Abstract
Melanoma is a type of skin cancer that often leads to poor prognostic responses and survival rates. Melanoma usually develops in the limbs, including in fingers, palms, and the margins of the nails. When melanoma is detected early, surgical treatment may achieve a higher cure rate. The early diagnosis of melanoma depends on the manual segmentation of suspected lesions. However, manual segmentation can lead to problems, including misclassification and low efficiency. Therefore, it is essential to devise a method for automatic image segmentation that overcomes the aforementioned issues. In this study, an improved algorithm is proposed, termed EfficientUNet++, which is developed from the U-Net model. In EfficientUNet++, the pretrained EfficientNet model is added to the UNet++ model to accelerate segmentation process, leading to more reliable and precise results in skin cancer image segmentation. Two skin lesion datasets were used to compare the performance of the proposed EfficientUNet++ algorithm with other common models. In the PH2 dataset, EfficientUNet++ achieved a better Dice coefficient (93% vs. 76%–91%), Intersection over Union (IoU, 96% vs. 74%–95%), and loss value (30% vs. 44%–32%) compared with other models. In the International Skin Imaging Collaboration dataset, EfficientUNet++ obtained a similar Dice coefficient (96% vs. 94%–96%) but a better IoU (94% vs. 89%–93%) and loss value (11% vs. 13%–11%) than other models. In conclusion, the EfficientUNet++ model efficiently detects skin lesions by improving composite coefficients and structurally expanding the size of the convolution network. Moreover, the use of residual units deepens the network to further improve performance.
Highlights
Melanoma is a type of skin cancer with high spread characteristics and a mortality rate of approximately 75% [1]
Former is an essential measure of the overlap between two samples. is measure ranges from 0 to 1, where a Dice coefficient of 1 means complete overlap, which is represented as equation (1), where N is the size of the pixels, pi is the predicted pixels, and yi is the test pixels
In BCE, if y 0 is much larger than y 1, the y 0 component of the loss function will dominate, making the model heavily biased towards the background, resulting in poor training results. erefore, the loss function in this study uses the combination of BCE and Dice loss; the formula is represented as equation (3), and the parameter α is used to control the weights of the BCE or Dice coefficients
Summary
Melanoma is a type of skin cancer with high spread characteristics and a mortality rate of approximately 75% [1]. Medical image segmentation and computer-aided vision techniques have recently been used to improve the diagnosis of cancer lesions [6,7,8]. Convolutional neural networks (CNNs) are the most commonly used algorithms in medical imaging [9,10,11] and are used for many tasks, including image classification [10, 12, 13], superresolution [14,15,16], object detection [17,18,19], and semantic segmentation [20,21,22]. Deep learning can be used to automatically extract features from images in different categories; this may improve the feature detection time and efficiency of traditional computer-aided detection by 10%
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 call
