“AI success is so much more than a working product.”
Arterys CEO, John Axerio-Cilies, PhD, on the role that infrastructure, clinical workflow, and behavioral change play in achieving AI success.
Imaging Wire Sponsors
- Arterys – Reinventing imaging so you can practice better and faster.
- Bayer Radiology – Providing a portfolio of radiology products, solutions, and services that enable radiologists to get the clear answers they need.
- GE Healthcare – Enabling clinicians to make faster, more informed decisions through intelligent devices, data analytics, applications and services.
- Healthcare Administrative Partners – Empowering radiology groups through expert revenue cycle management, clinical analytics, practice support, and specialized coding.
- Hitachi Healthcare Americas – Delivering best in class medical imaging technologies and value-based reporting.
- Novarad – Transformational imaging technologies that empower hospitals and clinicians to deliver clinical, operational and fiscal excellence.
- Nuance – AI and cloud-powered technology solutions to help radiologists stay focused, move quickly, and work smarter.
- Riverain Technologies – Offering artificial intelligence tools dedicated to the early, efficient detection of lung disease.
- Siemens Healthineers – Shaping the digital transformation of imaging to improve patient care.
- United Imaging – Our mission, Equal Healthcare for All, pushes us beyond conventional boundaries to help clinicians expand modern, digital, intelligent care to more people within their communities.
- Zebra Medical Vision – Transforming patient care with the power of AI.
The Imaging Wire
Who Will Pay for AI?
We’ve been talking/writing about imaging AI economics quite a bit lately, but just about all of these perspectives came from the folks on the making/selling side of the AI industry. That’s what makes Radiology: AI’s new buyer/user side perspective worth paying attention to. Here are some highlights:
- FDA Approval, the Short Hurdle – If you measure progress by FDA approvals, imaging AI is doing just fine, representing 72% of all FDA-approved healthcare AI devices so far.
- Narrow AI Reimbursements – However, there are only a handful of imaging AI reimbursement codes through MFPS (IDx-RX for diabetic retinopathy) and IPPS (NTAP payments for large-vessel occlusion strokes & FFR-CTs), and they don’t totally match how these products are used.
- Fee-For-Service Challenges – With that, the authors warned of a “challenging” path towards AI adoption with the current fee-for-service payment models.
- Value-Based Path – As value-based payment models mature, AI that maintains/improves quality without increasing costs should be well positioned. This could include AI tools that help radiologists achieve MIPS quality incentives (so radiologists adopt them) or population health tools that catch/predict more early-stage diseases (so health systems and payers adopt them).
- Who Will Pay for AI? – The authors believe that the “entity that receives the most benefit likely will pay for AI, and ultimately may consider this payment simply the cost of doing business.” That might not be as specific as some might hope, but it still makes sense.
Imaging Wire Q&A: Arterys’ AI Journey
In the latest Imaging Wire Q&A we sat down with Arterys’ CEO and co-founder, John Axerio-Cilies, PhD, to discuss medical imaging’s AI and cloud evolution and how Arterys works with its Center of Excellence partners to make AI real. Here are some of the big takeaways:
- AI Milestones – Back in 2011 the forward-looking part of the healthcare industry was mainly focused on precision health and big data, but cloud has since become a requirement for many institutions and AI continues to make its way beyond the early adopters.
- Arterys’ Platform Evolution – Over the last few years, Arterys expanded its platform to more use cases, modalities, and service lines, while making the platform almost self-serve for AI vendors and institutions.
- Evidence & Excellence – Targeting the imaging industry’s historic lack of clinical evidence, Arterys created its Center of Excellence program to help its customers achieve their AI adoption goals – and give them the data to guide this process.
- AI Adoption Advice – Institutions considering AI adoption should have a defined problem that hurts bad enough to solve, understand their goals, and commit to doing the work required to make AI real.
- Philips’ Disney Experience: Philips announced a collaboration with Disney to measure how broadcasting custom-made animated Disney stories within Philips’ Ambient Experience solution (integrates video, sound, lighting w/ MRI) helps reduce children’s fear/anxiety during MRI exams. The project will operate across six European hospitals through the second half of 2021, and feature six co-developed videos that were created with Disney’s characters/storytelling and Philips’ clinical guidance.
- Get Prepared for Cardiac CT & MRI: A new Radiology Journal editorial warned diagnostic radiologists at all institutions (not just academic) to get prepared to interpret cardiac CT and MRI studies, noting the exams’ low current radiologist adoption and recent/expected growth. The editorial encouraged residents to achieve level 1 training and suggested that current rads should seek cardiac CT/MRI education.
- SWE for Brain Tumor Surgery: A new Frontiers in Oncology study found that shear wave elastography ultrasound (SWE) could be a valuable tool to help neurosurgeons detect and remove brain tumors during surgery. The researchers performed intraoperative SWE scans on 26 patients before/during/after brain tumor removal surgery, detecting remaining tumor tissue with 94% sensitivity (vs. 36% surgeon estimates, 73% B-mode ultrasound) and 77% specificity (vs. 100% surgeon estimates, 63% B-mode ultrasound).
- HAP’s Surprise Billing Guide: With the U.S.A.’s long-awaited surprise billing legislation now official, Healthcare Administrative Partners just provided a helpful breakdown on how it will affect radiology practices when it goes live next year. Here are highlights: 1) There won’t be any changes in the 16 states that already have surprise billing rules; 2) Unlike some specialists, radiologists can’t use a consent waiver to bypass surprise billing rules; 3) Radiologists will have to be even more diligent obtaining coverage information and reviewing payments; 4) Practices should define their minimum acceptable payments so they can quickly identify which payments to resolve/dispute.
- LCS CAC Screening: A new Duke University study revealed that lung cancer screening reports could help identify patients at risk of atherosclerotic cardiovascular disease (ASCVD) and guide them to statin therapy. The team reviewed 5,495 individuals’ LCS exams, finding that 73.6% of the individuals (2,777/ 3,771) who were not already diagnosed with ASCVD were “statin eligible” based on their LCS CAC scores and other factors. Although 60.5% of these “statin eligible” individuals (1,681 / 2,777) were not receiving statin therapy at the time of their screening, they were twice as likely to begin treatment if their radiology reports noted high CAC scores.
- Philips & DiA Partner: Philips and DiA Imaging Analysis announced a partnership that combines Philips’ PoC ultrasound systems with DiA’s ultrasound automation solutions, while giving Philips a minority stake in the ultrasound AI company. DiA’s apps focus on automating cardiac and abdominal ultrasound capture/analysis, streamlining exams and expanding ultrasound to less-experienced clinicians. DiA also has ultrasound partnerships with GE Healthcare, Konica Minolta, and Terason, although Philips appears to be the first OEM to take a stake in the company.
- ADC Cuts Unnecessary Breast Biopsies by 32%: One third of unnecessary breast biopsies could be eliminated if clinicians used DWI MRI / CE-MRI apparent diffusion coefficient (ADC) measurements to define BI-RADS 4 lesions. A retrospective study out of Vienna (n = 657 patients, 696 BI-RADS 4 lesions on CE-MRI, 414 cancerous lesions) applied a DWI ADC cutoff for BI-RADS 4 lesions (≥1.5 × 10−3 mm2/second) that would have eliminated 32.6% of the unnecessary biopsies (92/282) while approving biopsies for 96.6% of cancerous lesions (400/414). Since DWI is widely available, this measurement technique could “immediately” be adopted “anywhere.”
- Enterprise Imaging Labeling Standards: A new HIMSS-SIIM white paper called for enhanced image labeling standardization in order for “enterprise imaging to be successful across specialties, systems, and sites.” The paper specifically proposed a standard based on the DICOM Anatomic Region Sequence codes, suggesting that this standardization would improve interoperability, patient care, patient privacy/safety, analytics, and AI development.
- Benign Bone Tumors, Common in Kids: Benign bone tumors are a common incidental finding in pediatric X-rays, and that’s because nearly 19% of asymptomatic children/adolescents have them. That’s from an Indiana University longitudinal study that reviewed 25.5k left-side extremity X-rays from 262 pediatric subjects, finding 35 tumors in 33 subjects, and estimating that 18.9% of the subjects would have benign bone tumors if other body areas were also scanned.
- The Free-Text Epidemic: Even though physicians use clinical decision support systems that prompt them to select a structured indication for advanced imaging orders, many orders aren’t scored for imaging appropriateness because many physicians chose to enter free-text indications. That’s from a new JACR study that reviewed 39.5k advanced imaging orders within a large health system, finding that 59% of the orders (23,267 of 39,533) did not receive AUC scores, and 71% of those unscored orders had free-text indications (16,440 of 23,267).
- Siemens Adds Epsilon Echoinsight: Siemens Healthineers will offer Epsilon Imaging’s Echoinsight strain imaging solution as part of its syngo Dynamics cardiovascular imaging and reporting solution. The distribution agreement will allow syngo Dynamics users to add strain imaging with automated cardiac function measurements to their echo workflows (and receive reimbursements for myocardial strain imaging).
- Chest CT COVID Outbreak Predictor: A Mednax team developed a machine learning NLP algorithm that could help spot regional COVID outbreaks based on the language in chest CT reports. Using chest CT reports from Mednax’s imaging platform (450k CT reports, 2,100 facilities, 50 states) the team developed three NPL algorithms to track CT features typical of COVID infections and compared their findings against official infection figures. The best performing algorithm showed a strong correlation with official daily COVID-19 cases, including on the state level (r2=0.82, p<0.005).
- Infection & Inflammation PET Coverage: CMS rescinded its 12-year-old national noncoverage policy for PET imaging for infection and inflammation, allowing Medicare Administrative Contractors (MACs) to decide whether the scans will be covered/reimbursed in their regions. The major radiology societies are now meeting with the local MACs to encourage infection/inflammation PET coverage, while SNMMI is working on appropriate use criteria for these scans (starting with fever of unknown origin).
The Resource Wire
– This is sponsored content.
- Smaller radiology practices (<20 employees) will have prioritized access to file for Paycheck Protection Program (PPP) loans through March 10th. Check out Healthcare Administrative Partners’ latest breakdown of this year’s PPP and how it will work for radiology practices.
- In this Novarad video, interventional oncologist Gary M. Onik, MD shares how Novarad’s AR surgical navigation system, OpenSight, helps his team accurately access and treat tumors.
- Learn how leveraging the right cardiology image and reporting platform drives performance and outcomes in this Hitachi white paper.
- Read Siemens Healthineers’ latest case study to see how one customer is pioneering MR-guided, non-invasive treatment for prostate cancer with MAGNETOM Skyra MRI.
- Check out how Arterys’ Lung AI automates chest CT workflows, cutting reading times by 45% and reducing missed nodules by up to 70%.
- Catch up with Zebra-Med at this week’s ECR meeting where they’re demoing a range of population health and diagnostic solutions.
- Learn how Yale New Haven Health improved its radiology efficiency, communications, and turnaround times when it adopted Nuance’s PowerScribe Workflow Orchestration and PowerConnect Communicator solutions.
- Catch Riverain’s discussion about using AI to detect lung disease at the Society of Thoracic Radiology conference this Sunday (March 7th) at 1PM Eastern.
- Check out this GE Healthcare report detailing how even after cutting MRI scan times by up to 50% with deep-learning MR image reconstruction, radiologists find it “just easier to read through these beautiful images.”
- This Bayer case study details how radiation benchmarking programs can help push CT dose exposure reduction initiatives from achieving compliance to driving quality.
- If you missed United Imaging a couple of months ago at RSNA, you missed their well-attended virtual presentation, Beyond the Status Quo. In part one of that video, they cover their product launches. (Look for part two in the next Imaging Wire.)