“This is a win-win for both the patient and the physician—it is taking advantage of modern healthcare technology at no additional cost to the patient.”
Kansas Assistant Insurance Commissioner, Clark Shultz, on the state’s decision to require that insurance plans support 3D mammography.
Intel and Philips Make Their Case for CPUs in AI
Intel and Philips announced a collaboration, combining Intel’s Xeon CPUs and OpenVINO toolkit with Philips’ medical imaging deep learning inference models. The collaboration yielded impressive speed improvements, reducing the time required for X-ray bone-age-prediction modeling by 188-times and cutting CT lung-segmentation modeling by 38-times versus baseline measures. However, this announcement was not about an AI inference modeling breakthrough. The goal of this announcement was to help reinforce CPU technology’s place in medical imaging AI. The companies argued that although GPUs work well with images, their inherent memory limitations create challenges when building some AI/ML models, while CPUs can handle more “memory-intensive models typically found in medical imaging” at lower costs than GPUs. For Intel, this means its cornerstone chip technology may pose more AI competition to NVIDIA’s GPU chips than some may think. For Philips, this means its HealthSuites Insights platform can deliver AI solutions from an online marketplace without requiring end-user investments in pricey GPU-based engines. It’s harder to say what this means for the overall medical imaging AI segment, which still seems to be driven by GPU technology – at least for now.
IAMERS’ International Evidence
IAMERS continued its campaign for service rights, publishing an editorial on dotmed.com that highlighted Siemens Healthineers’ recent anti-competitive violations against 3rd party servicers in Korea (and $5.7 million fine) and the Italian government’s ongoing investigation into similar activity from Siemens, GE, and Philips. The service lobbying group led-off by highlighting the Korean government’s decision that Siemens gave unfavorable terms (price, function, and service key turnaround times) to independent service organizations and sent “exaggerated notices” to hospitals about the risks of signing with ISOs. The Italian Competition Authority is still investigating a claim that the manufacturers tried to restrict/exclude service competitors by limiting access to service keys and other necessary maintenance materials (management software and service manuals), refusing to provide original replacement parts, and performing “denigratory actions” to make hospitals believe that only OEMs can properly deliver service. IAMERS is doing its job here, but there are certainly two sides to this debate, which is expected to continue in the courts and in the press.
The State of the MRI Market
An Imaging Roundtable discussion published on 24x7mag.com provided a range of insights into the state and future of the MRI market from the perspective of MR leaders at Philips Healthcare, Siemens Healthineers, and Zetta Medical Technologies.
- The executives attributed MRI’s growth to increased demand for key procedures (musculoskeletal, body oncology, and cardiovascular), recent technology innovations improving patient experience and image quality, rising demand from the aging population, and increased adoption of MRI in emerging markets.
- When asked about the biggest innovations taking place in MRI, the executives cited precision medicine, artificial intelligence, the commercialization of new combined technologies like PET/MRI, new clinical use 7T MRIs, and efficiency-creating acceleration techniques.
- Discussing the biggest challenges facing MR and efforts to overcome these issues, the panelists cited helium shortages (addressed by monitoring systems), MRIs’ inherent site, time, cost, and service disadvantages versus other modalities (addressed by new technological advancements), provider profitability issues (addressed by improved efficiency), and patient comfort issues (addressed by focused improvements on the patient experience).
- From a service perspective, the executives emphasized the growing benefits of remote diagnostics and service, the responsibility of providers’ staff to keep MRI scanners secure (this is a loaded one), and the need to establish solid preventative maintenance processes.
- Looking to the future of the MRI market, the panelists forecast that MRI AI will continue to grow in importance, new automation and quantitative imaging techniques will become more prevalent, MRI exams will become faster helping expand the modality to new settings like the ER, MRIs will become more common in interoperative and radiation oncology settings, and the systems will take on a greater role in precision diagnostics and precision medicine.
- Guerbet launched a partnership with Imalogix, offering the company’s radiation dose management platform alongside its own imaging technology, helping its healthcare clients understand and improve patient radiation dose management.
- Kansas announced that starting in 2019 all health insurers in the state will be required to cover 3D mammograms, making it the 9th state to adopt this requirement and following a similar policy change in neighboring Missouri last month (DBT also covered in: Arkansas, Connecticut, Illinois, Maryland, New York, Pennsylvania, and Texas).
- Illinois became the 36th state to enact a breast density notification law, following similar moves from Wisconsin, Utah, Washington, and Florida since March. In addition to expanding awareness, these laws appear to be driving more imaging volume, particularly for follow-up ultrasounds.
- RadNet announced a 5-year full-risk “capitation contract” with EmblemHealth, assuming imaging responsibilities for some EmblemHealth AdvantageCare enrollees and taking over imaging operations at 26 AdvantageCare locations. The partnership is RadNet’s first full-risk capitation contract on the East Coast, which the company highlighted as “an important strategic development” after operating similar capitated relationships with dozens of large medical groups on the West Coast for over 20 years. Here is the Wiki for capitated relationships.
- Brain imaging AI company, MaxQ-AI, will seek an $8 million IPO on the Nasdaq, using the funding to achieve FDA clearance for its Accipio software portfolio. MaxQ-AI gained European CE clearance for its AccipioIx solution in May and has agreements in place with GE, Samsung NeuroLogica, and IBM Watson, but is not yet cleared to offer any of its solutions in the US.
- Fujifilm Sonosite launched the SonoSite iViz C60v deep application-targeted transducer (abdominal, lung and obstetrics) and the SonoSite iViz L25v superficial application-targeted transducer (ophthalmic, arterial, venous, lung and nerve), both of which are intended for use with the company’s SonoSite iViz point-of-care ultrasound system. The transducers first launched in Japan in May and appear to be well on their way to a global rollout.
- ViewRay announced a $150M common stock offering, making 1.6 million shares available for $9.25 per share, and using the funds for working capital, R&D, commercial expenses, and infrastructure expenses. The stock offering follows continued quarterly losses and comes on the heels of a C-level leadership overhaul in late July.
- Georgetown University made a cache of brain cancer biomedical data available to worldwide researchers. The REMBRANDT (REpository for Molecular BRAin Neoplasia DaTa) dataset features information on 671 adult patients collected from 14 contributing institutions, and includes genomic, diagnostic (imaging), treatment, and outcomes data.
- Conavi Medical announced the first clinical use of its Novasight Hybrid System, which allows simultaneous imaging of coronary arteries using both intravascular ultrasound (IVUS) and Optical Coherence Tomography (OCT) on a single catheter. Conavi claims that simultaneous IVUS and OCT coronary imaging has been “highly anticipated in the interventional cardiology community” due to IVUS/OCT’s expected ability to improve characterization of atherosclerotic tissue and guide coronary interventions.
- Researchers from UCLA found that integrated CT CAD workflow reduce radiologist reading times by between 7% and 44%, compared to current workflow that generally requires radiologists to view studies in CAD applications and clinical workflow systems at separate workstations. The study also suggests that this inherent inefficiency has reduced overall demand for CAD software.
The Resource Wire
– Future home of sponsored content. Maybe yours.
- Agfa Enterprise Imaging integrates clinical videos and images into a single application
- Analogic’s BK Ultrasound bkSpecto urology ultrasound system combines image quality, speed, usability, and small footprint
- Canon Aquilion Precision, the world’s first Ultra-High Resolution CT system
- Caresteam touts benefits of weight bearing CT imaging
- Fujifilm’s FDR AQRO mobile x-ray combines high sensitivity with small size and low-dos
- GE’s Every woman. Every body. Every town. roadshow brings the company’s breast health technologies and message to centers across the country.
- Hitachi white paper on the business benefits of high-field MRI.
- Hologic’s 3Dimension Mammography System leads with 3D imaging and patient comfort.
- IBM Watson Imaging Clinical Review cognitive data review tool for more accurate patient records.
- Konica Minolta publishes latest Kontours issue.
- Philips’ new HealthSuite Insights AI platform curates and analyzes healthcare data and offers tools to build and operate AI-based solutions.
- Purview survey assesses medical image management effectiveness.
- Samsung’s RS85A premium ultrasound combines image quality with advanced ergonomics
- Shimaduz MobileDaRt Evolution MX8 premium mobile x-ray
- Siemens’ new Somatom Go CTs add cardiac and image-guided capabilities