Project Nightingale | Fast & Accurate On-Site FFR-CT

“I hope to bring closer these unsung medical imaging scanners to the world, reintroduce them to anxious patients like kids, make it fun and interactive.”

UK-based MRI tech, Apollo Exconde, on the inspiration behind his Open MRI LEGO toy design.


Imaging Wire Sponsors

  • Carestream – Focused on delivering innovation that is life changing – for patients, customers, employees, communities and other stakeholders
  • Focused Ultrasound Foundation – Accelerating the development and adoption of focused ultrasound
  • Medmo – Helping underinsured Americans save on medical scans by connecting them to imaging providers with unfilled schedule time
  • Nuance – AI and cloud-powered technology solutions to help radiologists stay focused, move quickly, and work smarter
  • Pocus Systems – A new Point of Care Ultrasound startup, combining a team of POCUS veterans with next-generation genuine AI technology to disrupt the industry
  • Qure.ai – Making healthcare more accessible by applying deep learning to radiology imaging

The Imaging Wire


Project Nightingale

Google spent the first half of this week in the hot seat after it was revealed that the company is involved in a partnership with Ascension that gives Google access to tens of millions of patients’ healthcare data.

Project Nightingale – Google’s alliance with Ascension (the 2nd largest U.S. health system) began last year, providing Ascension with a range of Google services (storage, analytics, etc.) and giving Google access to relatively complete patient histories (lab results, radiology scans, doctor diagnoses, hospitalization records, names, birth dates, and more) without informing the patients or doctors involved.

Google & Ascension’s Motives – The Project Nightingale system analyzes patient data and provides Ascension with insights into how it could improve treatment and administrative operations (e.g. procedures, billing, staffing, etc.). However, Google was also using the project to design future AI-based healthcare data solutions that it will sell to other hospitals — and Google reportedly didn’t charge Ascension for much for development. It’s quite possible that Google’s deals with other major healthcare providers have a similar structure and strategy.

Legal & Compliant – Although Project Nightingale drew plenty of ethics-based criticism, both Google and outside privacy experts confirm that this was legal and compliant. Hospitals can share data with business partners without telling patients as long as the information is used to help the hospital “carry out its health care functions,” plus Google had the necessary agreements in place. Also, despite widespread coverage calling this a “secret project,” Alphabet technically revealed the partnership in its Q2 2019 earnings call (it just didn’t get much headlines then).

Google’s Not Alone – While Project Nightingale is perhaps the largest effort yet, patient data is at the center of most tech giants’ (Amazon, Apple, Microsoft, etc.) healthcare expansion strategies. That’s a pretty good approach given these companies’ traditional strengths, the need to improve the way healthcare manages and uses data, and the opportunities to build solutions based on healthcare data. However, as proven by this week’s reaction to Project Nightingale, these strategies will lead to privacy concerns even when rules are followed.

Fast & Accurate On-Site FFR-CT

New research in Cardiothoracic Imaging found that on-site FFR-CT is superior to coronary CT angiography for the identification of functionally significant coronary artery stenosis among patients with suspected/known CAD. This is relatively notable result and adds even more evidence supporting FFR-CT.

The retrospective study looked at 77 vessels in 57 patients who underwent CCTA up to 60 days before an invasive FFR, finding that on-site FFR-CT (not sent to a 3rd party site for analysis) was able to identify functionally significant stenosis far better than CCTA-alone (0.87 vs. 0.70 AUCs, 83.1% vs. 64.9% accuracy, 77.5% vs. 42.5% specificity, 89.2% vs. 89.2% sensitivity). The onsite FFR-CT model was also able to complete its analysis in about 36 minutes, versus several hours for offsite FFR-CT analysis.


The Wire

  • A UK-based MRI radiographer’s Open MRI Lego design is halfway to being accepted as an official Lego toy design after receiving over 5,170 votes. Created with young and anxious patients in mind, the MRI shows patients what happens in the different stages of the MRI scanning process (e.g. head, body, coil placement) and can also be modified to create Upright MRI, CT scan, and DEXA scan designs. The Open MRI needs 10k total votes within the next 6 months in order for it to become an official Lego design.
  • In other Google data privacy news. The University of Chicago asked to be dismissed from a class action lawsuit accusing Google and the University of Chicago Medical Center of improperly sharing patient data, arguing that the plaintiffs haven’t shown that any patients were harmed or even identified. Google and the University of Chicago Medical Center were sued in June for allegedly sharing hundreds of thousands of patients’ records that still had identifiable information (dates, notes).
  • A survey of Swiss radiologists (59), surgeons (56), and medical students (55) revealed a mix of expectations and concerns about the future of radiology. AI was of course a hot topic, as the majority of respondents viewed the use of AI as a radiology support system favorably (median 8 positive score out of 10). Medical students were more concerned than radiologists that AI could threaten diagnostic radiology careers, with 26% of students revealing that they don’t intend to specialize in radiology in part because of the threat of AI. Meanwhile, radiologists were more concerned about “turf losses” from other disciplines than medical students.
  • A new article in AJR highlighted the “profound impact” deep learning has had on quantitative cardiac MRI analysis, suggesting that it still holds “great promise for future use in clinical practice and scientific research.” The authors touted cardiac MRI DL’s flexibility (e.g. structure & function, myocardial scarring assessment) and ability to significantly reduce interpretation times, particularly benefiting less-experienced radiologists and high-volume imaging centers.
  • Carestream announced an updated version of its DRX-Revolution Mobile X-ray System, applying customer feedback to improve its ergonomic design (lighter & balanced tube head and collimator, quieter breaks/engine/tube head, better sealing for cleaning, lockable detectors) and user workflow (responsive display screens at the tube head and main display, new functional LED lighting, in-bin detector charging).
  • A new study out of Singapore revealed that the first AI-powered ultrasound-guided automated spinal landmark identification system (uSINETM) improved needle insertion accuracy and success rates during spinal anesthesia. uSINETM uses anatomical landmarks during ultrasound scans, alerting the anesthetist in real-time when the right location and angle for insertion are found with a 92% first attempt success rate (n = 100 women). With this study wrapped up, the team will next look to improve upon uSINETM and investigate its use with high-risk patients, while also working to commercialize the system.
  • Hologic announced the FDA clearance of its 3DQuorum Imaging Technology, a new algorithm that works with Hologic’s Clarity HD technology and is based on the company’s new Genius AI artificial intelligence platform. 3DQuorum reconstructs 3D imaging data into 6mm “SmartSlices” and helps identify clinically relevant regions of interest, allowing a 66% reduction in image volume and a 12.5% reduction in interpretation time.
  • Chinese researchers developed a new CT coronary artery calcium score (CACS) quantification algorithm that could replace the manual cardiac risk assessment process, potentially improving efficiency and accuracy. Using CT CACS scan data from 530 patients (300 for training, 90 validation, 140 testing), the team developed a deep learning algorithm to automatically quantify CAC scores. The algorithm produced Agatston, mass, and volume scores that were statistically similar to the manual method, but in far less time and requiring much less radiologist labor.
  • Guerbet announced a partnership with InterSystems, using the company’s “IRISTM for Health” data management platform to better integrate Guerbet’s Contrast&Care contrast media injection management solution into hospital and imaging center IT systems.
  • A new report reveals that three patients died and five other patients suffered permanent harm after London’s St George’s University Hospitals failed to identify diseases in scans or appropriately follow-through on radiology findings. A review found that staff missed cancers in scans, failed to escalate incidental cancer findings, improperly reported results, and sent diagnoses to unmonitored inboxes, before implementing changes to its radiology reporting structure and workflow.

The Resource Wire

  • This Carestream Special Report details how providers can get the greatest ROI from their X-ray technology as radiography demands increase and budgets head the other direction.
  • Yale University research reveals that the average patient drives past SIX lower-cost providers on the way to an imaging procedure, due in large part to patients’ and physicians’ limited cost consciousness. Medmo helps address this issue by letting patients enter what they can afford for their scan, then booking them at a nearby imaging center willing to accept that rate.
  • U.S. News’ recent “Medical Marvels” article told the story of five patients who addressed their chronic pain through focused ultrasound therapy.

Leave a Reply

Basic HTML is allowed. Your email address will not be published.

Subscribe to this comment feed via RSS

Get every issue of The Imaging Wire

Delivered right to your inbox. We will never sell or share your information.