When Time Really Matters: One Physician's Journey to Leveraging Data Science for Better Outcomes

Dataiku Product, Scaling AI Josh Hewitt

While medicine and treatment options for patients are constantly evolving and improving, the healthcare field as a whole grapples with seamlessly incorporating data science, machine learning, and AI into their workflows. By making AI accessible to a wider audience and equipping medical professionals with the technology and processes that ensure AI efforts are collaborative, transparent, and explainable, there is significant potential to decrease costs and improve patient outcomes.

Dr. Lawrence Richer, Associate Dean of Clinical/Translational Research in the Faculty of Medicine and Dentistry at the University of Alberta in Canada sat down with members of the Dataiku academics team to explain his own journey with Dataiku and data science to help expand his team's ability to empower physicians to make better and more informed diagnoses. Specifically, they wanted to see if they could create an algorithm that could help raise awareness of stroke in children presenting to the emergency room.

Specializing in pediatric neurology, Dr. Richer and the team at the Women and Children’s Health Research Institute strive to help physicians overcome some of the clinical challenges they face in diagnosing and treating rare conditions such as strokes in children. Because strokes are so rare among children, many physicians do not think of them as a possibility for their patients. While adults who experience a stroke are typically diagnosed and treated within a matter of hours, the time between when a child experiences a stroke to diagnosis is much longer and, unfortunately, sometimes days after the window during which most treatment options can be applied.

Dr. Richer came across Dataiku and was attracted to its low barrier to entry. Shortly after, with a Dataiku academic license in hand, Dr. Richer and his team — a diverse array of profiles including only one data scientist — immediately began to see and appreciate the real application that data science, machine learning, and AI could have on healthcare.

Dr. Richer, who has a Master’s degree in clinical epidemiology, found the platform to be quite intuitive and leveraged Dataiku’s tutorials and sample projects to inform his training and the project as a whole. Using a relatively small sample dataset, the team was able to achieve its objective. After applying machine learning, they created an algorithm that, if used today, would cut the number of CT scans in half and help lead to better outcomes for pediatric patients.

But, that was just a small data sample. Dr. Richer knew that there was room for improvement. Through data augmentation, the team went from 2,500 records to a million to allow for more meaningful and accurate predictions.

Looking Ahead

Of late, Dr. Richer is no longer just any physician. With his newly acquired data science skills, he is ready to leverage Dataiku’s capabilities even further. The next step for Dr. Richer and his team is to push the model they created out of the lab and move on to the real-life application, educating their network of doctors to effectively reduce the time from presenting symptoms to diagnosis.

They are also excited to expand their use cases and identify other ways that Dataiku can help them drive tangible impact in the healthcare field. Dataiku’s collaborative platform will continue to enable clinicians — including Dr. Richer and his team — to collect, clean, and wrangle data and transform it into actionable insights that will guide physicians in their decision-making processes across the greater physician community.

Dr. Lawrence Richer is the Associate Dean for Clinical & Translational Research, Director of the Northern Alberta Clinical Trials and Research Center, Professor and Division Director of Pediatric Neurology, and Associate Director of the Women and Children’s Health Research Institute at the University of Alberta.

You May Also Like

A Sneak Peek of Season 5 of the Banana Data Podcast: Humanizing Data Science and AI

Read More

A Day in the Life of a Data Scientist at Pfizer

Read More

Understanding the Power of Data

Read More

How to Perform Basic ML Scoring With Scikit-Learn, Docker, Kubernetes

Read More