Protosurv: A multimodal survival prediction framework based on prototype learning and gated hierarchical attention

Puntos clave

• The multimodal framework offers enhanced survival prediction accuracy, integrating diverse data types.
• Key evidence shows improved predictions with prototype learning using various patient data inputs.
• Analysis utilizes machine learning techniques with gated hierarchical attention to prioritize key features.
• This method highlights the importance of accurate survival prediction, potentially guiding clinical decision-making.