- Collect and curate a dataset consisting of lung cancer pathological slides and radiological images. This dataset may include histopathological slides of lung cancer tissues, as well as relevant radiological imaging such as CT scans
- Develop and train a deep learning model capable of extracting relevant features from both pathological and radiological images. Using convolutional neural networks (CNN) and other deep learning techniques, the model will learn to classify and predict based on different image features.
- Validate and evaluate the performance of the constructed deep learning model. By utilizing a held-out test set, evaluate and analyze the accuracy and stability of the model's predictions.
- Compare the predictions of the deep learning model with traditional prognosis assessment methods. Evaluate the advantages and effectiveness of the deep learning model in predicting lung cancer patient prognosis by comparing its results with traditional clinical assessment methods.
- Ultimately, establish a predictive model that can be used in clinical practice, providing accurate prognosis predictions to healthcare professionals to assist in clinical decision-making and improve patient treatment and management strategies.

Chen Zhiqiang School of Basic Medical Sciences, Fudan University