A trio of firms have launched new advances in synthetic intelligence designed to enhance affected person outcomes and scale back clinician burnout.
RapidAI introduced a brand new software that helps frontline clinicians make choices when minutes rely in acute stroke care, the place “time is mind,” with out ready for added imaging.
In the meantime, Clew Medical is unveiling new FDA-cleared machine studying algorithms that not solely scale back affected person situation alerts however have been proven to extend accuracy in detecting scientific deterioration.
Lastly, a brand new dataset on weight problems from Dandelion based mostly on longitudinal affected person information might unlock insights on efficacy and unintended effects quicker than trial-and-error prescribing of a category of well-liked diabetes and weight reduction medication, the corporate stated.
Stroke AI imaging
The U.S. Meals and Drug Administration has cleared San Mateo, California-based RapidAI’s AngioFlow module, the corporate stated.
The module delivers perfusion imaging evaluation instantly inside its interventional platform to boost scientific confidence, streamline imaging workflows and doubtlessly, enhance affected person outcomes.
Time is a stroke affected person variable at each small or rural hospitals in addition to giant well being programs the place transfers can run up a stroke affected person’s clock, threatening life and future high quality of life. Actual-time AI evaluation on the level of care can present caregivers with speedy data that may make an enormous distinction in what occurs to sufferers after a stroke.
For care groups working within the Fast AI interventional suite, they’ll now generate qualitative perfusion maps inside minutes to assist assess ischemic change in mind areas with diminished cerebral blood circulation, saving time and decreasing redundant imaging, the corporate stated.
“By avoiding pointless scans, stroke sufferers can obtain the well timed care that may be the distinction between having the ability to stroll out of a hospital to their houses versus being discharged to a talented nursing facility,” stated Dr. Abhishek Singh of the Creighton College College of Drugs in Omaha, Nebraska in an announcement from RapidAI.
“We will now help stroke AI imaging alongside all the affected person pathway, from the preliminary non-contrast CT scan all the way in which to the interventional suite,” added Karim Karti, RapidAI CEO.
Affected person deterioration AI
Boston-based Clew Medical, a scientific surveillance platform, introduced that its second-generation, AI/ML fashions predicting affected person deterioration have obtained 510(okay) clearance from the FDA.
“AI and machine-learning know-how on this house should endure the identical degree of scrutiny and diligence in design, improvement, testing and validation as different medical gadgets utilized by clinicians,” Paul Roscoe, Clew’s CEO, stated in an announcement Monday.
Clearance included the FDA’s approval of Clew’s pre-authorized change management plan, which can enable sure future adjustments to the system enter knowledge set with no need to file for a brand new clearance, the corporate famous.
The Clew platform, which obtained preliminary clearance as a medical machine in 2021, provides well being programs an early identification of a affected person’s threat of degradation that’s 5 occasions extra correct than alerts, based on a research by UMass Memorial Medical Middle and WakeMed Well being & Hospitals revealed in November in CHEST.
The place false alerts are infamous for overburdening hospital workers, the researchers needed to know, “Do ML alerts, telemedicine system-generated alerts or biomedical screens have superior efficiency for predicting episodes of intubation or administration of vasopressors?”
The algorithms have been educated to foretell intubation and vasopressor initiation occasions amongst critically ailing adults, and the research evaluating them to conventional alerts discovered that on common, 98% of bedside monitoring alarms have been false positives, based on Clew Medical.
“ML-derived notifications for clinically actioned hemodynamic instability and respiratory failure occasions characterize an advance as a result of the magnitude of the variations of accuracy, precision, misclassification price and pre-event lead time is giant sufficient to permit extra proactive care and has markedly decrease frequency and interruption of bedside doctor workflows,” the researchers stated within the revealed summary.
Weight problems knowledge trove
Leveraging its consortium of non-academic medical middle well being system companions, Dandelion Well being has launched a brand new library of all scientific knowledge – structured and unstructured knowledge that’s refreshed quarterly – for a category of medicines used to deal with sort 2 diabetes and weight problems.
In keeping with the New York-based startup, the dataset accommodates the total longitudinal affected person information for tens of millions of sufferers – hundreds of which have taken Glucagon-like peptide-1 agonists and 200,000 of whom are on prescribed GLP-1 agonists – and never simply from digital well being information.
The library consists of picture and waveform knowledge island content material in scientific notes, providing unprecedented perception into affected person journeys and the affect of GLP-1s, the corporate stated in its announcement Tuesday.
“By making the wealthy, multi-dimensional knowledge supplied by unstructured modalities accessible and accessible – and linked to real-world remedy patterns and outcomes – individuals who use the library can reply key questions concerning the essential position that GLP-1 based mostly therapies will play in scientific care, stated Shivaani Prakash, Dandelion’s head of knowledge.
The info might allow medical researchers to grasp how GLPs have an effect on every part from the bodily construction of the guts to particular side-effect profiles, Dandelion stated, noting that the corporate is engaged on proofs of idea. One educational medical middle analysis companion developed an algorithm that segments stomach CTs to quantify fats loss and muscle and bone preservation.
Utilizing the dataset researchers can consider the standard of weight reduction by way of biomarkers present in physique scans, evaluate the efficacy of therapies, reveal secondary advantages, quantify unintended effects related to taking GLP-1s and develop precision medication instruments that match sufferers to more practical remedy plans.
“Our GLP-1 dataset will assist cardiometabolic illness enter its precision medication period,” Elliott Inexperienced, co-founder and CEO of Dandelion Well being, stated in an announcement.
Weight problems care lags far behind that of immunology, oncology and different specialties, he famous.
“What bought these markets to the place they’re right this moment was knowledge – knowledge that exposed underlying mechanisms of illness, how people’ illnesses look completely different and consequently, how they may reply to remedy otherwise.”
Andrea Fox is senior editor of Healthcare IT Information.
E mail: [email protected]
Healthcare IT Information is a HIMSS Media publication.