CT and Guidelines based Human-machine Fusion Model for predicting resectability of the Pancreatic Cancer (Preprint)
Autor: | Yi Gao, adilijiang yimamu, Jun Li, Haojie Zhang, Lidu Liang, Feng Lei, Yi Wang, Chenjie Zhou, Shulong Li |
---|---|
Rok vydání: | 2023 |
DOI: | 10.2196/preprints.48884 |
Popis: | BACKGROUND Accurate preoperative prediction of the Pancreatic Cancer resectability is an important work. OBJECTIVE To develop a hybrid Machine Learning model for predicting resectability of the Pancreatic Cancer which based on CT and NCCN Guidelines. METHODS We retrospectively studied 349 patients. 171 cases from Center 1 and 92 cases from Center 2 were used as the primary training cohort, 66 cases from Center 3 and 20 cases from Center 4 were used as the independent test dataset. Semi-automatic module of ITK-SNAP software was used to assist CT image segmentation to obtain 2D layer by layer tumor ROI, which was then stacked into 3D ROI. 788 3D artificial features were extracted using Pyradiomics, through T-test, LASSO, SVM-SFS filtered the features, then connected to the SVM to construct the cRad model. 3D ROI was used to unify the resolution by 3D spline interpolation method, and the 3D tumor image tensor was constructed (interpolated 3D ROI was placed in the center, and zero was filled around). Using 3D tumor image tensor as input, KSTM and ResNet models were constructed. cRad, KSTM and ResNet are fused into a Multi-classifier Fusion ML model based on CFS-ER. Two experts with more than 10 years of clinical experience were invited to reevaluate each patient based on their CECT following the NCCN Guidelines to obtain resectable, unresectable, and border line resectable diagnoses. The three results were converted into probability values of 0.25 ,0.75 and 0.50 according to the traditional empirical method. Then it is used as an independent classifier and integrated with Multi-classifier Fusion ML model to obtain a Human-machine Fusion ML(HMfML) model. RESULTS Multi-classifier Fusion ML model`s AUC:0.8610, ACC:80.23% SEN:78.95%, SPE:80.60%, which is better than cRad, KSTM, and ResNet based single classifier models and their two classifier fusion models. Meaning that three different models have mined complementary CECT feature expression from different perspectives, and can be integrated through CFS-ER, so that the fusion model has better performance. HMfML`s AUC:0.8845, ACC:82.56%, SEN:84.21%, SPE:82.09%. Meaning that ML models might learning extra information from CECT that experts cannot distinguish, thus complementing expert experience and improving the performance of hybrid ML models. CONCLUSIONS Our HMfML can predict PC resectability with high accuracy. CLINICALTRIAL not applicable |
Databáze: | OpenAIRE |
Externí odkaz: |
načítá se...