Simplifying the workflow: Artificial intelligence in MRI

When radiology departments talk about their workflow, they are typically referring to the patients they see and how long it takes to scan each of them, including pre-scan or set-up time, imaging the patient and post-processing time. The ideal workflow for magnetic resonance imaging (MRI) would prioritize patients, as well as optimizing the scan time for each patient. This would allow the MR technologist to focus on what they want to focus on throughout the scan with the help of an intelligent or automated workflow. These intelligent workflows may increase the comfort and confidence of the technologist as well as the patient. This is especially true when the workflow is combined with other techniques and innovations to increase comfort and accuracy.

Prioritizing MRI patients

Unfortunately, the current workflow for many radiology and MRI departments may not be as efficient as they wish it were. This is because some departments want to scan more patients than is currently possible. In cases where the patients are labeled "stat", which can mean the scan is urgent, the department must see them and provide the results as quickly as possible. They also want to make sure they are accommodating and imaging as many of these patients as they can.

As it is, patients may be "stat" patients, meaning they should be scanned as soon as possible, in two different situations.1 One of these situations is in reference to an emergency or urgent care patient who should be scanned as soon as possible, such as someone who has had a stroke. The other may refer to someone who has an appointment the same day and whose results are needed quickly.

With an improved, intelligent workflow that uses artificial intelligence (AI), the system may be able to use phrase-detection to determine which of these two situations the patient is in.1 This could help determine who should be scanned first or more quickly than the current method, as well as adjusting the schedule to allow the radiology department to see as many patients as possible. This is still being researched, but the potential for this type of software is astounding.

Optimizing MR scan time

Alongside the need to scan more patients in the course of a day, radiology departments may find that they need to have a more optimized workflow throughout the appointment. Some appointments can take longer than others, because the patient needs to be positioned multiple times or because the patient moves during the scan. Radiology departments may try to accommodate these potential obstacles in their scheduling. However, the scan time during the session may be increased as the technologist goes through the steps of positioning the patient, scanning and then re-positioning.

The radiology department at Kawasaki Saiwai Hospital in Kawasaki, Kanagawa, Japan, focuses on patient-centered healthcare.2 The Chief Technologist, Takafumi Naka, who is a registered radiology technologists for MR, personally tested their new software and coils that were purchased in February. Both the software and the coils have shown promise in increasing patient satisfaction and comfort, as well as saving time throughout their workflow.

The new software includes a program that reduces the chance for human error by automatically detecting and centering the new coil.2 Naka has had to center the patient manually in the past. He says that with this new software, the department has a better patient positioning workflow than they have had in the past, and it saves the technologist time in the process.

The new anterior array coil for the MR system has increased versatility. It is able to scan a wider variety of patients because of its flexibility and durability.2 The design of the new coils allow technologists to position the coil closer to the patient's body and, at times, can be wrapped around the body part being imaged.

Many radiologists may see a benefit to their patient-centered healthcare centers with these new intelligent workflows that can help to aid departments in their prioritization and scanning. Patients who are scanned may also be more at ease, because of the confidence the radiologist may have and the increased comfort of the new type of coils employed at Kawasaki Saiwai hospital, like Takafumi Naka has already experienced.



  1. Improving Efficiency With an Intelligent Worklist and Rules Engine. itnonline.com Last accessed July 10, 2019.
  2. Ultra-Flexible AIR Technology Suite making a difference in the technologist's workflow. SIGNA Pulse of MR Last accessed July 10, 2019.
  3. Deep learning improves radiologist workflow, efficiency determining musculoskeletal MRI protocol. HealthImaging.com Last accessed July 10, 2019