Evaluation of Enhanced Learning Techniques for Segmenting Ischaemic Stroke Lesions in Brain Magnetic Resonance Perfusion Images Using a Convolutional Neural Network Scheme

Abstract

Magnetic resonance (MR) perfusion imaging non-invasively measures cerebral perfusion, which describes the blood's passage through the brain's vascular network. Therefore, it is widely used to assess cerebral ischaemia. Convolutional Neural Networks (CNN) constitute the state-of-the-art method in automatic pattern recognition and hence, in segmentation tasks. But none of the CNN architectures developed to date have achieved high accuracy when segmenting ischaemic stroke lesions, being the main reasons their heterogeneity in location, shape, size, image intensity and texture, especially in this imaging modality. We use a freely available CNN framework, developed for MR imaging lesion segmentation, as core algorithm to evaluate the impact of enhanced machine learning techniques, namely data augmentation, transfer learning and post-processing, in the segmentation of stroke lesions using the ISLES 2017 dataset, which contains expert annotated diffusion-weighted perfusion and diffusion brain MRI of 43 stroke patients. Of all the techniques evaluated, data augmentation with binary closing achieved the best results, improving the mean Dice score in 17% over the baseline model. Consistent with previous works, better performance was obtained in the presence of large lesions.

Keywords: computer vision; convolutional neural networks; deep learning; deepmedic; ischaemic stroke; medical image analysis; segmentation.

Figure 1

E0 - Segmentation metrics of validation and train subjects during training. The graphs shown are the averages of all 5 folds. The light gray area illustrates ±1 standard deviation. Full segmentation on training cases was not performed by DeepMedic, reason why the lower-right graph is empty.

Figure 2

E0 - Error bars. Each metric for each post-processing step and lesion category is presented. A fourth column, representing all subjects, is also included. Base results are present in all rows for comparison purposes (semi-transparent black). Each marker represents the mean value, and the upper and lower limits represent the 95% confidence interval. The metrics shown are: Dice similarity coefficient (DSC), Matthews correlation coefficient (MCC), True positive rate (TPR), True negative rate (TNR), and Positive predicted value (PPV).

Figure 3

E0 - Volume Bland-Altman analysis. Each lesion category (lacunar/subcortical, small cortical and big cortical) and post-processing step (THT0, THT1, FH, and base) are included. Base is also included in each post-processing step row for comparison purposes (semi-transparent black). Each point represents one subject. The solid line is the mean difference, whereas the dotted-line represents the limits of agreement, computed as mean±1.96 Standard deviation (STD). The x axis is the average volume between the predicted segmentation and the ground truth, whereas the y label is the difference.

Figure 4

DSC error bars of all experiments for the base prediction and FH and each lesion category.

Figure 5

Performance curves of E0-E3. The gray lines indicate the iso-F1S curves, the value of DSC for each point in the graph. The AP metrics are also included.

Figure 6

E1 - Visual segmentation comparison of lacunar/subcortical lesions. The examples include the predicted lesions after each post-processing step. Images are 2D slices, their cut coordinate in the z axis is included, as well as the volume of each segmentation and the DSC achieved.

Figure 7

E1 - Visual segmentation comparison of small cortical lesions. The examples include the predicted lesions after each post-processing step. Images are 2D slices, their cut coordinate in the z axis is included, as well as the volume of each segmentation and the DSC achieved.

Figure 8

E1 - Visual segmentation comparison of big cortical lesions. The examples include the predicted lesions after each post-processing step. Images are 2D slices, their cut coordinate in the z axis is included, as well as the volume of each segmentation and the DSC achieved.