“the people most impacted—those whose lives are changed by the outcome of an algorithm—have no power”
Boston Review’s Rachel Thomas on how the growing role of healthcare AI is disempowering patients and providers.
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.
- 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.
- Zebra Medical Vision – Transforming patient care with the power of AI.
The Imaging Wire
CT/US Fusion Biopsies
A University of Cambridge-based research team developed a precision tissue sampling technique that combines CT radiomics and ultrasound guidance to improve biopsies, suggesting that this technique could eventually allow “virtual biopsies.”
- The Technique – The new technique superimposes CT radiomic tumor habitat images with real-time ultrasound images before performing targeted biopsies. Physicians then perform ultrasound-guided biopsies, relying on the pre-biopsy CT/US fusion images to guide tissue sampling.
- The Study – In a prospective study on six patients with suspected high-grade serous ovarian cancer (HGSOC), an experienced gynecological radiologist used the fused images to obtain up to six biopsies per patient.
- The Results – The clinical team was able to successfully co-register the CT-based radiomic tumor habitats with US images in all six patients. The new technique achieved a 0.53 median DSC accuracy score (range 0.79 to 0.37), posting the highest accuracy with larger pelvic tumors (DSC: 0.76–0.79) and lower accuracy with smaller omental metastases (DSC: 0.37–0.53). Two of the patients’ biopsies yielded insufficient materials for diagnosis.
- Takeaways – Although more research is needed, this is a promising first step for a technique that could improve biopsy accuracy and efficiency, and can be “can be seamlessly integrated” into clinical workflows.
- DIY X-Ray: Researchers from Finland’s University of Oulu are developing an automated compact X-ray system that could change where and how X-ray imaging is performed. The extremity X-ray is highlighted by its automated patient positioning guidance system, potentially allowing patients to perform their own scans, while its small form factor (~ 20x20x51”) and low costs could make it an option for non-traditional settings (e.g. ski resorts).
- Rethinking Patient Consent: Considering that healthcare AI was barely an idea when HIPAA guidelines and patient consent forms were drafted, healthcare providers need to modernize their consent practices. That’s from a new JACR editorial that proposes: 1) Patients understand what they are consenting to; 2) Informatics teams understand all aspects of data-contracting and data-sharing; and 3) All staff “provide a trustworthy setting for patient data.” The authors also proposed a tiered consent system that allows patients to decide which data can be shared (e.g. images), how it can be used (e.g. AI development), and with whom it can be shared (e.g. for-profit AI developers).
- QT Imaging in EMEA: QT Imaging’s QTscan transmission ultrasound breast imaging system is on its way to the European, Middle Eastern, and North African markets through a distribution agreement with Freedom Ventures. Freedom Ventures will also assist QT Imaging’s CE Mark approval and regional registration processes.
- Patient Portal Availability & Adoption: A Clinical Imaging survey of 31 U.S. academic hospital systems revealed that 94% of these institutions have patient portals, 78% of which automatically post radiology reports (avg. delay: 4 days; range: 0-7). These portal availability numbers are pretty high, but 80% of the institutions revealed that less than half of their patients actually use their portals. The survey found no correlation between patient participation and length of the portals’ radiology report delays.
- Home Breast Cancer Screening Spotlight: The Blue Box home breast cancer screening device became a mainstream news topic in recent weeks (Vogue, Yahoo News, Daily Mail, more) after its inventor received the James Dyson Award for her efforts to allow home screening. The device scans urine samples for breast cancer biomarkers, which are analyzed by a cloud-based algorithm, and the results are sent to the user’s phone. The early-stage device doesn’t technically involve imaging, but if it reaches clinical use it would live within a very imaging-focused screening continuum.
- HAP’s 2021’s Code Change Breakdown: Healthcare Administrative Partners provided a super detailed breakdown of the coding changes impacting radiology practices this year. The good news for many of you is that these changes will have a minimal impact on diagnostic radiologists’ daily work (IRs and radiation oncologists face more changes), and DRs avoided fee reductions for all services except low-dose CT Lung Screening (fee schedule -38% to $150.74).
- CathWorks’ $30M: Angiogram image analysis firm CathWorks just wrapped up a $30M Series D funding round that it will use to support the commercialization of its FFRangio system. The CathWorks FFRangio System transforms routine angiogram images into physiology information, including color-coded 3D renderings of blood flow, to support coronary artery disease decision making.
- About Your ML Problem: Boston Review shared its take on “Medicine’s Machine Learning Problem,” suggesting that the growing role of ML in healthcare is further disempowering patients and providers. To address this, the article suggests that we should: 1) Acknowledge that medical data can be wrong (incomplete, incorrect, missing, and biased); 2) Understand that ML systems can centralize power at the expense of patients and providers; 3) Consider how ML will impact a medical system “that is already disempowered;” 4) Accept that machine learning must not replace the domain expertise of patients and doctors; and 5) “Move the conversation around bias and fairness to focus on power and participation.”
- Rad Res’ Low Stipends: A new JACR study highlighted diagnostic radiology residents’ financial challenges (n = 118 DR resident programs, 77 cities), noting that current stipend levels ($49k to $72k per year) can feel much lower when adjusted for regional living costs ($30k to $69k). The authors encouraged incoming residents to consider cost of living in their residency decisions, noting that residencies in large cities had the greatest negative cost of living impacts (NYC = -53.8%) while cities with lower living costs can increase the value of stipends by as much as 21.7%.
- Ultromics’ EchoGo Pro FDA: Ultromics announced the FDA 510(K) clearance and U.S. launch of its EchoGo Pro cardiac decision support tool (already has CE Mark), which uses AI to analyze echo exams for coronary artery disease. Ultromics will offer EchoGo Pro as a module in its cloud-based EchoGo suite, alongside its already-available EchoGo Core tool (performs systolic function and strain analysis).
- Limiting Radiologist Workloads: A JACR opinion piece proposed implementing national radiologist workload limits that would cap shifts at 10 hours and limit hourly interpretation volumes. Although the author expects resistance (financial & freedom objections), he suggests that these changes will reduce medical errors that have been found to correlate with higher caseloads and longer shifts.
- Nanodroplets & Ultrasound Drills: A North Carolina State research team developed a new technique that uses nanodroplets and an ultrasound “drill” (a small transducer that fits into a catheter) to break up especially dense retracted blood clots. The nanodroplets are made out of small lipid spheres filled with liquid perfluorocarbons, which penetrate clots and convert to microbubbles when they are exposed to ultrasound. The ultrasound waves then cause the microbubbles to vibrate and help break-up the clots, reducing their size by an average of 40% when combined with drug treatments (30% on their own). This isn’t our type of ultrasound, but it is another example of ultrasound technology’s expanding role in medicine.
- Neurodiagnostics’ High Costs: A new University of Michigan study detailed the rising out-of-pocket costs of neurologic diagnostic services, warning that these increases may cause patients to forgo evaluations. Using data from 3.7M patients who had a neurologic visit or diagnostic test from 2001-2016, the researchers found significant increases in the ratio of patients who paid some out-of-pocket costs for diagnostic tests (23.1% to 69.5%) and a similar increase in the cost of out-of-pocket MRI exams ($84 to $242). Other non-imaging neurologic services similarly increased, adding to this trend.
The Resource Wire
– This is sponsored content.
- Zebra-Med’s AI1 Imaging Analytics Engine can be fully embedded in PACS/RIS systems, making clinical insights natively accessible in the PACS viewer to support interpretations or seamlessly in the worklist to support prioritization.
- In this Nuance Q&A, Nines CEO David Stavens, PhD, and medical director Dr. Jean-Paul Dym, MD discuss their experiences with PowerScribe One and their perspectives as a Nuance innovation partner.
- Register today for this webinar from Healthcare Administrative Partners where they will examine how the 2021 Medicare Physician Fee Schedule (MPFS) and Quality Payment Program (QPP) final rules will impact physicians, with a focus on radiologists.
- Watch Jared Christensen, MD, MBA explain how Duke University Health uses Riverain Technology’s ClearRead CT Vessel Suppress and ClearRead CT Detect in its daily practice.
- This UCSD study shows that fully automated ventricular segmentation using Arterys’ Cardio AI provides contours and volumes that can be used to aid expert segmentation.
- Learn how GE-enabled rapid breast cancer diagnosis clinics are reducing diagnoses-to-consultation turnarounds “from weeks to hours,” and how they are helping patients and clinicians in this new GE Healthcare Insight.
- Siemens Healthineers detailed the six steps that radiology practices can follow to build resiliency, providing a roadmap detailing the workflow, leadership, and financial changes required to successfully respond to the pandemic and come out stronger because of it.
- Learn how Windsong Radiology Group used Bayer’s MEDRAD Stellant FLEX CT injection system to drive cost efficiencies and standardization across its imaging centers.
- Easy-to-use customizable templates, remote access, EMR integration. These are three of the five Cardiovascular Information Systems “must haves” detailed in this new Hitachi Healthcare Americas blog.