ESTRO 2024 - Abstract Book

S5062

Physics - Radiomics, functional and biological imaging and outcome prediction

ESTRO 2024

for all stages were as follows: 0.79, 0.04, 0.045, 0.0031, and 0.00085, respectively. In the READ dataset, the p-values for survival analysis from stage I to IV and for all stages were: 0.15, 0.98, 0.18, 0.9, and <0.0001, respectively. In COAD, the single-factor analysis on the training set with 5-fold validation yielded a 95% confidence interval (CI) of 0.6009 0.6547, and the p-value for multi-factor analysis was 0.047, which is less than 0.05. Therefore, the proportions of fibroblasts in each category after classification in a single WSI are independent prognostic features

Conclusion:

This research demonstrates that deep learning and machine learning can provide valuable assistance and meaningful insights into the prognosis prediction of colorectal cancer patients. The results based on this model's identification of fibroblast cells and subsequent classification using physical features can be used to predict patient prognosis.

Keywords: Deeplearning,prognosis predict,colorectal cancer

References:

Fleming, M., Ravula, S., Tatishchev, S.F. & Wang, H.L. Colorectal carcinoma: Pathologic aspects. Journal of gastrointestinal oncology 3, 153 (2012).

Royston, P. & Altman, D.G. External validation of a Cox prognostic model: principles and methods. BMC medical research methodology 13, 1-15 (2013).

Hanley, C.J. et al. Single-cell analysis reveals prognostic fibroblast subpopulations linked to molecular and immunological subtypes of lung cancer. Nature communications 14, 387 (2023)

1614

Digital Poster

MRI V/Q for assessing radiation-induced changes in regional lung function in lung cancer patients

Bilal A Tahir, Paul JC Hughes, Joshua R Astley, Alberto M Biancardi, Helen Marshall, Jim M Wild, Matthew Q Hatton

The University of Sheffield, Division of Clinical Medicine, Sheffield, United Kingdom

Purpose/Objective:

Radiotherapy (RT) is a cornerstone in treating non-small cell lung cancer (NSCLC). Although there have been advancements in RT techniques, such as the introduction of stereotactic ablative RT, survival rates remain poor. A contributing factor to this is radiation-induced lung damage, which is further aggravated by pre-existing diminished pulmonary function prior to RT [1]. Presently, NSCLC patients' lung function prior to RT is evaluated using standard