Uncategorized

Imaging Wire #300

“What sounds like sci-fi is now becoming a reality.”

UK NHS chief executive, Sir Simon Stevens, on diagnostic “pill cameras” becoming a reality.


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


The Case for MRI-First Prostate Screening

MRI’s role in prostate cancer screening just got another boost with a new JAMA study that found MRI-first screening to be more effective (clinically and financially) than biopsy-first screening.

  • The Study – The researchers used a hypothetical cohort of 4.48m 55-69yr men to compare MRI-first and biopsy-first screening, either based on patients’ age or risk profile.
  • Age-Based Results – With the age-based screenings, the MRI-first method reduced prostate cancer deaths by 0.9%, cut overdiagnosis by 14.9%, and reduced biopsies by 33.8% (compared to biopsy-first).
  • Risk-Based Results – MRI-first screening was even more effective with a risk-based approach (compared to MRI-first, age-based). Performing MRI-first screenings on men with ≥2% or ≥10% prostate cancer risks, reduced overdiagnosis by 10.4% & 72.6%, while leading to 21.7% fewer MRIs and 53.5% fewer biopsies over 10 years.
  • The MRI-First Sweet Spot – The researchers proposed MRI-first screening for men with ≥3.5% risk of developing prostate cancer in 10 years, suggesting that it would prevent 16.5% of prostate cancer deaths while reducing overdiagnosis by 27%.
  • Prostate MRI Momentum – MRI prostate cancer screening is on a roll, as this study comes just a few weeks after a separate study found that Short MRI might also be a better screening option than PSA testing.



AI Buyers Guide

European Radiology just published a helpful set of guidelines that radiology practices and departments can follow when evaluating imaging AI solutions. Here’s what they recommend:

  • Make Sure it’s Relevant – AI buyers should make sure they actually need the solution they’re evaluating by defining the problem the solution solves and confirming how it’s used, its output, benefits, and risks.
  • Confirm it’s Performance – They should confirm that the solution is well made by reviewing its design specifications, training/validations/testing process, and documented biases.
  • Usability and Integration Check – Radiology practices/departments should then make sure the solution would actually work for them, by evaluating its workflow fit, IT requirements, interoperability, data accessibility, and required interpretation tools (i.e. visualization).
  • Regulatory Review – Then (or maybe as a first step) they should make sure the solution has local regulatory approval and complies with their country’s data protection rules.
  • Financial Fit – After all that, radiology practices/departments can finally look into the solution’s economic and support factors, evaluating the AI vendor’s licensing model (and cost implications) and how they’ll support training/onboarding, maintenance, and future malfunctions.
  • The Takeaway – Although the main takeaway of this one is pretty straightforward (these guidelines might be useful for AI buyers), it’s worth noting this is part a recent influx of articles focused on buying and/or integrating AI (here are a few previous papers). Even if we’re still making our way toward the commercial stage of the AI hype curve, these recent papers suggest we’re getting closer.

The Wire

  • PillCam on Trial: The UK NHS is trialing PillCam “pill cameras” that capture two images per second as they pass through patients’ digestive tract, potentially allowing remote digestive exams (bowel cancer, Crohn’s disease). The trial will initially include 11k patients, evaluating the PillCam’s accuracy and whether it could make the NHS’ digestive screening more efficient and scalable. We’re definitely seeing momentum in the capsule endoscopy arena, which also includes Check-Cap (low-dose X-ray) and InsideOut (GI tract videos).
  • The MRI CT Bone Scan: Belgium’s Ghent University Hospital performed what they’re calling the first radiation-free “CT bone scan,” using MRIguidance software to transform MRI images into CT-quality 3D bone scans. Ghent and MRIguidance are excited about this breakthrough, calling its benefits (no radiation, soft tissue + bone imaging w/ 1 scan) a “game changer” and suggesting that it will become “the standard” within a few years.
  • Chest CT AI’s Significant Value: New research out of the Medical University of South Carolina (MUSC) highlighted the “significant added value” that AI could provide chest CT reporting workflows. The researchers applied Siemens Healthineers’ AI-Rad Companion Chest CT to 100 consecutive patients’ non-contrast CTs and compared the AI findings against the original radiology reports. The AI solution outperformed the cardiothoracic radiologists’ reports for identifying aortic dilatation (AI AUC 0.89 vs. Report AUC 0.63) and coronary artery calcifications (0.95 vs. 0.85), while falling short of the reports for identifying pulmonary lesions (0.88 vs. 0.99) and vertebral compression fractures (0.82 vs. 1.0).
  • Imaging AI’s $2.6B: Medical imaging AI companies raised more than $2.6b in venture funding since 2014 (>185 companies), with about 70% of that funding happening during the last three years (2018 = $790m, 2019 = $450.2m, 2020 = $592.3m). That’s from a new Signify Research report that also noted a recent funding shift to later-stage AI firms (Series B and higher), as VCs focus more on growing established AI companies rather than nurturing early-stage startups. Signify also highlighted Chinese AI companies’ rising funding share (45% of 2020 total vs. 26% of 2019 total) and slowing funding for Western firms.
  • CTC Advantage: CT colonography with a 10mm threshold is the most effective noninvasive colorectal screening method, improving detection and decreasing unnecessary colonoscopies. That’s from an AJR review of 125 previous studies that measured advanced neoplasia detection using multi-target stool-DNA (mt-sDNA), fecal immunochemical test (FIT), and CT colonography (CTC-10mm and CTC-6mm). Although CTC-6mm achieved the highest cancer detection rate (4.8% vs. 4% CTC-10mm, 3.4% mt-sDNA, 2% FIT), CTC-10mm had by far the highest positive predictive value for advanced neoplasia (61% vs. 34.4% CTC-6mm, 31.8% FIT, 26.9% mt-sDNA).
  • Tube Teamwork: Here’s a study that can be pretty straightforward in some countries and might lead to interpretation turf debates in the US. A UK-based study highlighted the benefits of relying on radiographers to immediately interpret pediatric X-rays for correct nasogastric feeding tube placement. After training radiographers on a nasogastric tube position interpretation / communication pathway, the researchers reviewed all pediatric nasogastric tube exams over a 13-month period (n = 282). The study found that the radiographers correctly followed the pathway with 92.9% of the exams, while 97% of the 240 radiographer comments were accurate. Over the study period, the radiographers identified four (1.7%) misplaced tubes, leading in their removal or re-siting.
  • Sectra Signs Emory: Sectra signed a five-year enterprise imaging contract with Emory Healthcare that will provide a range of modules and tools (radiology, all imaging subspecialties, orthopedics, image archiving, visualization, zero-footprint viewer, analytics) across Emory’s network (11 hospitals, 250 clinics), while linking Emory with the Grady Health System.
  • The SubtlePET Effect: A team from major European imaging center company, Affidea, released a study detailing how Subtle Medical’s SubtlePET PET/CT reconstruction solution allowed them to reduce 18F-FDG PET/CT radiopharmaceutical dosage without affecting interpretative quality. The researchers performed one low-dose CT and two sequential PET scans (“PET-processed” scans with 66% 18F-FDG dose & “PET-native” standard-dose scans) on 61 patients, and then used SubtlePET to enhance “the PET-processed” scans. They found no significant difference between the lower-dose “PET-processed” and standard-dose “PET-native” scans’ detection ability, with “almost perfect” inter-reader agreement, and no false positives.
  • Imaging Informatics Growth: Frost & Sullivan forecasts that the global imaging informatics market will grow at a 3.5% compound rate over the next five years, reaching $10.4b in 2025 (vs. $8.5b in 2019). The firm attributed this growth to the evolution of AI and cloud technologies, providers’ changing administrative needs, new regulations, and a post-COVID rebound.
  • LUS for Pediatric Pneumonia: New research out of Italy provided more evidence supporting lung ultrasound’s (LUS) value for diagnosing and monitoring community‐acquired pediatric pneumonia (CAP), even if chest x-ray remains the gold standard. The researchers performed LUS and CXR scans on 68 children (41 w/ CAP, 27 healthy) finding that LUS correctly diagnosed 40 of the 41 CAP patients (97% sensitivity, 96% specificity) and only required 5-10 minutes per exam. LUS’ sensitivity improved to 100% when performing 30-day follow-ups, although specificity declined to 94%.
  • Stroke Ambulance Struggles: The American Heart Association warned that the role/availability of stroke ambulances will remain limited without reimbursement changes, revealing that “nearly all” mobile stroke units rely on private gifts, grants, or institutional support to operate. The AHA called for CMS to create reimbursements for mobile stroke tests (including CT scans) and treatments in order to expand the country’s fleet of stroke ambulances.
  • ULDCT + DLNR, an Emphysema CT Alternative: Ultra-low-dose CT combined with deep learning noise reduction software can be used to measure emphysema in COPD patients with similar effectiveness as standard CT, while reducing radiation exposure by 16.6% to 84%. That’s from a new EJR study that performed ULDCT+DLNR and SDCT scans on 49 COPD patients, finding that applying DLNR to ULDCT scans also reduced variability by 24%-27% (vs. SDCT scans).
  • Delayed CT Lawsuit in Tampa: Bayfront Health St. Petersburg and Tampa Bay Radiology Associates are facing a medical malpractice lawsuit after a 6.5-hour turnaround for a stroke patient’s “STAT” head CT interpretation led to a “catastrophic” 12-hour treatment delay. The male patient initially underwent a mechanical thrombectomy for a large ischemic penumbra, and when his follow up CT revealed a small ganglia hemorrhage (no shift or IVH) his care team ordered continued CT monitoring. However, a “STAT” CT order at 2:00pm that afternoon was not performed until 3:05pm (it was supposed to be within 45 minutes) and the scan was not interpreted until 8:35pm. That study revealed mass effect and an 8mm right to left midline shift.

The Resource Wire

– This is sponsored content.

  • Every decision United Imaging makes, every innovation they choose to pursue, comes down to their mission: Equal Healthcare For All. Hear more about their mission.
  • This Riverain Technologies case study details how Duke University Medical Center integrated ClearRead CT into its chest CT workflows, reducing read times by 26% and improving nodule detection by 29%.
  • 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.
  • CD burning issues? Check out this one minute video showing how Novarad’s CryptoChart image sharing solution allows patients to easily access and share their medical images using personalized, highly secure QR codes (rather than CDs).

You're signed up!

It's great to have you as a reader. Check your inbox for a welcome email.

-- The Imaging Wire team