#376 – The Wire

  • The Medical Imaging 13: CBInsights released its 2021 Digital Health 150 list, which included 13 medical imaging companies, and brought some interesting changes from last year. The analyst firm slashed its field of AI triage/detection companies (from 7 to 3; now Lunit, Qure.ai, Shukun), while going all-in on POCUS startups (from 1 to 5; now DiA Imaging Analysis, Caption Health, Exo, Ultromics, See-Mode). CBInsights also expanded its list of imaging firms focused on disease assessment/management (from 1 to 3; now Cleerly, Elucid, Perspectum), and added its first companies from the “picks and shovels” (Centaur Labs) and non-pixel (Rad AI) sides of radiology AI.
  • Physician Gender Disparities: A new Health Affairs study found that female physicians in the U.S. make $2M less than their male colleagues during a 40-year career (n = 80k, $6.2M vs. $8.3M), even after adjusting for factors like hours worked, specialty, and practice type. Female physicians’ earning disparities were greatest for surgical specialists and nonsurgical specialists (-$2.5M & -$1.6M), while female radiologists made nearly $1M less than their male counterparts. 
  • Subtle + Cortechs.ai Synergies: A new study in the American Journal of Neuroradiology found that Subtle Medical’s SubtleMR image enhancement software allows 60% faster brain MRIs, without affecting the performance of Cortechs.ai’s NeuroQuant quantitative volumetric analysis software. The researchers performed standard and accelerated brain MRIs on 40 patients (enhancing the accelerated MRIs with SubtleMR), finding that NeuroQuant performed the same with the SubtleMR-enhanced accelerated MRIs and standard MRIs. A pair of neurorads also found that the Subtle-enhanced scans’ image quality was superior to the standard MRIs.
  • Medicare Reprieve: American providers breathed a sigh of relief after Congress passed legislation that will soften upcoming 2022 Medicare cuts. Once president Biden signs it into law, the 2022 Medicare conversion factor rate for physicians will increase by 3%, the automatic 2% Medicare payments cuts will be suspended through June 30th (and then reduced to 1%), and the 15% reduction in clinical diagnostic lab tests will be eliminated for 2022.
  • Exablate Prostate Cleared: Insightec announced the FDA clearance of its Exablate Prostate MR-guided Focused Ultrasound prostate tissue treatment system. Exablate Prostate uses Focused Ultrasound to destroy prostate tissue without surgery (using MR for targeting and temp monitoring), reducing tissue damage and potentially allowing patients to delay or avoid radical therapy. Given the achievements of Insightec’ Exablate Neuro system (global regulatory approvals, widespread plan coverage), Exablate Prostate is worth keeping an eye on.
  • The Incidental Recommendation Gap: A new study out of NYU found that adopting a structured recommendation and closed-loop tracking/communication system for incidental lung nodules (ILNs), didn’t sufficiently improve their follow-up rates. The researchers analyzed ILN follow-ups before and after adopting the system (n = 255 & 1,046), revealing that follow-up imaging rates did increase (60% vs. 70%). However, follow-ups were largely associated with non-reporting factors (e.g. race, imaging site, nodule size) and 30% of follow-ups still didn’t happen. The authors suggested that improving patient communication during the discharge process might be more effective.
  • The Fixed XR & Flouro Rebound: After COVID-driven declines in 2020, Signify Research forecasts solid 2021 revenue rebounds for the global fixed digital radiography (from -14% in 2020 to +10% in 2021) and fluoroscopy markets (from -19% to +8%). Fixed X-ray’s 2021 revenue rebound came at the expense of the mobile X-ray market (from +77% to -39%), and quickly returned the fixed segment to pre-pandemic levels. Meanwhile, fluoroscopy spending will remain largely procedure and hospital budget dependent, gradually returning to pre-COVID levels by 2024.
  • Modest AI Generalization: UCSF and UPenn researchers found that adding a “modest” amount of relevant external data to an otherwise all-internal DL training set significantly improves generalizability. The researchers first trained a brain MRI abnormality segmentation model using 293 exams from a single institution (IN1) and then tested it against 51 patients from a second intuition (IN2), finding that performance declined with IN2 data. Then, after evaluating different training data variations (IN1, IN2, and multi-institution data), they found that training the model with 90% IN1 data and 10% IN2 data allowed equal performance when tested with larger IN1 and IN2 datasets.
  • AHN & Penn State’s Reading Partnership: Allegheny Health Network (AHN) and Penn State Health announced that the Penn State radiology department is now providing imaging study reading and interpretation support for AHN (currently 1-3 rads per day). The partnership will allow AHN to absorb its significant imaging volume growth, while supporting the Penn State radiology division’s goal to expand to external health care providers. There might be other academic hospitals with this type of alliance but it’s the first we’ve ever seen, and it’s an interesting alternative to working with telerads or local private practices.
  • AMRA’s Muscle Assessment 510(k): AMRA Medical announced the FDA clearance of its MAsS Scan muscle assessment solution. MAsS Scan utilizes AMRA’s rapid neck-to-knee MRI protocol, analyzing muscle and fat biomarkers, and producing reports containing patients’ Muscle Assessment Scores, body composition measurements (with insights), and segmented images.

You're signed up!

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

-- The Imaging Wire team