Inter-vendor performance of deep learning in segmenting acute ischemic lesions on diffusion-weighted imaging: a multicenter study

Abstract

There is little evidence on the applicability of deep learning (DL) in the segmentation of acute ischemic lesions on diffusion-weighted imaging (DWI) between magnetic resonance imaging (MRI) scanners of different manufacturers. We retrospectively included DWI data of patients with acute ischemic lesions from six centers. Dataset A (n = 2986) and B (n = 3951) included data from Siemens and GE MRI scanners, respectively. The datasets were split into the training (80%), validation (10%), and internal test (10%) sets, and six neuroradiologists created ground-truth masks. Models A and B were the proposed neural networks trained on datasets A and B. The models subsequently fine-tuned across the datasets using their validation data. Another radiologist performed the segmentation on the test sets for comparisons. The median Dice scores of models A and B were 0.858 and 0.857 for the internal tests, which were non-inferior to the radiologist's performance, but demonstrated lower performance than the radiologist on the external tests. Fine-tuned models A and B achieved median Dice scores of 0.832 and 0.846, which were non-inferior to the radiologist's performance on the external tests. The present work shows that the inter-vendor operability of deep learning for the segmentation of ischemic lesions on DWI might be enhanced via transfer learning; thereby, their clinical applicability and generalizability could be improved.

Figure 1

The patient selection process in the current study.

Figure 2

Residual two-dimensional convolutional long short-term memory (ConvLSTM) U-net. The stack of high b-value diffusion-weighted images and corresponding apparent diffusion coefficient maps are fed into the network per-patient. Two important modifications are made to the 2D U-net. First, residual layers are utilized for each convolutional block, which allows unimpeded propagation of information throughout the network and mitigates the vanishing gradient problem. Second, bi-directional ConvLSTM layers are implemented on top of each convolutional block of the encoder network to allow communication of the feature maps. Consequently, it enables the network to consider all of the slices of an examination before delineating an ischemic lesion’s borders. Therefore, we suggest that this architecture, to some extent, mimics how radiologists assess images, which involves sequential assessment of all slices of an examination before making the final diagnosis or, in this context, performing segmentation.

Figure 3

The flowchart of the deep learning experiments. (A) The deep learning models trained on datasets A and B were referred as models A and B. The segmentation performance of models A and B were first assessed on the internal test sets that consisted of images from the scanners of the same manufacturer. (B) Subsequently, the models’ performances were evaluated on the test partition of the other dataset, which was referred as external test. (C) The validation parts of each dataset were used to fine-tune the pre-trained model using transfer learning. These models were referred as fine-tuned models A and B, respectively. Subsequently, the fine-tuned models’ performances were assessed on the external test sets. A single expert radiologist made segmentations on the test partitions of the datasets for performance comparisons with the deep learning models.

Figure 4

Acute ischemic lesion segmentation of model B on the external test. Figures shows images of a 64-year-old male patient with acute ischemic lesion in the right cerebellar hemisphere. (A) Ground-truth segmentation mask created by the neuroradiologist. (B) The segmentation of native model B. Note the incorrect contours, which is more profoundly marked in the lateral part of the lesion. (C) The segmentation by the radiologist. (D) The segmentation by fine-tuned model B. Note that the fine-tune model demonstrates similar segmentation performance to that of the radiologist, while delineating the borders of the lesion much correctly than the native model.