Abstract
In recent years, several deep learning networks are proposed to segment 2D or 3D bio-medical images. However, in liver and lesion segmentation, the proportion of interested tissues and lesions are tiny when contrasting to the image background. That is, the objects to be segmented are highly imbalanced in terms of the frequency of occurrences. This makes existing deep learning networks prone to predict pixels of livers and lesions as background. To address this imbalance issue, several loss functions are proposed. Since no researches are having made a comparison among those proposed loss functions, we are curious about that which loss function is the best among them? At the same time, we also want to investigate whether the combination of several different loss functions is effective for liver and lesion segmentation. Firstly, we propose a novel deep learning network (cascade U-ResNets) to produce liver and lesion segmentation simultaneously. Then, we investigate the performance of 5 selected loss functions, WCE (Weighted Cross Entropy), DL (Dice Loss), WDL (Weighted Dice Loss), TL (Teverskry Loss), WTL (Weighted Teversky Loss), with our cascade U-ResNets. We further assemble all cascade U-ResNets trained with different loss functions together to segment livers and lesions jointly on the liver CT (Computed Tomography) volume. Experimental results on the LiTS dataset 1 showed our ensemble model can achieve much better results than every individual model for liver segmentation. 1 https://competitions.codalab.org/competitions/17094#.
Highlights
Liver and lesion segmentation are to delineate a liver and its lesions in medical images
We can see that all similarity-based loss functions (i.e., Dice Loss (DL) and Tversky Loss (TL)) perform much better than WCE, even though the WCE loss takes the data imbalance into consideration
Between the two similarity-based loss functions (i.e., DL and TL), TL performs slightly better than DL
Summary
Liver and lesion segmentation are to delineate a liver and its lesions in medical images (such as CT, MRI, PET images). In computer-aided detections and diagnoses, precise automatic segmentation of the liver is meaningful, but manually delineating liver outline for millions of medical image slices is time-consuming. Automatic liver segmentation is one of the most difficult tasks in computer vision because of the diverse shapes of livers and low contrast with nearby tissues. Applications of deep learning on medical image analysis are soaring. Deep convolutional neural networks can learn high-level features automatically and give reasonable output. Different from common image classification, liver and lesion segmentation is a pixel-level classification task, where a classification model needs to assign a label to each pixel and output the same size mask
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
