CCNI | Clinical Computational Neuroimaging Group


Our stroke research focuses on advancing acute ischemic stroke patient care through the development of imaging biomarkers that can be used for triaging patients for stroke trials, for current interventions, for evaluating novel therapies or for understanding mechanistic pathways of stroke. To do this, we use MRI coupled with machine learning algorithms and investigate novel MRI methods that can further improve our understanding.

Patient Selection

Advancing Acute Ischemic Stroke Patient Treatment

We showed that MRI tissue “signatures” can be used to expand the number of patients who can be safely treated with a clot-busting drug (aka thrombolytic). Previous acute ischemic stroke guidelines did not recommend thrombolysis for patients whose stroke symptom onset were unwitnessed or who had strokes on wake-up, approximately 1/4 of all stroke patients.. Our landmark study showed that using our MR WITNESS algorithm, we were able to use the MRI to serve as “witness” when no humans were around. Our study was followed by another study in Europe which showed that a similar approach was beneficial. This led to changes in current acute stroke treatment guidelines. 

Additional Information

      • MR WITNESS: A phase 2a open-label, prospective, multi-center, safety study of intravenous thrombolysis in acute ischemic stroke patients with unwitnessed stroke onset.

      • Late Breaking Oral and Plenary sessions at the 2016 International Stroke Conference

    Evaluating Treatments

    We previously have shown that machine learning algorithms can be used to evaluate therapeutic efficacy of stroke interventions by predicting patient outcome if the patient was not given an intervention. This allows for virtual cross-over studies, potentially improving efficiency of clinical trials. Below are examples for evaluating thrombolysis with alteplase to show that this method can work and another example evaluating normobaric oxygen treatment.

    Intravenous Alteplase

    Normobaric Oxygen

    Machine Learning

    We use machine learning for many applications in our stroke research ranging from segmentation, classification and prediction.


    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 has important clinical implications as acute infarct volumes are increasingly being used for triage of stroke patients presenting for late window interventions. We showed that an ensemble of convolutional neural networks trained on multiparametric MRI maps outperforms single networks trained on solo MRI parametric maps. You can read our paper, “Ensemble of Convolutional Neural Networks Improves Automated Segmentation of Acute Ischemic Lesions Using Multiparametric Diffusion-Weighted MRI,” here:


    We were able to use this algorithm to segment a “big data” set of acute ischemic stroke data to make physiological inferences regarding stroke lesion sizes and stroke subtypes. Ultimately we plan to use this data to study genetic pathways underlying stroke. You can view our original article, “Big Data Approaches to Phenotyping Acute Ischemic Stroke Using Automated Lesion Segmentation of Multi-Center Magnetic Resonance Imaging Data”, here:



    Fundamental advances in stroke care will require pooling imaging phenotype data from multiple centers, to complement the current aggregation of genomic, environmental, and clinical information. Sharing clinically acquired MRI data from multiple hospitals is challenging due to inherent heterogeneity of clinical data, where the same MRI series may be labeled differently depending on vendor and hospital. Furthermore, the de-identification process may remove data describing the MRI series, requiring human review. However, manually annotating the MRI series is not only laborious and slow but prone to human error. We developed deep learning models using transfer learning techniques to automatically classify MRI modalities based purely on imaging features.



    We used machine learning algorithms to combine multiparametric MRI to predict on an individual patient basis how much tissue is likely to die without intervention. This can allow physicians to judge whether the risk of aggressive intervention is worth the potential benefit of rescuing tissue likely to die without treatment. You can read our landmark paper, “Predicting tissue outcome in acute human cerebral ischemia using combined diffusion- and perfusion-weighted MR imaging” published in 2001 in Stroke, here: Since then, we have extended our approach to deeper networks, as shown below.


    MRI Biomarkers

    Diffusion Kurtosis Imaging

    Diffusion Kurtosis Imaging may provide additional insight with regards to microstructural alterations after an acute ischemic stroke.

    Perfusion Weighted Imaging