P2Echo is a single model that performs echocardiography segmentation across multiple cardiac views using text prompts. It uses a Dynamic Image-Text Alignment module to modulate decoder features based on the specific requested view and target structure.
HOW THIS AFFECTS YOU
●
builderYou can integrate text-controlled segmentation into medical imaging pipelines using a single unified framework.
●
healthYou can use a single model to segment different parts of the heart just by providing text instructions.