This study describes a deep learning algorithm model that facilitates accurate prediction of grading and staging of renal clear cell carcinoma and also promotes the optimization of individualized treatment plans.
Deep learning algorithms were used to develop a model that allows the prediction of the grading and staging of renal clear cell carcinoma to facilitate the tailoring of individual treatment plans. The study employed deep learning algorithms to predict the cancer’s pathological grading and staging using preoperative clinical variables in 878 renal clear cell carcinoma patients. The study concluded that the proposed algorithm model was accurate. The study findings are published in the Journal of International Medical Research.
Performance of Deep Learning Algorithm Models
The preoperative variables utilized in the deep learning algorithm model included symptoms, chronic disease, sex, age, preoperative blood cell count, tumor information, and body mass index. The study indicated no significant differences between the verification and test sets based on the patient characteristics. The performance of the deep learning algorithm models (BiLSTM, CNN-BiLSTM, and CNN-BiGRU) was evaluated using the F1-score, accuracy, and precision. Both the T-stage and G-grade prediction models had good stability.
CNN-BiGRU Model for Staging and Grading of Tumor
According to the results of the current study, the area under the receiver operating characteristic curve (AUC) was 0.948 and 0.77 for tumor staging (T) and grading (G), respectively. This indicated a high degree of accuracy of this model for the preoperative grading and staging of renal cancer. Furthermore, the T-stage prediction model was associated with better prediction results than the G-grade prediction model, based on the CNN-BiGRU algorithm.
Limitations in the Application of Deep Learning Algorithm Model
The current study included data from a single center. In addition to limited data, the application of the deep learning algorithm model was also limited. For future research to improve the stability of the model, multicenter collaboration is needed in order to increase sample size and perform further clinical and external validation.
The current study demonstrated the role of deep learning algorithm models in predicting preoperative grading and staging of renal clear cell carcinoma.
Source
Wen-Zhi, G., Tai, T., Zhixin, F., Huanyu, L., Yanqing, G., Yuexian, G., & Xuesong, L. (2022). Prediction of pathological staging and grading of renal clear cell carcinoma based on deep learning algorithms. J Int Med Res, 50(11), 3000605221135163. https://doi.org/10.1177/03000605221135163
