Dynamically aggregating MLPs and CNNs for skin lesion segmentation with geometry regularization

BACKGROUND AND OBJECTIVE

Melanoma is a highly malignant skin tumor. Accurate segmentation of skin lesions from dermoscopy images is pivotal for computer-aided diagnosis of melanoma. However, blurred lesion boundaries, variable lesion shapes, and other interference factors pose a challenge in this regard.

METHODS

This work proposes a novel framework called CFF-Net (Cross Feature Fusion Network) for supervised skin lesion segmentation. The encoder of the network includes dual branches, where the CNNs branch aims to extract rich local features while MLPs branch is used to establish both the global-spatial-dependencies and global-channel-dependencies for precise delineation of skin lesions. Besides, a feature-interaction module between two branches is designed for strengthening the feature representation by allowing dynamic exchange of spatial and channel information, so as to retain more spatial details and inhibit irrelevant noise. Moreover, an auxiliary prediction task is introduced to learn the global geometric information, highlighting the boundary of the skin lesion.

RESULTS

Comprehensive experiments using four publicly available skin lesion datasets (i.e., ISIC 2018, ISIC 2017, ISIC 2016, and PH2) indicated that CFF-Net outperformed the state-of-the-art models. In particular, CFF-Net greatly increased the average Jaccard Index score from 79.71% to 81.86% in ISIC 2018, from 78.03% to 80.21% in ISIC 2017, from 82.58% to 85.38% in ISIC 2016, and from 84.18% to 89.71% in PH2 compared with U-Net. Ablation studies demonstrated the effectiveness of each proposed component. Cross-validation experiments in ISIC 2018 and PH2 datasets verified the generalizability of CFF-Net under different skin lesion data distributions. Finally, comparison experiments using three public datasets demonstrated the superior performance of our model.

CONCLUSION

The proposed CFF-Net performed well in four public skin lesion datasets, especially for challenging cases with blurred edges of skin lesions and low contrast between skin lesions and background. CFF-Net can be employed for other segmentation tasks with better prediction and more accurate delineation of boundaries.

Automatic vessel crossing and bifurcation detection based on multi-attention network vessel segmentation and directed graph search

Analysis of the vascular tree is the basic premise to automatically diagnose retinal biomarkers associated with ophthalmic and systemic diseases, among which accurate identification of intersection and bifurcation points is quite challenging but important for disentangling complex vascular network and tracking vessel morphology. In this paper, we present a novel directed graph search-based multi-attentive neural network approach to automatically segment the vascular network and separate intersections and bifurcations from color fundus images. Our approach uses multi-dimensional attention to adaptively integrate local features and their global dependencies while learning to focus on target structures at different scales to generate binary vascular maps. A directed graphical representation of the vascular network is constructed to represent the topology and spatial connectivity of the vascular structures. Using local geometric information including color difference, diameter, and angle, the complex vascular tree is decomposed into multiple sub-trees to finally classify and label vascular feature points. The proposed method has been tested on the DRIVE dataset and the IOSTAR dataset containing 40 images and 30 images, respectively, with 0.863 and 0.764 F¹-score of detection points and average accuracy of 0.914 and 0.854 for classification points. These results demonstrate the superiority of our proposed method outperforming state-of-the-art methods in feature point detection and classification.

Comparison of 2 methods for prediction of liver dosimetric indices in hepatocellular cancer IMRT planning

To evaluate the performance of 2 methods predicting liver dosimetric indices for hepatocellular cancer patients treated with intensity modulated radiation therapy (IMRT). Two predicting methods were implemented to build correlation between geometric and dosimetric information of hepatocellular cancer IMRT plans. One method used Principle Component Analysis method to simplify information and Support Vector Machine (SVM) regression method to build correlation. The other method used a simple linear function to map volumes of certain regions to dosimetic indices. Thirty eight hepatocellular cancer IMRT plans were randomly selected to train the 2 methods. The effectiveness of the 2 methods was validated using another 8 plans. Liver dosimetric indices V₁₀, V₂₀, V₃₀, and mean dose of the 8 plans in validation cohort were calculated using the predicting methods. The predicted indices were compared with the indices derived from the clinically accepted treatment plans. The absolute differences of V₁₀ calculated with the SVM method and the actual values of treatment plans had a mean value of 7.87%. And the values for V₂₀, V₃₀, and mean dose were 4.37%, 4.41%, and 4.20%, respectively. The absolute differences of V₁₀ calculated with the linear formulation method and the actual values of treatment plans had a mean value of 6.09%. And the values for V₂₀, V₃₀, and mean dose were 5.28%, 5.05%, and 5.92%, respectively. These 2 methods have similar accuracy in predicting dosimetric indices V₁₀, V₂₀, V₃₀ of livers. The predicting of V₃₀ and V₂₀ is more accurate than the predicting of V₁₀. The SVM method is more accurate in predicting mean liver dose.