Prediction of the epidermal growth factor receptor gene mutations in lung adenocarcinoma based on CT imaging

The epidermal growth factor receptor (EGFR) is a receptor protein located on the cell membrane that regulates cellular behavior. Due to this characteristic, EGFR can serve as a critical target for cancer diagnosis and treatment. Timely detection of EGFR mutations in patients provides valuable opportunities for disease management and treatment. Given the existing research gap in utilizing EGFR for diagnosing lung cancer, this study explored the application of EGFR in this context, offering valuable therapeutic opportunities for alleviating and treating patient conditions, thereby potentially improving their health to a certain extent. Computer Tomography (CT) scans were employed to detect lesions in patients, and regions of interest were manually defined. Subsequently, feature selection was performed based on the Lasso regression method. Image features with predictive capabilities were chosen, and an EGFR personalized prediction model was established using these features, enabling the prediction of EGFR gene mutations. The parts involving model establishment and computations were implemented on the Linux platform using the C++ programming language. 70 patients enrolled were categorized as 2 groups with 50 patients as training set and 20 patients in the experimental validation set. The results showed that the personalized EGFR gene mutation prediction model displayed an area under the curve (AUC) of 0.894 in the training set and 0.889 in the validation set. In addition, its sensitivity and specificity in the training set were 0.67 and 0.86, respectively, while the values of these two indicators in the validation set were 0.94 and 0.58, respectively. The results indicated the effectiveness of this model in predicting EGFR gene mutations in lung adenocarcinoma patients. This study suggested that CT imaging detection before the deterioration of a patient's condition and predicting disease progression provided a new approach for diagnosing EGFR mutations in lung adenocarcinoma patients. © (2024), (Bio Tech System). All Rights Reserved.