Comparative assessment of the capability of machine learning-based radiomic models for predicting omental metastasis in locally advanced gastric cancer

Clinical featuresAmong the 478 eligible sufferers, 419 sufferers had no omental metastasis, and 59 sufferers had omental metastasis, with an prevalence fee of 14.1%. The coaching cohort consisted of 262 sufferers: 189 males and 79 females; 229 had no omental metastasis, and 33 had omental metastases. The take a look at cohort consisted of 112 people, 83 males, and 29 females; 13 of them had omental metastases, and 99 didn’t. The validation cohort consisted of 104 sufferers, 75 males, and 29 females; 13 sufferers had omental metastases, and the remaining 91 sufferers didn’t. The scientific options of the coaching, take a look at, and validation cohorts confirmed no important statistical variations (Table 1), making certain the reliability of the outcomes obtained from the take a look at and validation cohorts.Table 1 LAGC affected person traits in the coaching, take a look at, and validation cohorts.Radiomic function and scientific function choiceWe extracted 864 radiomic options from the arterial part CT pictures. Detailed radiomic options are proven in Additional File 1. Features with an intra-class correlation coefficient (ICC) larger than 0.75 have been thought of steady options, and 548 radiomic options have been in the end chosen. The ICC values for radiomic options are detailed in Additional File 2. The radiomic options of the coaching cohort have been analyzed utilizing LASSO regression. As the lambda worth elevated, the absolute worth of the function coefficients regularly decreased and finally approached 0 (Fig. 2A). As the lambda worth elevated, the bias share first regularly decreased after which regularly elevated. The optimum lambda worth corresponds to the minimal bias share (Fig. 2B). The chosen radiomic options based mostly on the optimum lambda worth have been diagnostics picture authentic imply (DIOM), authentic form most 2D diameter slice (OSMDS), authentic form most 3D diameter (OSMD), authentic first order kurtosis (OFK), wavelet LH first order kurtosis (WLFK), and wavelet HLH Gldm massive dependence excessive grey stage emphasis (WHGL). Similarly, the scientific options of the coaching cohort have been subjected to LASSO regression to pick distinct scientific options. The scientific options chosen based mostly on the optimum lambda worth have been CA125 and scientific N staging (Fig. 2C, D).Figure 2Screening of radiomic function and scientific function. (A) Relationship between the lambda values and radiomic function coefficients; (B) relationship between the lambda values and bias share of radiomic options; (C) relationship between the lambda values and scientific function coefficients; (D) relationship between the lambda values and bias share of scientific options.SVM-based predictive mannequinIn the omental metastasis prediction mannequin constructed utilizing SVM, we decided that the mannequin had the highest accuracy when the vector quantity was 7 (Fig. 3A). The multidimensional information of LAGC sufferers have been transformed into two-dimensional information, and omental metastasis and non-omental metastasis teams had a extra distinct concentrated distribution in the two-dimensional area (Fig. 3B). In the coaching cohort, the AUC worth of the predictive mannequin was 0.844; sensitivity and specificity have been 0.849 and 0.704, respectively; PPV and NPV have been 0.292 and 0.790, respectively (Fig. 3C). In the take a look at cohort, the AUC of the predictive mannequin was 0.735; sensitivity and specificity have been 0.769 and 0.622, respectively; PPV and NPV have been 0.212 and 0.953, respectively (Fig. 3D). In the validation cohort, the AUC of the predictive mannequin was 0.741; sensitivity and specificity have been 0.833 and 0.756, respectively; PPV and NPV have been 0.385 and 0.974, respectively (Fig. 3E).Figure 3SVM-based capability assessment of the predictive mannequin. (A) Number of vectors and mannequin accuracy of the SVM; (B) distribution of omental metastasis and non-omental metastasis sufferers in the two-dimensional area; (C) the ROC curve of the coaching cohort; (D) the ROC curve of take a look at cohort; (E) the ROC curve of the validation cohort.DT-based predictive mannequinIn DT, we decided the quantity of tree cut up nodes to be 3 (Fig. 4A). The variable options of the constructed prediction mannequin have been ranked in accordance with their significance, with the high six options being scientific N staging, CA125, DIOM, WLFK, OFK, and OSMD, respectively (Fig. 4B). Based on the DT cut up node quantity of 3, we chosen the high three essential options to assemble the DT prediction mannequin (Fig. 4C). In the coaching cohort, the AUC of the predictive mannequin was 0.759; sensitivity and specificity have been 0.606 and 0.883, respectively; PPV and NPV have been 0.426 and 0.940, respectively (Fig. 4D). In the take a look at cohort, the AUC of the predictive mannequin was 0.624; sensitivity and specificity have been 0.231 and 0.990, respectively; PPV and NPV have been 0.227 and 0.910, respectively (Fig. 4E). In the validation cohort, the AUC of the predictive mannequin was 0.658; sensitivity and specificity have been 0.400 and 0.917, respectively; PPV and NPV have been 0.400 and 0.917, respectively (Fig. 4F).Figure 4DT-based capability assessment of the predictive mannequin. (A) Relationship between the quantity of splitting factors of DT and complexity parameters; (B) the significance of scientific options and radiomic options for DT-based predictive mannequin; (C) determination tree diagram; (D) the ROC curve of the coaching cohort; (E) the ROC curve of the take a look at cohort; (F) the ROC curve of the validation cohort.RF-based predictive mannequinIn the RF predictive mannequin, when the quantity of timber in the mannequin is 5, the out-of-bag (OOB) error reaches the minimal worth of 0.122 (Fig. 5A). We then decided the tree cut up nodes, and when the quantity of cut up nodes is 6, the error reaches the minimal worth of 0.318 (Fig. 5B). Additionally, we carried out function significance evaluation, and amongst the eight chosen options, OSMD, OSMDS, and N staging performed essential roles in prediction accuracy. N staging, OSMDS, and OFK performed essential roles in decreasing the Gini coefficient of the predictive mannequin (Fig. 5C). In the coaching cohort, the AUC worth of the predictive mannequin was 0.995, sensitivity and specificity have been 0.970 and 0.965, respectively; PPV and NPV have been 0.800 and 0.995, respectively (Fig. 5D). In the take a look at cohort, the AUC worth of the predictive mannequin was 0.750, sensitivity and specificity have been 0.769 and 0.663, respectively; PPV and NPV have been 0.233 and 0.956, respectively (Fig. 5E). In the validation cohort, the AUC worth of the predictive mannequin was 0.808, sensitivity and specificity have been 0.750 and 0.800, respectively; PPV and NPV have been 0.308 and 0.964, respectively (Fig. 5F).Figure 5RF-based capability assessment of the predictive mannequin. (A) Relationship between the quantity of timber and OOB error in RF, “0” represents the omental metastasis group, and “1” represents the non-omental metastases group; (B) relationship between the quantity of cut up factors and error in the tree; (C) significance of radiomic and scientific options in enhancing the mannequin accuracy and decreasing the Gini coefficient; (D) the ROC curve of the coaching cohort; (E) the ROC curve of the take a look at cohort; (F) the ROC curve of the validation cohort.KNN-based predictive modelThrough the hyperparameter optimization of the KNN operate, we discovered that the finest kernel operate for the predictive mannequin was “triangular,” and the finest k-value was 14 (Fig. 6A). We constructed the predictive mannequin based mostly on these circumstances. In the coaching cohort, the AUC worth of the predictive mannequin was 0.759, sensitivity and specificity have been 0.714 and 0.827, respectively; PPV and NPV have been 0.370 and 0.952, respectively (Fig. 6C). In the take a look at cohort, the AUC worth of the predictive mannequin was 0.797, sensitivity and specificity have been 0.909 and 0.598, respectively; PPV and NPV have been 0.227 and 0.980, respectively (Fig. 6D). In the validation cohort, the AUC worth of the predictive mannequin was 0.611, sensitivity and specificity have been 0.500 and 0.852, respectively; PPV and NPV have been 0.333 and 0.920, respectively (Fig. 6E).Figure 6KNN and LR-based capability assessment of the predictive models. (A) Screening for the finest kernel operate and k-value; (B) finest predictive mannequin and every feature’s AIC worth; (C) the ROC curve of the coaching cohort in the LR-based predictive mannequin; (D) the ROC curve of the take a look at cohort in the LR-based predictive mannequin; (E) the ROC curve of the validation cohort in the LR-based predictive mannequin; (F) the ROC curve of the coaching cohort in the KNN-based predictive mannequin; (G) the ROC curve of the take a look at cohort in the KNN-based predictive mannequin; (H) the ROC curve of the validation cohort in the KNN-based predictive mannequin.LR-based predictive modelThrough LR evaluation, we discovered that when the predictive mannequin consists of N staging, CA125, DIOM, and WHGL, the regression mannequin has an optimum becoming state, with an AIC worth of 155.24. When the predictive mannequin removes these options or provides different options, the AIC worth will improve (Fig. 6B). In the coaching cohort, the AUC worth of the regression predictive mannequin was 0.864, sensitivity and specificity have been 0.879 and 0.774, respectively; PPV and NPV have been 0.358 and 0.978, respectively (Fig. 6F). In the take a look at cohort, the AUC worth of the predictive mannequin was 0.808, sensitivity and specificity have been 0.615 and 0.888, respectively; PPV and NPV have been 0.210 and 0.941, respectively (Fig. 6G). In the validation cohort, the AUC worth of the predictive mannequin was 0.782, sensitivity and specificity have been 0.692 and 0.846, respectively; PPV and NPV have been 0.250 and 0.953, respectively (Fig. 6H).Comparison of predictive talents of numerous modelsThe predictive talents of omental metastasis models in LAGC constructed utilizing numerous machine studying strategies are proven in Table 2. In the coaching cohort (Fig. 7A), the RF predictive mannequin had higher accuracy, AUC, sensitivity, specificity, PPV, and NPV in comparison with LR, SVM, DT, and KNN. The RF predictive mannequin achieved a big enchancment in PPV in comparison with the different 4 predictive models. The DT predictive mannequin had a decrease sensitivity in comparison with the different 4 predictive models, with a sensitivity of solely 0.606. In the take a look at cohort (Fig. 7B), all 5 machine learning-constructed predictive models had a comparatively low PPV. The analysis indicators of the DT predictive mannequin have been extra considerably totally different in comparison with the different 4 predictive models, with a sensitivity of 0.231 and specificity of 0.990 for the DT predictive mannequin. The LR-constructed predictive mannequin had the lowest PPV of 0.210 in comparison with the different 4 predictive models. In the exterior validation cohort (Fig. 7C), the analysis indicator outcomes of the predictive models have been usually just like these in the take a look at cohort.Table 2 A comparability of omental metastatic models’ prediction capacity.Figure 7Line graphs depicting the predictive energy of machine learning-based models. (A) The coaching cohort; (B) the take a look at cohort; (C) the validation cohort.

https://www.nature.com/articles/s41598-024-66979-x