Mock-up of the MRIdle software tool

MRIdle solves the problem of underutilization of MRI machines in the radiology department. This underutilization is caused by rudimentary appointment planning and the fact that a sizable chunk of patients don’t show up for their appointments. With the goal of improving the occupancy of the MRI machines, MRIdle will harness historical data to recommend optimal appointment schedules and to predict and mitigate no-show patients.

The general objective of the project is to improve the efficiency of the USZ Radiology department by improving patient attendance and reducing machine idle time. Over the next two years, we aim to achieve this goal by building the MRIdle smart scheduling tool that fulfills the following specific objectives:

  • Power targeted appointment reminders to reduce the frequency with which patients miss their MRI appointments. The scheduling tool will identify patients that are at risk for not showing to their appointments. With these predictions, the scheduling staff may proactively reach out to these patients to increase the likelihood of them attending their appointments.

  • Improve schedule density with a smart scheduling application that assists staff in scheduling appointments. The smart scheduling tool’s AI-powered algorithms will recommend optimal durations and timings of appointments to guide the scheduling staff in creating an optimal schedule with minimal idle time.

  • Make the Radiology department more robust to changes in staff. In our interviews with the radiology department’s scheduling staff, we learned that scheduling the department’s many kinds of appointments requires experience and expertise. Today, that knowledge is stored only in the minds of the department’s long-time staff. MRIdle’s machine learning models learn from years of appointment scheduling data to make new staff just as powerful schedulers as long-time ones.

Laura Kinkead
Senior Research Software Engineer

Building software tools for Data Science.