Butterfly iQ+ | RadRes Reps | Psych AI


“. . . one step closer to realizing the full potential of bedside point-of-care ultrasound as the stethoscope of the future—a true window into the body,”

Butterfly Network’s Chief Medical Officer, Dr. John Martin, on the new iQ+ ultrasound and the company’s big goals.


If you or a colleague made outstanding contributions to radiology practice this year, please take three minutes to nominate them (or you) for a Imaging Wire Awards.


Imaging Wire Sponsors

  • 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.

The Imaging Wire



Butterfly iQ+

Butterfly Network just announced its new Butterfly iQ+ handheld ultrasound system, adopting the same tough-to-match $2k price point as the original iQ, while making a range of improvements.

  • More Power – The Butterfly iQ’s launch had a much greater focus on technology than we typically see with medical products, using terms like “Power meets Performance” and beginning most of its upgrade headlines with “More” (More: Clarity, Durability, Power, Control).
  • What More Means – The iQ+’s upgrades definitely have plenty of clinical relevance. Its clarity improvements (+15% frame rates, +60% pulse repetition frequency) directly support Butterfly’s live imaging value proposition, which has been a key target of competitor criticism. Meanwhile, its smaller form factor (-15% head, -10% length), lower power consumption (+20% battery, up to 2x longer runtime), and greater durability (4ft drop, replaceable cord) directly address the needs of real world clinical users.
  • New Needle Viz – Butterfly also added to its anesthesia, emergency, bedside, and MSK value prop with its new Needle Viz technology, which helps clinicians visualize a needle to support a range of procedures.
  • Butterfly Marketing – Butterfly remains among the imaging industry’s best marketers, creating impressive buzz for its launch (with plenty of retweets), holding a solid live launch event, and posting a sleek B2C-styled iQ+ product page. Butterfly might be better suited for this kind of marketing (just one product to promote, unique consumerized healthcare angle, massive funding), but that doesn’t take away from the fact that they do a great job with this kind of stuff.

The Wire

  • COVID & Resident Reps: A new Henry Ford study revealed that radiology residents’ image reading training volumes declined by an average of 62.8% during the COVID-19 pandemic, when the trainees lost a median of 14.5 days of reading room experience. The pandemic’s double whammy of lower imaging volumes and social distancing policies created the largest declines for R1s and R2s (-87.3% & -64.3%) and mammography, MRI, and nuclear medicine studies (-92%, -73.2%, -73%).
  • Psych AI: Georgia State researchers received an $875k grant to continue developing a fMRI-based machine learning system that could help psychiatrists predict mental health patients’ response to different medications (currently 90% accuracy). They will use the grant to refine the algorithm with scans from a more diverse set of patients and fMRI machines.
  • RBMA Warning: RBMA became the latest radiology group to warn CMS that its proposed 2021 Medicare payment changes (radiology down ~11%) would hurt practices and patient care, calling for CMS to delay the changes and/or wave budget neutrality requirements. RBMA supported this letter with a practice survey (n = 155) indicating that these changes would force practices to cut staff and benefits (70%), consider furloughs (>50%), delay equipment upgrades (63.8%), consider closing imaging centers or reducing hours (34%), or consider restricting service to Medicare beneficiaries (34%).
  • Epic’s COVID AI: A University of Minnesota team, in collaboration with M Health Fairview and Epic, developed a CXR-based COVID-19 detection algorithm that they will provide at no charge within Epic’s App Orchard (starting with 12 M Health Fairview hospitals). The algorithm (trained w/ 100k negative and 18k positive CXRs) automatically evaluates CXRs taken when patients arrive at the ED, estimating the likelihood that the patient is COVID-positive within “seconds.”
  • NVIDIA & MGH’s CORISK Model: NVIDIA and MGH researchers developed an AI model that can predict whether an ED patient with COVID-19 symptoms will require supplemental oxygen with a 0.94 AUC. NVIDIA and MGH developed the ‘CORISK’ model using a combination of CXR images and health records sourced from a 20-hospital federated learning initiative (using the NVIDIA FL framework).
  • Solis & UChicago: UChicago Medicine and Solis Mammography will open two new breast health centers, delivering Solis’ trademark “peace of mind mammogram” experience (more comfortable mammograms, spa-like centers) and staffed with UChicago Medicine’s radiologists and technologists. Solis Mammography operates about 70 imaging centers across the country and has a number of similar hospital partnerships.
  • SyntheticMR & GE: GE Healthcare will integrate SyntheticMR’s SyMRI quantitative MRI software into its MR platform, calling the new solution ‘MAGiC Neuro.’ The new MAGiC Neuro solution expands upon GE Healthcare’s previous MAGIC solution (also developed by SyntheticMR), adding support for tissue volumes, myelin-correlated volumes, and quantitative data.
  • Gadavist’s CMR Support: Bayer highlighted results from a pair of Phase III trials (n = 376 & 388 adults) that supported the effectiveness of its Gadavist contrast agent for cardiac MRI. The studies found that Gadavist had “high diagnostic accuracy” with myocardial perfusion (stress, rest) and late gadolinium enhancement (LGE) among adult patients with known or suspected coronary artery disease (CAD). Because of results like this, last year Gadavist became the first contrast agent approved by the FDA for cardiac MRI.
  • Dyad’s $3.5M: Cardiac imaging AI startup, Dyad Medical, announced $3.5M in new funding and revealed plans to use the capital to support its R&D and FDA clearance efforts. Dyad Medical’s cloud-based platform would provide cardiologists with “instantaneous cardiac image interpretations” using scans from any modality, on any part of the heart, and from any device.
  • CycleGAN Augmentation: NIH researchers developed a generative adversarial network (CycleGAN) to augment labeled CT datasets, which could create larger and more-generalizable sets for AI training. The team trained the CycleGAN to transform contrast CT images into non-contrast images using a large image database and then combined the original contrast images and synthetic non-contrast images. They compared the CT segmentation performance of a U-Net trained on the original dataset with a U-Net trained on the combined dataset (contrast + synthetic non-contrast CT), finding that some CT segmentation tasks improved significantly with the augmented dataset.
  • Faster DBT Reads: A University of Michigan team developed a DBT reconstruction protocol that could reduce interpretation times (a key challenge vs. DM) without affecting diagnostic performance, by using thinner slabs (6mm vs. 10mm), a narrower overlap (3mm vs. 5mm), and eliminating 1mm slices. To study this method, four blinded radiologists read 122 DBT exams, finding that the mean interpretation time was 12% to 18% shorter for three of the readers, while reducing mean image count by 72%.
  • Prenatal 4D Blood Flow MRI: A King’s College London team developed a 4D MRI prenatal blood flow visualization and quantification method that could improve physicians’ ability to detect prenatal congenital heart disease. The 4D MRI fetal heart imaging system uses MRI velocity-encoding, combined with a motion-robust reconstruction framework (creates 4D images with multiple 3D scans), to show both heart structure and blood flow.
  • ConVIRT Learning: A Stanford-led team proposed a self-supervised learning method that contrasts true image-text pairs for AI pre-training (vs. random), highlighting its advantages over ImageNet and its potential to reduce AI training’s image labeling labor barrier. The team leveraged their new ConVIRT (Contrasive VIsual Representation learning from Text) method to pretrain chest and bony image encoders and then used it for four classification tasks, finding that ConVIRT only required 10% as much labeled data as an ImageNet-based model and performed as well or better.
  • DystoniaNet: A Massachusetts Eye and Ear team developed a deep learning system that can accurately detect dystonia (a neurological movement disorder) in brain MRIs in less than a second. The team used their DystoniaNet AI system to compare brain MRIs of 612 people (392 w/ dystonia), achieving 98.8% accuracy and interpreting scans in 0.36 seconds. This could lead to the first diagnostic test for dystonia, which would be “transformative” since the disorder often goes undetected for up to 10 years.

The Resource Wire

– This is sponsored content.

  • This Nuance blog details how it helped Northwell Health quickly expand its use of PowerShare to support New York City radiologists in their response to COVID-19.
  • Did you tune into the Siemens Healthineers MRI Summer Webinar Series? If not, you still have the chance. The series featured 5 leading MRI experts that discussed hot topics within the magnetic resonance imaging world. Register for access to watch the webinar recordings at your own convenience.
  • Learn how GE Healthcare’s new women’s ultrasound auto recognition tool automates physician tasks and improves confidence.

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.

Join thousands of imaging professionals.