Learning to Predict Ischemic Stroke Growth on Acute CT Perfusion Data by Interpolating Low-Dimensional Shape Representations

Abstract

Cerebrovascular diseases, in particular ischemic stroke, are one of the leading global causes of death in developed countries. Perfusion CT and/or MRI are ideal imaging modalities for characterizing affected ischemic tissue in the hyper-acute phase. If infarct growth over time could be predicted accurately from functional acute imaging protocols together with advanced machine-learning based image analysis, the expected benefits of treatment options could be better weighted against potential risks. The quality of the outcome prediction by convolutional neural networks (CNNs) is so far limited, which indicates that even highly complex deep learning algorithms are not fully capable of directly learning physiological principles of tissue salvation through weak supervision due to a lack of data (e.g., follow-up segmentation). In this work, we address these current shortcomings by explicitly taking into account clinical expert knowledge in the form of segmentations of the core and its surrounding penumbra in acute CT perfusion images (CTP), that are trained to be represented in a low-dimensional non-linear shape space. Employing a multi-scale CNN (U-Net) together with a convolutional auto-encoder, we predict lesion tissue probabilities for new patients. The predictions are physiologically constrained to a shape embedding that encodes a continuous progression between the core and penumbra extents. The comparison to a simple interpolation in the original voxel space and an unconstrained CNN shows that the use of such a shape space can be advantageous to predict time-dependent growth of stroke lesions on acute perfusion data, yielding a Dice score overlap of 0.46 for predictions from expert segmentations of core and penumbra. Our interpolation method models monotone infarct growth robustly on a linear time scale to automatically predict clinically plausible tissue outcomes that may serve as a basis for more clinical measures such as the expected lesion volume increase and can support the decision making on treatment options and triage.

Artificial intelligence to diagnose ischemic stroke and identify large vessel occlusions: a systematic review

“…The majority of current studies describe AI-mediated detection of core stroke lesion volume on CT and MRI scans, and some use these features to predict acute stroke growth 16–19. For non-contrast CT analysis specifically, automated ASPECTS from ML used in conjunction with clinical presentation accurately associates NIHSS score and can be used to select patients for ET 20 21.…”

The Current Research Landscape of the Application of Artificial Intelligence in Managing Cerebrovascular and Heart Diseases: A Bibliometric and Content Analysis

“…Various techniques such as robotics, machine learning, and natural language processing have been applied to the study of these cardiovascular diseases. Some cutting edge applications of machine learning models include: predicting the presence of a high-risk plaque or an absence of coronary atherosclerosis, using biomarkers in patients with suspected coronary artery disease [4], selecting suitable elderly patients for endovascular therapy to reduce intracerebral hemorrhage after thrombectomy [5], grading of coronary artery stenosis and extent of myocardial ischemia [6,7,8,9,10], as well as stroke lesion outcome prediction [11,12,13,14,15,16,17,18]. Some authors have explored the potential of image-based AI applications in the scoring of non-contrast computerized tomography scans [19,20] as well as machine learning in the prediction of mortality in coronary artery disease and heart failure patients based on echocardiography [21].…”

Utility of CT Perfusion Imaging in Patients With Vertebral Artery Stenosis Treated With Balloon Expandable Stent

“…In clinic, vertebrobasilar artery stenosis was a common disease leading to recurrent PC strokes, the main treatment for this disease was stent implantation. At present, the indications of stenting for vertebral artery stenosis are usually determined by whether the degree of stenosis is more than 50% which is detected by CTA or DSA examination before operation, and corresponding clinical symptoms such as dizziness, walking unsteadily, or the other symptoms ( 10 – 13 ). However, there was no objective assessment of the blood flow in the posterior circulation.…”