data.JUN <- (Flood Models to AI in Health, Andrew Goodwin)

Working with Large, Noisy Datasets - From Flood Models to AI In Healthcare

Andrew is a PhD student who’s work focuses on making sense of the high frequency physiological signals in a pediatric and neonatal intensive care environment. He has a background in environmental engineering, and began his career running noise and flood models for the mining industry in the hunter valley. These environmental models often require remotely sensed geospatial data including 3D point clouds derived from LiDAR, and high-resolution imagery from satellites and airborne sensors. Some of Andrew’s work in this area lead to the establishment of Newcastle based geospatial analytics company Anditi.

Andrew spent five years in Ontario, Canada from 2011 and worked at a large children’s hospital in downtown Toronto. While working as a data scientist and researcher at the hospital he designed and implemented a physiological waveform capture system that ingests over 120 million rows of data per hour. He works closely with doctors, nurses, biomedical specialists, computer scientists, and machine learning specialists to build predictive clinical models and deploy them on streaming data in a challenging environment. Over the past few years the waveform database has collected over 650,000 patient-hours of data and has grown to contain over 2.25 trillion rows.
Andrew has now returned to Newcastle and is working with the University of Sydney and the Neonatal ICU at Westmead Hospital to introduce some of the systems and techniques developed by the Canadian research group.

The presentation will cover some of the tools and techniques Andrew uses in his work and research. He will cover a wide range of challenges related to extracting insight from large, noisy datasets, and will give an overview of the current directions and limitations of AI in healthcare.

The talk will use examples drawn from Andrew’s two decades of data science experience to cover a broad range of topics including data cleaning, data compression, indexing, and visualization.

Online Profiles
https://scholar.google.com.au/citations?user=ayvh6V4AAAAJ&hl=en https://www.linkedin.com/in/andrewgoodwin/ https://twitter.com/ndrewg

Laussen Labs Website (Canadian Research Group)
https://laussenlabs.ca/

Anditi
http://anditi.com/

Links to some of online visualisations
https://www.youtube.com/watch?v=iDlCQ_JottU https://www.youtube.com/watch?v=PcuTmzSGx5A https://www.youtube.com/watch?v=b6vl2sMkAWs https://www.youtube.com/watch?v=Z76_Wbd2cj0