Delving into the data at Hack Aotearoa 2020
This Friday marks the return of Hack Aotearoa, the globally recognised Massachusetts Institute of Technology (MIT) Critical Data event.
Building on the success of the 2019 event, Hack Aotearoa 2020 continues to innovate and will include a health datathon in addition to the main Artificial Intelligence in Healthcare conference. During the datathon, local and international data scientists, academics and healthcare professionals will have the chance to work with the MIT eICU Collaborative Care and MIMIC Critical Care datasets.
The Multiparameter Intelligent Monitoring in Intensive Care (MIMIC) database was the first resource of this kind. Developed and maintained by the MIT Laboratory for Computational Physiology (LCP), it is now in its third iteration as the Medical Information Mart for Intensive Care (MIMIC-III) database. The MIMIC-III database contains high-resolution and multi-modal de-identified data from the electronic health records (EHRs) of over 50,000 hospital admissions. MIMIC-III includes vital sign recordings and waveforms, laboratory data, clinical notes, diagnostic reports, and administered interventions, including medications.
In addition to MIMIC-III, MIT Critical Data partnered with Philips to release de-identified data from the Philips eICU Research Institute. The eICU Collaborative Research Database (eICU-CRD) is a multi-centre critical care database comprised of more than 200,000 intensive care unit admissions from across the United States.
Conference organiser Dr Mataroria Lyndon, senior lecturer in the School of Medicine at the University of Auckland, says Hack Aotearoa delegates should be excited about the two datasets available.
“Often there’s limited access to large healthcare datasets, so it’s a fantastic opportunity for clinicians and data scientists to be able to work with the MIMIC and eICU data – large-scale datasets that are well curated, high resolution and have many variables. This sort of data allows much more sophisticated analyses to be performed, compared with smaller or less well-curated datasets,” says Dr Lyndon.Dr Mataroria Lyndon, Hack Aotearoa 2020 organiser
The Hack Aotearoa 2020 datathon is an ideal opportunity to introduce new researchers to the MIMIC and eICU data. It also provides a platform for continuous development of code and research, as the data use agreement requires that MIMIC code be shared.
Clinicians and data scientists will be grouped into teams and will work together to solve a clinical problem or challenge. Dr Lyndon says that datathon participants will learn about large datasets, data mining, programming language, and applying data science methods to answering clinical problems.
“They’ll ‘learn through doing’, by working together in interdisciplinary teams to analyse health datasets to answer a clinical problem. They’ll learn from each other and learn together,” he says.
The benefits are two-fold: healthcare professionals can learn about data extraction and model development, while data scientists gain invaluable insights into clinical data capture and decision-making. In short, the event is collaborative in scope with the ultimate aim to improve the diagnosis and treatment of critical care patients.
This weekend’s datathon is focused on six clinical challenges:
- Comparing the mortality of patients admitted to the intensive care unit on a weekend with those admitted on a week day
- Predicting in-hospital mortality of intensive care patients using decision trees
- Comparing methods for identifying patients with sepsis
- Evaluating the reproducibility of mortality prediction studies that use the MIMIC-III database
- Optimising treatment strategies for sepsis using Artificial Intelligence (AI)
- Identifying the association of hypokalemia (low blood potassium) with an increased risk for medically treated arrhythmia (irregular heartbeat)
Precision Driven Health data scientist Dr Ning Hua is looking forward to attending the PDH-sponsored event.
“The datathon will be a good chance to work really closely with health professionals and get tips on the sorts of problems they’re interested in solving. I’m really looking forward to two days of intensive cooperation!” says Ning.
Hack Aotearoa 2020 will take place 17th to 19th January at Owen G Glenn Building, University of Auckland. Find out more at www.hackaotearoa.co.nz.
Learn more about the datasets: