A Novel Hybrid Deep Learning Model with Receptive Field-Enhanced Skip Connections and Adaptive Loss for Medical Image Segmentation

Medical image analysis, crucial for disease diagnosis and treatment, often suffers from the challenge of class imbalance, where the area of normal tissue significantly outweighs that of abnormal regions. Furthermore, the varying class ratios across different images within a dataset complicate the application of uniform loss adjustments. To address these issues and advance automated segmentation, this study proposes a novel deep learning model integrating the strengths of YOLO Version ۸'s efficient feature extraction modules (SPPF and C۲F) within a U-shaped architecture enhanced by a Receptive Field Enhancement (RFE) module. The RFE module, acting as an advanced skip connection, strategically fuses multi-scale features from corresponding and subsequent encoder layers processed through SPPF and C۲F to enrich feature transfer and improve receptive field. To specifically tackle the class imbalance and the diversity of class distributions across images, we introduce a novel Adapt Exponential Loss function. This pixel-level loss dynamically adjusts class weights for each image based on its individual lesion-to-total-pixel ratio (k). We evaluated our proposed model and loss function on challenging skin lesion datasets: ISIC ۲۰۱۸, ISIC ۲۰۱۷, and PH۲. Our method achieved significant segmentation performance with IoU scores of ۸۶.۴۷%, ۸۵.۶۷%, and ۹۳.۱۳%, and Dice scores of ۹۱.۶۳%, ۹۰.۱۹%, and ۹۶.۰۲% on ISIC ۲۰۱۸, ISIC ۲۰۱۷, and PH۲, respectively, demonstrating its effectiveness in accurately delineating skin lesions despite class imbalance and varying lesion proportions. This work contributes a robust framework for medical image segmentation, facilitating more reliable diagnostic tools in dermatology.

کلیدواژه ها:

Medical image segmentation ، Loss function ، Data imbalance ، Skin lesion ، Receptive field enhancement