ISLES 2016 and 2017-Benchmarking Ischemic Stroke Lesion Outcome Prediction Based on Multispectral MRI
See where this sits in the topic map →Summary AI-generated
- TL;DR
- Comparing newly developed machine learning tools is often difficult because model performance depends heavily on both the algorithm and the dataset used.
- Problem
- Without a common basis for comparison, researchers must implement others' algorithms themselves to establish a benchmark, or a direct comparison of new and old techniques becomes infeasible.
- Method
- The Ischemic Stroke Lesion Segmentation (ISLES) challenge addressed this by providing a uniformly pre-processed multispectral MRI dataset, allowing researchers worldwide to apply their algorithms directly.
- Results
- Nine teams participated in ISLES 2015 and 15 teams participated in ISLES 2016, with their performances evaluated transparently to identify the state-of-the-art; top-ranked teams almost always used convolutional neural networks (CNNs).
- Contributions
- Not specified in the abstract.
- Limitations
- Despite these efforts, ischemic stroke lesion outcome prediction remains challenging.
- Takeaways
- The annotated dataset remains publicly available, and new approaches can be compared directly through an online evaluation system to serve as a continuing benchmark.
- Applications
- Ischemic stroke lesion outcome prediction using multispectral MRI.
- Topics
- Machine Learning; Medical Imaging; Ischemic Stroke Lesion Outcome Prediction; Multispectral MRI; Benchmarking
- For industry
- Healthcare and medical technology.
- Why it matters
- Provides a standardized benchmarking platform and public dataset to improve the comparability of stroke lesion outcome prediction models.
Abstract
Performance of models highly depend not only on the used algorithm but also the data set it was applied to. This makes the comparison of newly developed tools to previously published approaches difficult. Either researchers need to implement others' algorithms first, to establish an adequate benchmark on their data, or a direct comparison of new and old techniques is infeasible. The Ischemic Stroke Lesion Segmentation (ISLES) challenge, which has ran now consecutively for 3 years, aims to address this problem of comparability. ISLES 2016 and 2017 focused on lesion outcome prediction after ischemic stroke: By providing a uniformly pre-processed data set, researchers from all over the world could apply their algorithm directly. A total of nine teams participated in ISLES 2015, and 15 teams participated in ISLES 2016. Their performance was evaluated in a fair and transparent way to identify the state-of-the-art among all submissions. Top ranked teams almost always employed deep learning tools, which were predominately convolutional neural networks (CNNs). Despite the great efforts, lesion outcome prediction persists challenging. The annotated data set remains publicly available and new approaches can be compared directly via the online evaluation system, serving as a continuing benchmark (www.isles-challenge.org).