Excitement brews with Watson’s capabilities, but companies already provide artificial intelligence to radiology.
Imagine this: your hospital administrator asks you to help reduce the length of inpatient stays and they need a plan within a week. Could you do it?
Chances are, most of you couldn’t. But, the technology to mine and analyze your data does exist. Much like your daily Google searches, it’s possible to input your search criteria, click Enter, and have answers at your fingertips in seconds.
Doing so is part of radiology’s push toward using big data, said Woojin Kim, MD, director of innovation at Montage Healthcare Solutions, Inc.
“Radiology doesn’t yet have big data like other industries, but that’s changing rapidly. People want access to data to be able to turn insight into action,” he said. “Providers want quick easy access to radiology reports, and they want to mine through data intelligently for research and quality and performance improvements.”
That’s why Kim, along with William Boonn, MD, Curtis Langlotz, MD, and Rajan Agarwal, MD, co-founded Montage. Its products use proprietary natural language processing to search and pull information from your RIS and electronic medical record (EMR), pinpointing clinical findings to augment business performance and clinical-quality analytics.
How It Started
The idea behind Montage was borne out of a moment of frustration in 2009, when Kim realized he could use an Internet search engine to answer any question in seconds, but he couldn’t search his own patient data the same way.
“I had a personal need for that patient information, and there was no easy way to find it if I didn’t remember the patient’s name,” he said. “I thought not having that ability was unacceptable.”
Although health care, overall, uses many forms of advanced technology, it’s decades behind other industries in how it uses analytics. But, the delay will be a benefit as health care can learn from and side-step mistakes that other groups have made.
From inception, Montage’s initial impetus for catching up was getting quick and easy access to information morphed into a focus on developing intuitive ways for radiologists to mine and analyze their own data. The biggest emphasis has been on creating clearer, more informative radiology reports that include better quality metrics and RVU details without requiring users to be analytics experts.
Fundamentally, Kim said, Montage wants its technologies and tools to be as user friendly as possible.
“When I do demonstrations for Montage, I tell people if they can make their own airline reservations online, they can use our analytics solution,” he said.
Who Are the Customers and What Do They Want?
Currently, Kim said, Montage serves 150 customers at approximately 500 sites in the United States and Canada. Each day, thousands of clients log on, running more than 200 million radiology reports so far.
The company’s customer base is largely filled with hospitals, groups, or practices looking to abide by Meaningful Use reporting requirements. They’re interested in increasing reimbursements, improving their compliance rates, decreasing errors, and strengthening their abilities to mine their own outcomes data.
To reach that goal, he said, clients have specific needs they’re looking to fulfill:
• Better ways to track and monitor critical results and notification documentation
• Methods to detect errors in radiology reports, such as laterality or gender errors
• Strategies for monitoring PQRS compliance reports
• Tactics for improving how facilities follow up on radiologists’ recommendations
Growing Popularity[[{"type":"media","view_mode":"media_crop","fid":"47692","attributes":{"alt":"big data radiology","class":"media-image media-image-right","id":"media_crop_2506148633954","media_crop_h":"0","media_crop_image_style":"-1","media_crop_instance":"5643","media_crop_rotate":"0","media_crop_scale_h":"0","media_crop_scale_w":"0","media_crop_w":"0","media_crop_x":"0","media_crop_y":"0","style":"height: 200px; width: 200px; border-width: 0px; border-style: solid; margin: 1px; float: right;","title":"©phipatbig/Shutterstock.com","typeof":"foaf:Image"}}]]
While customers spread opinions about Montage word-of-mouth, much of its popularity has come from its close alignment with the American College of Radiology’s Imaging 3.0 initiative. The company’s products and mission support the push toward value-based care, Kim said, giving customers a leg-up on meeting fulfilling industry and patient expectations.
But, the tools don’t forego the need to continue reading studies and producing reports at high levels.
“Despite the emphasis on value, we can’t ignore volume, and as a practicing radiologist myself, the challenge is greater than before because we have the additional pressure of high volume while providing high quality,” Kim said. “We have to demonstrate the quality of our work.”
What Montage Offers
In an effort to meet those requests, Kim said, Montage offers workflow upgrades in four areas.
1. Enterprise Search: This search function facilitates exam dose and ordering pattern analysis, as well as clinical decision support and clinical quality improvement. It can also improve business performance through analysis of clinical report instructional information.
2. Radiology-Pathology Correlation: For the first time, diagnostic imaging and pathology reports can be linked in ways that improve quality and encourage clinical follow up. Doing so helps providers examine pathological disease progress and identify any reporting discrepancies.
3. Quality Analytics: This clinical quality monitoring can identify and resolve any undocumented critical results that affect JCAHO communication and compliance, opening the door to correcting errors that affect reimbursement and quality reporting.
4. Business Analytics: Through interactive dashboards, graphs, and tables, this tool helps practices identify their operational gaps, and pinpoint departmental and staff productivity issues that impact customer service and patient care. According to Kim, it sheds light on how modality, exam type, ordering physician, reading physician, and patient type can impact report volume, RVUs, and turn-around-time.
Montage’s Impact on Clients
So far, several clients have worked with Montage to make significant changes to their productivity, quality, and workflow.
For example, Radiology Associates of Canton worked with Montage to reduce their length of stay by three days at Aultman Hospital in Canton, Ohio, saving approximately $360,000. Administrators at Elkhart General Hospital in Elkhart, IN, completed 14 quality improvement projects in 90 days.
And, in coming weeks, customers will be able to learn directly from one another when Montage releases a new tool. This product will allow clients to share step-by-step details about what worked and what didn’t with any Montage-related initiatives.
Ultimately, Kim said, practices, providers, and facilities interested in solidifying their place in the industry could benefit from what Montage has to offer.
“If you want to succeed in a world of declining reimbursements, increasing regulations, and increasing competition,” he said. “If you want to ask questions on your own, mine your data to evaluate outcomes and quality, enhance research, reduce length of stay, improve compliance, increase revenue, decrease errors and medicolegal risks, and optimize productivity and efficiency – take a look at what Montage has to offer.”
New Study Examines Short-Term Consistency of Large Language Models in Radiology
November 22nd 2024While GPT-4 demonstrated higher overall accuracy than other large language models in answering ACR Diagnostic in Training Exam multiple-choice questions, researchers noted an eight percent decrease in GPT-4’s accuracy rate from the first month to the third month of the study.
FDA Clears AI-Powered Ultrasound Software for Cardiac Amyloidosis Detection
November 20th 2024The AI-enabled EchoGo® Amyloidosis software for echocardiography has reportedly demonstrated an 84.5 percent sensitivity rate for diagnosing cardiac amyloidosis in heart failure patients 65 years of age and older.
FDA Clears Updated AI Platform for Digital Breast Tomosynthesis
November 12th 2024Employing advanced deep learning convolutional neural networks, ProFound Detection Version 4.0 reportedly offers a 50 percent improvement in detecting cancer in dense breasts in comparison to the previous version of the software.