“. . . vendors of medical imaging AI were already battling with convincing payers prior to the pandemic, and there is no reason to suspect that all of a sudden these tools have become any more accurate or useful.”
Hardian Health’s Dr. Hugh Harvey, encouraging health systems to avoid unproven CV19 AI and instead look for solutions based on proven lung imaging AI.
Morning, everyone. Hope you’re all well and staying safe. Because nearly every story is about CV19 in some way, we’re moving away from including a dedicated COVID-19 section in each issue. Take care, hooray for the helpers, and cheers to everyone staying home.
Imaging Wire Sponsors
- Focused Ultrasound Foundation – Accelerating the development and adoption of focused ultrasound.
- GE Healthcare – Providing point of care ultrasound systems, from pocket-sized to portable consoles, designed to support your clinical needs and grow along with your practice.
- Healthcare Administrative Partners – Empowering radiology groups through expert revenue cycle management, clinical analytics, practice support, and specialized coding.
- Nuance – AI and cloud-powered technology solutions to help radiologists stay focused, move quickly, and work smarter.
- Qure.ai – Making healthcare more accessible by applying deep learning to radiology imaging .
- Riverain Technologies – Offering artificial intelligence tools dedicated to the early, efficient detection of lung disease.
The Imaging Wire
CV19 AI Works
There’s been no shortage of criticism directed towards the use of CT scans for COVID-19 and it’s a safe bet that CV19 AI criticism is coming. However, a team of Wuhan-based researchers recently found that a commercial AI solution can indeed use CT scans to support COVID-19 assessments and monitoring.
- The Study – The retrospective study analyzed chest CTs from 126 CV19 patients using a deep learning algorithm, finding significantly different CT lung opacification percentages for mild, moderate, severe and critical cases (all P < 0.01). The algorithm also revealed changes in lung opacification percentages between the baseline and first follow up (3.6% vs. 8.7%), but not between the first follow up and second follow up (8.7% vs. 6%).
- The Takeaway – AI can analyze CT scans to support CV19 assessments and ongoing monitoring, helping improve and streamline COVID-19 imaging interpretation.
Swiss healthcare startup IcosaMed and students from Swiss Federal Institute of Technology Lausanne developed an ultrasound-equipped “SmartBra” that can be used to diagnose early stage breast cancer. Here are some details:
- The Tech – The SmartBra uses piezoelectric sensors to emit ultrasound waves that create echography scans and the system alerts the wearer if it detects an abnormality.
- Applications – The team will first target women with breast cancer for ongoing monitoring, with plans to expand it to women who have a high genetic risk of cancer, and eventually offer it to all women.
- Big Ambitions – IcosaMed has big goals for its new wearable technology, suggesting that it “could lead to a completely revolutionary approach to cancer prevention.” They even think it could be used to prevent cancer by subjecting women to cancer-killing ultrasound waves while they wear the SmartBra. The company also plans to expand the technology to other garments and cancers in the future.
Some physicians may not like this idea, but cancer-detecting bras have been proposed and researched for years, and other companies are entering this space too.
Opening the CV19 Bottleneck
AuntMinnie.com contributor, Michael J. Cannavo (aka the PACSman), detailed how interoperability could eliminate many of the healthcare bottlenecks that COVID-19 is exposing (long result turnaround times, CV19 exposure risks, overwhelmed staff). Cannavo suggests that these challenges could be corrected if we:
- Create a “national clearinghouse” platform with all CV19 patient data and the ability to schedule and analyze lab/imaging tests
- Schedule patients based on their location, using the platform for patient communication
- Once lab/imaging tests are completed, the information is sent back to the platform for analysis
- Radiologists and/or lab specialists perform final diagnosis
- Results are archived in a VNA and provided to patients/physicians
- CV-positive patients are supported via telehealth
- Data from the platform is then used for epidemic tracking and management.
Some might see this as hard to achieve, but it’s going to take some big ideas and big efforts to fight one of the great health emergencies in modern history.
- Fujifilm Gauze Recognition: Fujifilm Japan launched new surgical gauze recognition software for its DR CALNEO AQRO mobile X-ray system. The new software is intended for use following surgical procedures, allowing surgery teams to identify surgical gauze before it’s left in patients.
- Volumes Still Falling: The latest global CT volume data from Aidoc found that CV19-related volume declines started to stabilize between March 23rd and April 1st, falling just 4% after a 39% drop during the previous week (3/16 – 3/22). The U.S. still endured a 14% decline last week (vs. -20% the week before) and CT declines started hitting new regions including Australia (-22% last week, -32% overall) and South America (-23% last week, 46% overall).
- Practices Cut Costs: Facing these major reductions in imaging reads, radiology practices are introducing some big changes in an effort to cut costs and maximize their staff. This AuntMinnie.com forum discussion revealed that some practices are cutting partner salaries by as much as 50%, encouraging unpaid vacation days, furloughing mid-levels and admin staff, and shifting weekend and overnight studies in-house (away from telerads), in addition to taking advantage of the CARES act.
- New MD MRI Protocol: AMRA Medical and Fulcrum Therapeutics created a standardized whole-body MRI protocol to evaluate skeletal muscle composition in patients with facioscapulohumeral dystrophy (FSHD, a slowly progressing muscular dystrophy). The companies positioned the new protocol as an alternative to current neuromuscular disease progression tests (e.g. 6-Minute Walk Test and single-slice MRI), which are prone to variability, suggesting that the protocol might shorten clinical trials.
- A Call for Density Law Research Standards: DenseBreast-Info.org issued a public call for new research standards to ensure that studies evaluating the effectiveness of breast density notification laws are reaching the right conclusions. The group noted that some recent studies (that found the laws to be ineffective) had flawed results because they didn’t distinguish state laws that provide “general or incomplete info” about the recipient’s personal density, the benefits of screening, and the risks that come with dense breasts.
- Incontinence US Acquisition: Swedish hygiene and health product company, Essity, expanded its incontinence product portfolio with its acquisition of Novioscan for SEK 70m ($6.8mm). Although this isn’t technically an imaging story, Novioscan produces an interesting product that uses wearable ultrasound tech to monitor users’ urine levels and lets them know when it’s time to empty their bladder.
- Generally Multispecialists: A new Academic Radiology study revealed that the majority of general radiologists in the U.S. actually practice as multispecialists and nearly all have at least one subspecialty focus area. The study of 12,438 radiologists who do >50% of their work within a single subspecialty (making them “generalists) found that 85% have at least two specialties that each represent >25% of their work (making them also “multispecialists”), only 14.6% have one focus area that represents >25% of their work, and 0.4% had no focus area (all <25%).
- Mo-99 Supply Concerns: CV19 air travel restrictions out of South Africa led to disruptions in the Mo-99 supply chain, prompting imaging groups to warn of potential downstream shortages of the crucial radioisotope. However, other global producers are still in operation and they are currently working on other export options from South Africa (e.g. charter flights).
- NC CT Takes the Lead: Non-contrast CT procedures outnumbered contrast-enhanced CT procedures for the first time in the U.S. last year (52% vs. 48% of 91.4m procedures), following years of non-contrast CT growth (from 22.6m in 2007 to 47.3m in 2019). That’s from an IMV Medical Information Division report published on AuntMinnie.com that attributed the shift to: 1. Improved CT resolution; 2. Reimbursement changes; and 3. Public concerns about contrast use.
- Aussie Center Outbreak: Six staff members at one of I-MED Radiology’s Melbourne-area imaging centers tested positive for COVID-19, prompting the major imaging company to close the location for disinfection. News of this story highlighted the infection risks at medical facilities that are not usually considered to be on the CV19 frontlines.
- brainQuan’s CE Mark: BrainTale announced the CE Marking of its brainQuan software, a SaaS-based solution that uses brain MRI scans to quantitatively measure brain microstructure and predict traumatic brain injury recovery.
- Incidental CV19 Findings: A new study detailed how PET/CT and SPECT/CT scans (taken for non-CV19 oncologic exams) can result in incidental COVID-19 findings, even in asymptomatic patients. The study reviewed nuclear imaging tests from 77 Italian patients (65 PET/CT, 12 SPECT/CT), revealing signs of interstitial pneumonia in six patients who underwent PET/CT (9%) and one patient who underwent SPECT/CT. Five of these seven patients later tested positive for COVID-19.
The Resource Wire
– This is sponsored content.
- The introduction of ultrasound into musculoskeletal care has been a game-changer, revolutionizing the level of precision MSK physicians can bring to patient care. This GE Healthcare profile details how one physician used point of care ultrasound to help improve performance and effectiveness.
- CMS has extended the deadline for reporting 2019 MIPS data and relaxed the criteria for avoiding a penalty in 2021. Learn more in this blog post from Healthcare Administrative Partners.
- ClearRead CT from Riverain Technologies is the first FDA-cleared system for the automatic detection of all lung nodule types, allowing radiologists to reduce search and reporting time and improve nodule detection rates.
- In this quick video, Einstein Healthcare Network’s Peter Wang, MD describes how Nuance’s embedded clinical guidance delivers streamlined radiology workflow and improves report quality.
- Qure.ai’s qXR tool was included in a study that found that deep learning algorithms can help identify TB-associated abnormalities in chest radiographs and are recommended for TB programs with limited resources.
- This Boca Raton Observer column details the recent advances in Parkinson’s research, including the role of focused ultrasound treatment.