News | Stroke | July 19, 2023

New Geometric Deep Learning Model for Detecting Stroke Lesions

The proposed approach outperforms other neural network architectures by leveraging rich geometric information to segment brain stroke lesion images

The proposed approach outperforms other neural network architectures by leveraging rich geometric information to segment brain stroke lesion images

GFCN-8s, the proposed geometric deep learning network with three pooling layers and eight-fold upsampling, can extract rich geometric information from computed tomography scans of the brain. This enables it to outperform the state-of-the-art models (U-Net, FCN-8s, and PointNet++) in detecting stroke lesions. Iporre-Rivas et al., doi 10.1117/1.JMI.10.4.044502. 


July 19, 2023 — Ischemic stroke, which occurs when a blood vessel in the brain gets blocked by a clot, is among the leading causes of death worldwide. Fortunately, surgeons now have access to advanced imaging techniques that allow them to visualize the interior of a patient’s brain during a stroke. This helps them pinpoint the location of the clot and analyze the extent of damage to the brain tissue. 

Computed tomography-perfusion (CT-P) is one of the most useful imaging modalities in the early stages of an acute stroke. However, it is challenging to accurately identify segmentation—the outline of stroke lesions—in a CT-P scan, and the final diagnosis depends greatly on the surgeon’s expertise and ability. To address this issue, scientists have come up with various machine learning models that perform automatic segmentation of CT-P scans. Unfortunately, none of them has reached a level of performance suitable for clinical applications. 

Against this backdrop, a team of researchers from Germany recently developed a new segmentation algorithm for stroke lesions. As reported in their study published in the Journal of Medical Imaging, the team built a geometric deep learning model called “Graph Fully-Convolutional Network” (GFCN). The internal operations performed by their geometric algorithm differ fundamentally from those of the more widely used Euclidean models. In their study, the researchers explored the benefits and limitations of this alternative approach. 

A key advantage of the proposed model is that it can better learn and preserve important features inherent to brain topology. By using a graph-based neural network, the algorithm can detect complex inter-pixel relationships from different angles. This, in turn, enables it to detect stroke lesions more accurately. 

In addition, the team adopted “pooling” and “unpooling” blocks in their network structure. Put simply, the pooling operations, also called “downsampling,” reduce the overall size of the feature maps extracted by the network from input images. This reduces the computational complexity of the algorithm, enabling the model to extract the most salient features of the CT-P scans. In contrast, the unpooling operations (or “upsampling”) revert the pooling operations to help properly localize the detected features in the original image based on contextual cues. By combining these two operations, the network structure can extract richer geometric information. 

The team conducted a series of analyses to determine the effect of each component of GFCN on its segmentation performance. They then compared the performance of the proposed algorithm against the state-of-the-art models, all trained using the same public dataset. Interestingly, although their model used basic unpooling techniques and a simple input configuration, it performed better than the conventional models under most conditions. 

Notably, GFCN-8s, with three pooling layers and eight-fold upsampling, achieved a Dice coefficient score—a metric indicating the overlap between the predicted and actual lesion areas—of 0.4553, which is significantly higher than other models. Moreover, the proposed model could adapt to irregular segmentation boundaries better than the state-of-the-art models. 

Overall, the findings of this study showcase the potential of geometric deep learning for segmentation problems in medical imaging. Further research on similar strategies could pave the way for highly accurate models for automatic stroke diagnosis that could improve patient outcomes and save lives. 

Read the Gold Open Access article by A. Iporre-Rivas et al., “Stroke-GFCN: ischemic stroke lesion prediction with a fully convolutional graph network,” J. Med. Imag. 10(4) 044502 (2023), doi 10.1117/1.JMI.10.4.044502

For more information: https://spie.org 


Related Content

News | Cardiac Imaging

April 23, 2024 — CDL Nuclear Technologies, a pioneer in advanced diagnostic solutions, is proud to announce the launch ...

Home April 23, 2024
Home
Feature | Cardiac Imaging

Cardiac positron emission tomography (PET) is growing in popularity among cardiologists because it provides the ability ...

Home March 05, 2024
Home
News | Cardiac Imaging

PLEASE NOTE: This webinar has been postponed to a later date. A new date will be posted in the coming days. On March 25 ...

Home February 29, 2024
Home
News | Cardiac Imaging

February 28, 2024 — Royal Philips, a global leader in health technology, announced major enhancements to its Image ...

Home February 28, 2024
Home
News | Cardiac Imaging

February 12, 2024 — According to the American Journal of Roentgenology (AJR), free-breathing cine-deep learning (DL) may ...

Home February 12, 2024
Home
News | Cardiac Imaging

November 16, 2023 — Cardiovascular diseases rank among the top causes of death across the world, and cardiac ...

Home November 16, 2023
Home
News | Cardiac Imaging

September 21, 2023 — Declines in cardiovascular procedure volumes observed early in the COVID-19 pandemic greatly ...

Home September 21, 2023
Home
News | Cardiac Imaging

July 11, 2023 — Hyperfine, Inc., the groundbreaking medical device company that created the Swoop system, the world’s ...

Home July 11, 2023
Home
Feature | Cardiac Imaging | By Matthew Jay Budoff, MD, FACC, FAHA

Coronary computed tomography angiography (CCTA) is cheaper, easier to use and its results are easier to understand than ...

Home July 10, 2023
Home
News | Cardiac Imaging

June 29, 2023 — According to an accepted manuscript published in ARRS’ own American Journal of Roentgenology (AJR) ...

Home June 29, 2023
Home
Subscribe Now