Article
Author(s):
The StrokeViewer tool is currently in use in Australia and Europe and has led to reduced patient disability in the short-term and more clot removals performed.
Merel Boers, PhD
Nico.Lab’s StrokeViewer LVO, an artificial intelligence (AI) algorithm that detects image characteristics associated with large vessel occlusion (LVO), had its 510k application (FDA K200873) approved by the FDA for the fast triage of patients with stroke.1
The decision was based on results from a clinical study where the performance of the algorithm was retrospectively evaluated in 284 patients from multiple US stroke centers. An assessment of the data, which identified LVOs (ICA, M1, and M2) and compared them to the algorithm, exceeded the performance goal, according to an expert panel.
StrokeViewer LVO is a cloud-based solution that uses AI to support physicians in the emergency stroke setting by sending a notification to medical specialists minutes after a patient with stroke has arrived at the hospital. At that point, physicians are able to use their smartphones, both at an in-hospital and at-home setting, to inspect the computerized tomography (CT) scans images in a web viewer and diagnose stroke.
"With our first FDA clearance we are now able to show US physicians how impactful the combination of human and artificial intelligence is. And yes, more will follow as we are fully committed to unlocking the full healthcare potential for every patient,” Merel Boers, PhD, chief executive officer and co-founder, Nico.Lab, said in a statement.
READ MORE: Cerebral Small Vessel Disease Score Shows No Predictive Value in Stroke Models
StrokeViewer consists of a set of tools to support the entire stroke workflow, but StrokeViewer LVO is the first with FDA approval. The diagnostic tool has a multitude of functions, including the use of advance neural networks to reinforce the prediction confidence of LVO and to minimize false positives. This is done with expert-level sensitivity and specificity even for distal M2 occlusions.2
In order to quantify ischemic core and penumbra in the case of an occluded brain artery, perfusion maps on the DICOM Viewer help physicians make the best treatment decisions leading to better patient outcomes.
StrokeViewer also can detect hemorrhages within seconds of onset by analyzing original head CTs and displaying an overlay for interpretation. It can segment and quantify the most difficult cases and provides physicians with instant volumetric estimation.
With occlusion location, StrokeViewer precisely indicates where the LVO is, thus allowing stroke neurologists to rapidly make well-informed decisions and reduce the risk of missing patients eligible for treatment.
By combining on vessel segmentation with thrombus detection, collateral circulation filling is assessed on the StrokeViewer, making it the most precise evaluation to date, according to Nico.Lab.
The final feature of the StrokeViewer tool is the 3D reconstruction, which enables the interventional specialist to create a 3D image of the brain vasculature. This can assist in preparing the endovascular thrombectomy (EVT) procedure.
"Medical specialists are under enormous pressure to make fast decisions day and night, but it’s not easy. Complicated assessment and interhospital communication sadly make life difficult for physicians motivated to treat stroke victims effectively,” Boers added.
Nico.Lab noted that the tool is currently in use in Australia and Europe, where it has supposedly proven to reduce the time from hospital arrival to start treatment in patients with acute stroke. This has led to a reduced patient disability in the short-term and more clot removal treatments performed.