University of Utah Health provides care for residents of Utah and six surrounding states, with a referral area of about 10 percent to 15 percent of the continental United States.
Given this large area, and the increasing number of imaging studies being performed, the health system found it had far too much data.
“When I first started in radiology 25 years ago, I was looking at about 2,500 images a day, and today I look at around 250,000 images every day,” said Dr. Richard Wiggins, professor of radiology and imaging sciences at University of Utah Health. “With this amount of data increasing, we needed to figure out how to analyze this data in a more efficient way.”
So the health system decided to implement imaging vendor Philips’ Illumeo system, which includes adaptive intelligence that can interact with imaging data intelligently.
“The tool enhances physician expertise and efficiency to address these volume challenges,” Wiggins said. “By implementing this technology, we were hoping to deliver on goals such as improved turnaround times for more urgent cases, more efficient same-day care, streamlined screen real estate, more accurate suggestions of relevant imaging toolsets and improved accuracy in reporting.”
There is a wide variety of imaging vendors in the health IT market today. These vendors include Agfa Healthcare, Canon USA, Carestream Health, Fujifilm Medical Systems USA, GE Healthcare, Hyland Healthcare, Konica Minolta, Lifeimage, Novarad, Siemens and TeraRecon.
MEETING THE CHALLENGE
University of Utah Health had IntelliSpace PACS Radiology Workspace Solution in place, which provided worklist benefits to more intelligently route studies to the appropriate groups of radiologists.
“To build on this, we implemented Illumeo,” Wiggins explained. “Illumeo combines contextual cues and anatomy awareness capabilities with advanced data and image processing to help augment our routine activities. The technology supports interactions with imaging data comparisons and provides an integrated view of clinically relevant, case-related information from various sources, designed to optimize workflow and enhance care consistency.”
Wiggins has been using the new technology live in his clinical practice, and with Illumeo, he and his fellow radiologists have been able to move beyond hanging protocols, leveraging adaptive intelligence and analytics to pull in all similar series of studies and compare them side by side to see if lesions are increasing in size over time.
This has allowed the radiologists to further improve report accuracy, time savings and efficiencies so they can provide the best care to patients, he said.
“For example, an issue the implementation addressed was creating a more automated process for study promotion and prioritizing the most urgent studies,” he explained. “Before, most studies were being labeled as STAT. But this presents challenges when it comes to prioritization and was bogging down the system.
“When I first started in radiology 25 years ago, I was looking at about 2,500 images a day, and today I look at around 250,000 images every day.”
Dr. Richard Wiggins, University of Utah Health
“Now we can more intelligently label cases and set rules so those that need to be viewed in a timely manner can be, such as reports that need to be ready for same-day follow-up appointments,” said Wiggins.
Illumeo is integrated with the radiology information system and speech recognition systems at the University of Utah Health Sciences Center. The integrations allow for a bi-directional workflow in an academic setting such that a radiologist can choose a study that needs to be read from a filtered worklist, and Illumeo will then pass the information to the speech recognition system so that the radiologist will be ready to dictate that case.
“The in-training workflow is also available, so that I can choose a report in the speech recognition system, and when opening that report, the speech recognition system will then pass along information to Illumeo, so that the images are displayed and can be compared with the report generated by a student,” he said. “I can also move back and forth between these two workflows at any time.”
There are multiple administrative goals the health system is hoping to achieve with the adoption of the new imaging technology in conjunction with the PACS Radiology Workspace Solution. Some of these already are being delivered on, Wiggins said, including:
- Improved turnaround times – Radiologists can now prioritize the most urgent studies based on a more automated process for study promotion, instead of everything being labeled STAT.
- More efficient same-day care – By prioritizing cases, radiologists can successfully deliver reports in time for follow up appointments taking place later in the day. Since the health system covers such a large area, many patients travel from multiple hours away to benefit from same-day care.
- Streamlined screen real estate – Applications that are integrated in Illumeo give information in one window with similar user interfaces to minimize distractions for the radiologist.
- Relevant imaging toolsets – Illumeo is context-sensitive and anatomy-aware in the way the software intuitively suggests relevant tools for analysis.
- Exam open speed improvements – Illumeo displays images exactly the way radiologists need to read them, reducing setup time and interactions for a faster and more efficient reading.
ADVICE FOR OTHERS
“Adopting technology like this is very valuable,” Wiggins advised. “It can help radiologists anticipate what we need to do next, allowing us to interact with imaging data in a much more intelligent way. It is important that we begin to explore how we can use artificial intelligence and deep learning in our clinical practices.”
There are factors working against training of the best students, as currently medical students are being told not to go into radiology, as there will not be radiologists in five years, he added.
“However, most people looking at this research believe that AI and deep learning will be assisting us, not replacing us, for at least the next 50 years,” he explained. “Our jobs will be different, but there will still be a need for radiologists. It is true, however, that those who begin using these deep learning algorithms may replace radiologists who refuse to use these new technologies.”
Email the writer: [email protected]
Source: Read Full Article