Ischemic Stroke Lesion Segmentation in CT Perfusion Scans using Pyramid Pooling and Focal Loss

TL;DR

This paper introduces a convolutional neural network that uses pyramid pooling and focal loss to improve automatic segmentation of ischemic stroke lesions in CT perfusion scans, aiming to assist faster and more accurate treatment planning.

Contribution

The study presents a novel application of pyramid pooling and focal loss in a CNN for stroke lesion segmentation, outperforming traditional architectures like U-Net and V-Net.

Findings

Achieved top performance in ISLES 2018 challenge

Outperformed U-Net and V-Net architectures

Demonstrated effective lesion segmentation accuracy

Abstract

We present a fully convolutional neural network for segmenting ischemic stroke lesions in CT perfusion images for the ISLES 2018 challenge. Treatment of stroke is time sensitive and current standards for lesion identification require manual segmentation, a time consuming and challenging process. Automatic segmentation methods present the possibility of accurately identifying lesions and improving treatment planning. Our model is based on the PSPNet, a network architecture that makes use of pyramid pooling to provide global and local contextual information. To learn the varying shapes of the lesions, we train our network using focal loss, a loss function designed for the network to focus on learning the more difficult samples. We compare our model to networks trained using the U-Net and V-Net architectures. Our approach demonstrates effective performance in lesion segmentation and ranked among the top performers at the challenge conclusion.