Vision transformer supportedkolmogorov-arnold networks for survivalprediction in lung cancer patients

Lung cancer is one of the most common and deadly forms of cancer worldwide, influenced by environmental factors, lifestyle choices, and genetics. It is crucial to find new ways to predict survival outcomes for lung cancer patients in cancer research. The use of pathological whole slide images and omic data with deep learning techniques represents a significant advancement in this field. These methods improve survival estimation and provide more personalized treatments. However, these techniques often struggle to capture the complexity of histological and tabular features, leading to poor performance in many survival prediction applications. In this project, we aimed to enhance survival predictions for lung cancer patients by employing the latest deep-learning methods and novel approaches. By using publicly accessible datasets, leveraging clinical data along with omic data, advanced vision models, and a novel Kolmogorov-Arnold network for processing tabular data, we have showcased the superior capabilities of our proposed method in estimating the survival of lung cancer patients.