Fully convolutional networks (FCNs)-based segmentation method for colorectal tumors on T2-weighted magnetic resonance images

Abstract

Segmentation of colorectal tumors is the basis of preoperative prediction, staging, and therapeutic response evaluation. Due to the blurred boundary between lesions and normal colorectal tissue, it is hard to realize accurate segmentation. Routinely manual or semi-manual segmentation methods are extremely tedious, time-consuming, and highly operator-dependent. In the framework of FCNs, a segmentation method for colorectal tumor was presented. Normalization was applied to reduce the differences among images. Borrowing from transfer learning, VGG-16 was employed to extract features from normalized images. We conducted five side-output blocks from the last convolutional layer of each block of VGG-16 along the network, these side-output blocks can deep dive multiscale features, and produced corresponding predictions. Finally, all of the predictions from side-output blocks were fused to determine the final boundaries of the tumors. A quantitative comparison of 2772 colorectal tumor manual segmentation results from T2-weighted magnetic resonance images shows that the average Dice similarity coefficient, positive predictive value, specificity, sensitivity, Hammoude distance, and Hausdorff distance were 83.56, 82.67, 96.75, 87.85%, 0.2694, and 8.20, respectively. The proposed method is superior to U-net in colorectal tumor segmentation (P < 0.05). There is no difference between cross-entropy loss and Dice-based loss in colorectal tumor segmentation (P > 0.05). The results indicate that the introduction of FCNs contributed to accurate segmentation of colorectal tumors. This method has the potential to replace the present time-consuming and nonreproducible manual segmentation method.

Assessment of MRI-Based Radiomics in Preoperative T Staging of Rectal Cancer: Comparison between Minimum and Maximum Delineation Methods

“…First, VOIs were manually delineated instead of being semiautomatically/automatically segmented, thus making it difficult to avoid subjective errors and making it unsuitable for large-scale data processing [30,31]. Studies had indicated that semiautomated/automated segmentations can provide the reproducible and accurate estimates of the tumor [31][32][33][34]. However, similar to the previous studies, which used manual segmentation in RC patients, these studies described a semiautomated/automated delineated manner along the tumor's outer edge on each consecutive slice, with no precise definition of the border of the whole lesion.…”

Colorectal Tumor Segmentation of CT Scans Based on a Convolutional Neural Network With an Attention Mechanism

“…Several studies have been performed in colorectal tumor segmentation using different deep learning modules, and the results of segmentation accuracy are acceptable [9]- [11]. However, there are some concerns that the false positive rate is high, and the segmentation edge is rough.…”

Full convolutional network based multiple side‐output fusion architecture for the segmentation of rectal tumors in magnetic resonance images: A multi‐vendor study

“…This made the training of this model very slow and expensive. Jian et al proposed an FCN‐based colorectal tumor segmentation method, which provided good results. In their report, they mentioned the problem of class imbalance, but did not further study its impact on segmentation, and the input size of their network was fixed at 96 × 96 pixels.…”

“…However, it is not completely automatic, as manual intervention is needed during segmentation in some cases. [8][9][10][11] Trebeschi et al 8 proposed an automatic segmentation method for rectal tumors based on convolutional neural networks (CNNs). They extracted a fixed-size patch around each voxel and obtained the segmentation results of these voxels using a trained CNN instance.…”