1. Research questions
1.1 How do multimodal data present the dynamic trajectory of chronic wound healing?
Integrate multimodal data such as imaging, physiological, biochemical, genomic and lifestyle data to identify key stages and turning points in the process of chronic wound healing.
1.2 Which biomarkers are associated with different stages of chronic wound healing?
In multi-omics data, are there specific gene expression patterns, protein levels or metabolite concentrations associated with specific stages of wound healing?
1.3 Can multimodal data fusion improve the accuracy of chronic wound healing prediction?
Compared with single modality data, can the fusion of multiple data types more accurately predict the process and outcome of wound healing?
1.4 What factors affect the speed and quality of chronic wound healing?
Which lifestyle, environmental or genetic factors are significantly associated with the speed and quality of wound healing?
1.5 How to use multimodal data to establish a predictive model for chronic wound healing?
Based on multimodal data, how to build an effective machine learning model to predict the individual’s wound healing trajectory?
2. Research objectives
2.1 Integrate multimodal data to construct a dynamic trajectory model of chronic wound healing.
Use imaging, physiological, biochemical, genomic and lifestyle data to construct a time series model of the chronic wound healing process and identify key biological events and turning points.
2.2 Identify key biomarkers related to chronic wound healing.
By analyzing multi-omics data, screen out genes, proteins and metabolites closely related to each stage of wound healing, providing a basis for subsequent mechanism research and intervention strategies.
2.3 Develop a wound healing prediction model based on multimodal data.
Use deep learning methods to integrate multiple data types and build a high-precision wound healing prediction model to assist clinical decision-making.
