Building bridges with IT

Chenyang Lu is not a civil engineer.

For a computer scientist, however, he builds a lot of bridges. In particular between the fields of IT and healthcare.

Lu is a Fullgraf Professor in the Department of Computer Science and Engineering at the University of Washington at the McKelvey School of Engineering in St. Louis. His research focuses on the Internet of Things (IoT), cyber-physical systems and artificial intelligence, and he is particularly interested in how these technologies can improve healthcare.

As part of multiple teams of surgeons and doctors, Lu has tested Fitbit activity trackers in studies that have shown that these relatively inexpensive wearable devices can play a valuable role in improving health. patients.

“We can collect data such as step count, heart rate and sleep cycles, which we use with our machine learning models to predict the deterioration or improvement in a patient’s condition. . Lu said. “These efforts demonstrate tremendous potential for wearable learning and machine learning to improve healthcare.”

A graph from a pilot study published in 2020, in which researchers monitored ambulatory patients using wearable devices and developed machine learning models to predict readmissions of newly discharged congestive heart failure patients of Barnes-Jewish Hospital. The results demonstrated the feasibility of continuous monitoring of ambulatory patients using bracelets. Machine learning models based on multimodal data (steps, sleep and heart rate) have significantly outperformed the traditional clinical approach. A similar approach is used in a study to predict postoperative complications and readmissions of patients undergoing pancreatic surgery. (Photo: Chenyang Lu)

Using data from Fitbits, for example, Lu and coworkers demonstrated their ability to predict the surgical outcome of pancreatic cancer patients with greater success than the current risk assessment tool.

The goal is to improve healthcare, but where some of the most difficult problems arise, Lu finds engineered solutions. Obstacles can include sub-par data or just not enough data to get useful information from portable devices.

“You have to extract features using engineering techniques,” Lu said. “How to harness that noisy and ugly data from portable devices and extract robust, predictive features to generate something clinically meaningful and informative so that can we actually predict something? “

Obtaining useful information from messy data is one of the reasons his colleagues in the Faculty of Medicine appreciate his partnership.

“Chenyang has established himself as an expert in how to interpret and connect the dots of this large-dimensional data. That’s why I think he’s so prolific, ”said Philippe payne, Professor Janet and Bernard Becker and Director of the Institute for Informatics, Associate Dean for Health Information and Data Science and Chief Data Scientist in the Faculty of Medicine.

He may be prolific, but Lu is eager to do more.

“We can extend it to perfect the technology and broaden the reach so that it can be used on larger groups of different types of patients,” Lu said. “I look forward to collaborating with even more physicians and surgeons to extend this work. “

Comments are closed.