A Helping Hand
By Keith Loria
Radiology Today
Vol. 26 No. 5 P. 16
Business platforms help radiology practices manage.
Radiology practices today need to navigate complex challenges, including surging patient volumes, mounting administrative burdens, and increasing concerns over staff burnout. As imaging demands rise across health care settings, radiologists and administrators are under increasing pressure to maintain accuracy, efficiency, and patient-centered care.
In response, a range of digital platforms and workflow-enhancing technologies are emerging to help practices streamline operations, improve communication, and support clinical decision making, helping practices proactively manage these pressures.
For example, CARPL is an AI platform that allows radiologists to test, deploy, and monitor multiple AI applications through a single user interface, single data integration, and single procurement channel. “By allowing radiologists to use multiple AI solutions that can help across multiple use cases, CARPL helps them go through their reading lists faster, and with more confidence, since they know they have a pair of ‘AI’ eyes watching over them,” says Vidur Mahajan, MD, CEO of CARPL.ai.
Tennr is a health care automation platform that leverages AI and machine learning to streamline administrative workflows, particularly those involving the processing of medical documents. “There are so many problems facing the radiologists, but what we do is we really help the imaging centers, the places trying to get the patients to come in and be seen,” says Trey Holterman, cofounder and CEO. “We work with imaging centers that suffer from [scheduling] problems; if I send you 100 patients, you’re only going to get scheduled about 80% of those patients, and only about 80% of those are going to show up, so now you’re down to 64.”
What Tennr tries to do is deal with each of the steps where people drop off, thinking of the process as a funnel. “If you run a world-class operating system that takes in those patients, however they’re sent—faxes, email, e-portal—and you’re instantaneous at qualifying the patient to make sure you have all the right documentation to put together an ABN (advanced beneficiary disclosure) when it’s relevant, and you contact the patient immediately so you understand what their costs are going to be, you get a big tick-up in scheduled patients.” An ABN is a preservice disclosure that informs patients that Medicare may not pay for a service, why, and what it could cost.
In addition, when the correct reminders are going out and the correct documentation is delivered, patients tend to show up more, Holterman says. “We see people selecting the wrong payer—doing incorrect payer mapping,” he says. “Basically, when you read an insurance card as a provider, it’s super easy to accidentally enter the wrong plan for a given patient’s insurance into your billing system. Because all these payers have gobbled each other up, you have to read each plan closely, and you have to use reasoning logic to make sure that you have actually noted the right payer for each patient that comes through.”
Operational Efficiency
Holterman notes that it’s semicommon for people to make mistakes in either prescribing the wrong service for patients or entering the service into their RIS. “It may seem like something that wouldn’t happen often, but it happens a lot,” he says. “So, part of our ‘qualifications and authorizations’ model is making sure that the diagnosis codes the patient has been given correlate directly with the treatment they’ve been referred for.”
HealthLevel offers Foundations, a mobile platform for radiology business operations that assists organizations in achieving operational efficiency. “It is very clear that the lack of radiologists is hurting our business, and staffing is a huge problem,” says Parag Paranjpe, CEO and founder of Health- Level. “What that does is it pushes the existing staff to provide additional coverage, and that causes some pressure. They have to figure out how to optimize the result planning.”
Foundations helps practices, centers, and their staff streamline workflows with insights derived from clinical, financial, and operational data. “When someone is provided information like a personalized report or performance metrics, many times, administrators look at the data, and it’s not enough,” Paranjpe says. “Our platform has helped our customers by getting the insight and information in a very interactive, realtime way.”
Regulatory Compliance
A growing issue facing radiologists is regulatory compliance. Many of the new platforms are designed to assist radiology practices in staying compliant with ever-changing regulations while minimizing administrative burden.
“As radiology practices implement AI at scale, it becomes increasingly important for them to check the regulatory status of the AI that is being deployed,” Mahajan says. “It is difficult, expensive, and time consuming for radiology practices to do this with one AI application at a time. Instead, CARPL simply provides them a single window into the compliances of most of the world’s AI developers.” Along with FDA regulatory compliance, CARPL also handles security compliance on behalf of the AI developers—ISO 27001, SOC2 Type 2, etc.
Holterman says the most significant compliance risk is ensuring that providers pass the CMS’s Targeted Probe and Educate audit. “We’re seeing excited and renewed interest from the CMS in auditing. Not just Medicare Advantage plans, I think we’re going to see a big increase in auditing of providers, as well,” he says. “That means if you’re treating patients on Medicaid and Medicare plans, all of the documentation must be present. Even though you may not have to jump through all of the same authorization hoops with those plans, it’s extremely important that you cross your Ts and dot your Is; otherwise, it’s going to be a compliance nightmare.”
Dealing With Cyberthreats
Cybersecurity threats are a growing concern in health care, which is why these platforms incorporate protocols to safeguard sensitive patient data and ensure practice-wide security.
“CARPL is built from the ground up to be infrastructure agnostic, which means that CARPL, along with its third-party AI algorithms, can all be deployed within the firewalls of the health care provider, if that is what is needed by the customer,” Mahajan says. “This ensures that no data leaves the premises of the customer, providing maximum security.”
Additionally, CARPL over-indexes on compliance and was the first AI platform to get ISO 27001:2022 certification. It also has SOC2 Type 2 certification, FDA clearance, and ISO 13485, as well. “CARPL’s router, which acts as a connector between the health care provider and CARPL’s platform, has the ability to deidentify patient images, including the ability to remove protected health information that may be burned onto the image itself,” Mahajan says.
Paranjpe says HealthLevel’s platform has never had a breach, but potential security threats are horrifying. “The problem occurs when access is not properly managed and controlled because there is only so much encryption of patient data you can do before it loses functionality,” he says. “One of the most important things that should be done is something called ‘zero trust.’ Applications should not trust any access unless it is authenticated and authorized.”
Reducing Burnout
Workforce burnout remains a significant issue in radiology today. Automation within management platforms can help streamline workflows, reduce manual tasks, and support radiologists and staff in managing workloads.
“CARPL allows radiologists to interact with clinical AI applications through its universal AI widget, which sits in a corner on the radiologists’ workstation, acting as a ‘copilot,’ giving the radiologist additional information about scans they are viewing,” Mahajan says. “Additionally, the widget alerts radiologists to patients who need their attention. It allows them to see a quick snapshot of the AI outputs and then navigate directly to the critical case within their PACS. There are other functionalities like report autopopulation, normal batching, critical triaging, etc, that increase productivity of the radiologist.”
Tennr focuses more on the administrative side, helping staff locate documents and chart histories, eliminating a significant amount of unnecessary work. This allows people to focus on reviews and approving the work that’s being done.
“It’s a lot less work for a team, and it’s done in a way that’s high-quality and results in many fewer mistakes,” Holterman says.
Facilitating Interoperability
Interoperability between systems is critical for efficient practice management and should be nonnegotiable in any radiology practice, Paranjpe notes. “Why does this not happen, and why isn’t it being done?” He asks. “[Information is sometimes] lost in communication, and that’s something we work on with our customers. Interoperability and data exchange should be built into the radiologist service contracts.”
Mahajan says CARPL is built to integrate with existing IT systems using standards such as DICOM, HL7, and FHIR. “Most standards-based inputs can be configured into the system from the front end,” he says. “CARPL can even move AI outputs of different AI vendors from one format to another, allowing radiologists and IT teams to pick the best AI consumption mode that suits their needs across multiple AI systems.”
The most important third-party tool that Tennr recognizes is the fax line. “That’s how most patient records are still travelling today, via fax lines, email, and e-portals, and it’s always a mess of documentation because someone has to work with those records to make sure a patient gets from point A to point B,” Holterman says. “When we talk about interoperability, we don’t talk about fancy networks or protocols; we talk about how things are really sent today. We make it easy for centers dealing with those forms of data to be efficient and reduce their denials from these unstructured formats.”
Today and Tomorrow
As practices grow and data management becomes more complex, radiology administrators have numerous considerations to keep in mind when evaluating options to ensure scalability and data integrity. That’s why radiology administrators need a partner who has the tech savviness to be resilient to security threats, the scalability to cater to massive upticks in AI demand, and the flexibility to cater to changing needs of radiologists and IT teams, as more AI comes into clinical workflows.
“Flexible scalability is the term we use,” Mahajan says. “All said and done, AI is today in its infancy when it comes to large-scale adoption, which means that the methods and processes through which radiologists use AI will change rapidly over the next few years. This is all in addition to the regular architecture and security audits one has to do.”
Mahajan feels radiology administrators should pick an AI platform that can cater to the needs of not only today but also the future. “AI is ever evolving, so having a good sense of a platform’s vision for the future and the ‘why’ they are doing what they are doing is important,” Mahajan says. “Additionally, I believe that an all-encompassing approach to platforms, which allows radiologists to not only use AI but also build, test, deploy, and monitor these solutions, is a critical factor to consider. Lastly, deep integration with PACS is becoming a prerequisite for any AI deployment, so any AI platform that has a partnership with a PACS deployed at a radiology practice should be given higher weightage.”
Holterman warns there’s a great deal of “smoke and mirrors” when people talk about agentic AI and extremely high levels of overpromising about the ability of certain machine learning models to do this work. “There is this idea we can lump AI onto current processes, and it will automate work the way humans do. That’s a silly way to think about things,” he says. “It is a transformative force for a business, and you have to think about it more carefully on what it can really do.”
— Keith Loria is a freelance writer based in Oakton, Virginia. He is a frequent contributor to Radiology Today.