Retinal Blood Vessel Segmentation Using Attention Module and Tversky Loss Function

Sau, Paresh Chandra

doi:10.1007/978-981-19-0976-4_41

Cited by 2 publications

(2 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…One such cost function is the Tversky loss, which is an extension of the Dice loss, that includes an additional parameter to control the balance between false positives and false negatives. The Tversky loss has been shown to be effective in medical image segmentation tasks, where the class imbalance between foreground and background pixels can be significant [43][44][45]. Thus, in this study, we used the Tversky cost function, as described in Equation (1):…”

Section: Suggested U-net Modelmentioning

confidence: 99%

See 1 more Smart Citation

Spatial Attention Mechanism and Cascade Feature Extraction in a U-Net Model for Enhancing Breast Tumor Segmentation

Zarbakhsh

2023

Applied Sciences

View full text Add to dashboard Cite

In the field of medical imaging, the accurate segmentation of breast tumors is a critical task for the diagnosis and treatment of breast cancer. To address the challenges posed by fuzzy boundaries, vague tumor shapes, variation in tumor size, and illumination variation, we propose a new approach that combines a U-Net model with a spatial attention mechanism. Our method utilizes a cascade feature extraction technique to enhance the subtle features of breast tumors, thereby improving segmentation accuracy. In addition, our model incorporates a spatial attention mechanism to enable the network to focus on important regions of the image while suppressing irrelevant areas. This combination of techniques leads to significant improvements in segmentation accuracy, particularly in challenging cases where tumors have fuzzy boundaries or vague shapes. We evaluate our suggested technique on the Mini-MIAS dataset and demonstrate state-of-the-art performance, surpassing existing methods in terms of accuracy, sensitivity, and specificity. Specifically, our method achieves an overall accuracy of 91%, a sensitivity of 91%, and a specificity of 93%, demonstrating its effectiveness in accurately identifying breast tumors.

show abstract

Section: Suggested U-net Modelmentioning

confidence: 99%

“…The Tversky loss is a generalization of the Dice loss, where 'alpha = beta = 0.5'. By adjusting the values of 'alpha' and 'beta', one can control the trade-off between false positives and false negatives, and tailor the loss function to one's specific task [43][44][45].…”

Section: Suggested U-net Modelmentioning

confidence: 99%