over several years of experience in the UiT Machine Learning Group and our team in EHR health analytics and image analysis, and will focus on concrete interrelated prediction and prevention core ICT technologies in health related to gastrointestinal colon cancer surgery, taking advantage of the uniquely available data material.

A large fraction of the EHRs consists of clinical multivariate time series. Building models for extracting information from these is important for improving the understanding of diseases, patient care and treatment. Such time series are oftentimes particularly challenging since they are characterized by multiple, possibly dependent variables, length variability and irregular samples.

The UiT Machine Learning Group has published multiple works aimed at addressing such issue, with kernel methods and deep learning methodology forming the backbone of our approaches.

Highlighted Publications

1. Noisy multi-label semi-supervised dimensionality reduction (Pattern Recognition 2019)

Karl Øyvind Mikalsen, Cristina Soguero-Ruiz, Filippo Maria Bianchi and Robert Jenssen

2. Time series cluster kernel for learning similarities between multivariate time series with missing data (Pattern Recognition Volume 2018)

Karl Øyvind Mikalsen, Filippo Maria Bianchi, Cristina Soguero-Ruiz and Robert Jenssen