DPFR: Semi-supervised gland segmentation via density perturbation and feature recalibration

Puntos clave

• Improved gland segmentation accuracy with a semi-supervised machine learning approach, indicating its effectiveness in medical imaging tasks.
• The algorithm employs density perturbation and feature recalibration techniques, achieving a notable enhancement in segmentation performance.
• Experimental validation shows significant advancements in accuracy, supporting the potential for better diagnostic tools in pathology.
• This study highlights the need for further testing in diverse datasets to ensure robustness beyond initial validation.