News

BOLD-AIR Summit 2021: Emerging Solutions: Deidentification and Federated Learning

BOLD-AIR Summit 2021: Emerging Solutions: Deidentification and Federated Learning

August 4, 2021

In order for machine learning algorithms to work robustly and produce accurate results, large amounts of diverse data are necessary. The amount of data available for radiological applications at any single center is insufficient and thus pooling of data across centers is critical for robust AI. Traditional approaches for sharing imaging data involves each hospital transferring its curated and harmonized data sets to a central system who then releases the trained model back to the participating hospitals. However this approach is burdened by regulatory and privacy concerns, even with...

Read more about BOLD-AIR Summit 2021: Emerging Solutions: Deidentification and Federated Learning
Intravenous alteplase for stroke with unknown time of onset guided by advanced imaging can improve patient outcomes

Intravenous alteplase for stroke with unknown time of onset guided by advanced imaging can improve patient outcomes

November 8, 2020
A large meta-analysis of patients who have had a stroke with unknown time of onset with a DWI-FLAIR or perfusion mismatch showed that intravenous alteplase resulted in better functional outcome at 90 days than placebo or standard care. A net benefit was observed for all functional outcomes despite an increased risk of symptomatic intracranial haemorrhage. Although there were more deaths with... Read more about Intravenous alteplase for stroke with unknown time of onset guided by advanced imaging can improve patient outcomes
picture

Deep learning used to analyze Big Stroke Datasets

September 30, 2019

We evaluated deep learning algorithms’ segmentation of acute ischemic lesions on heterogeneous multi-center mutli-vendor clinical diffusion-weighted magnetic resonance imaging (MRI) data sets and explored the potential role of this tool for phenotyping acute ischemic stroke. The ensemble consisting of a mixture of data from 12 international stroke genetics research center  and single-center convolutional neural networks performed best, demonstrating the need for diverse training data sets for creating robust machine learning algorithms. 

You can read our...

Read more about Deep learning used to analyze Big Stroke Datasets
S Winzeck

Deep learning model for segmenting acute stroke lesions publication selected as AJNR's editorial choice

June 21, 2019

Accurate automated infarct segmentation is needed for acute ischemic stroke studies relying on infarct volumes as an imaging phenotype or biomarker that require large numbers of subjects. This also has important clinical implications as acute DWI infarct volumes are increasingly being used for triage of stroke patients presenting for late window interventions. A group of researchers from our department conducted a study to investigate whether an ensemble of convolutional neural networks trained on multiparametric DWI maps outperforms single networks trained on solo DWI parametric maps...

Read more about Deep learning model for segmenting acute stroke lesions publication selected as AJNR's editorial choice
mrwitness

MRI can expand treatment options for patients with unwitnessed stroke onset

April 24, 2018

In the MRWITNESS study, led by investigators from the CCNI laboratory, we showed that patients with unwitnessed acute ischemic strokes may be safely treated with a "clot busting" drug, alteplase. Under current guidelines, treatment with intraveous alteplase is approved up to 4.5 hours from when the patient was last known to be well. Approximately 25% of ischemic stroke patients have stroke onsets that are unwitnessed. Our trial used an MR witness to identify patients who might be able to be treated safely with alteplase...

Read more about MRI can expand treatment options for patients with unwitnessed stroke onset