A discrimination model by machine learning to avoid gastrectomy for early gastric cancer

Aim

Gastrectomy is recommended for patients with early gastric cancer (EGC) because the possibility of lymph node metastasis (LNM) cannot be completely denied. The aim of this study was to develop a discrimination model to select patients who do not require surgery using machine learning.

Methods

Data from 382 patients who received gastrectomy for gastric cancer and who were diagnosed with pT1b were extracted for developing a discrimination model. For the validation of this discrimination model, data from 140 consecutive patients who underwent endoscopic resection followed by gastrectomy, with a diagnosis of pT1b EGC, were extracted. We applied XGBoost to develop a discrimination model for clinical and pathological variables. The performance of the discrimination model was evaluated based on the number of cases classified as true negatives for LNM, with no false negatives for LNM allowed.

Results

Lymph node metastasis was observed in 95 patients (25%) in the development cohort and 11 patients (8%) in the validation cohort. The discrimination model was developed to identify 27 (7%) patients with no indications for additional surgery due to the prediction of an LNM-negative status with no false negatives. In the validation cohort, 13 (9%) patients were identified as having no indications for additional surgery and no patients with LNM were classified into this group.

Conclusion

The discrimination model using XGBoost algorithms could select patients with no risk of LNM from patients with pT1b EGC. This discrimination model was considered promising for clinical decision-making in relation to patients with EGC.

Roles of Lipid Profiles in Human Non-Small Cell Lung Cancer

Aims: This review aims to identify lipid biomarkers of non-small cell lung cancer (NSCLC) in human tissue samples and discuss the roles of lipids in tissue molecular identification, the discovery of potential biomarkers, and surgical margin assessment. Methods: A review of the literature focused on lipid-related research using mass spectrometry (MS) techniques in human NSCLC tissues from January 1, 2015, to November 20, 2020, was conducted. The quality of included studies was assessed using the QUADAS-2 tool. Results: Twelve studies met the inclusion criteria and were included in the review. The risk of bias was unclear in the majority of the studies. The contents of lipids including fatty acids, phosphatidyl choline, phosphatidyl ethanolamine, phosphatidyl inositol, cardiolipin, phosphatidyl serine, phosphatidyl glycerol, ceramide, lysophosphatidylethanolamine, lysophosphatidylcholine, and lysophosphatidylglycerol differed significantly between cancer and healthy tissues. The sensitivity or specificity of the discrimination model was reported in 8 studies, and the sensitivity and specificity varied among the reported methods. The lipid profiles differed between adenocarcinoma and squamous cell carcinoma NSCLC subtypes. Conclusion: In preclinical studies, MS analysis and multiple discrimination models can be combined to distinguish NSCLC tissues from healthy tissues based on lipid profiles, which provides a new opportunity to evaluate the surgical margin and cancer subtype intraoperatively. Future studies should provide guidance for selecting patients and discrimination models to develop an improved method for clinical application.

Rapid Identification of Kudzu Powder of Different Origins Using Laser-Induced Breakdown Spectroscopy

The rapid identification of kudzu powder of different origins is of great significance for studying the authenticity identification of Chinese medicine. The feasibility of rapidly identifying kudzu powder origin was investigated based on laser-induced breakdown spectroscopy (LIBS) technology combined with chemometrics methods. The discriminant models based on the full spectrum include extreme learning machine (ELM), soft independent modeling of class analogy (SIMCA), K-nearest neighbor (KNN) and random forest (RF), and the accuracy of models was more than 99.00%. The prediction results of KNN and RF models were best: the accuracy of calibration and prediction sets of kudzu powder from different producing areas both reached 100%. The characteristic wavelengths were selected using principal component analysis (PCA) loadings. The accuracy of calibration set and the prediction set of discrimination models, based on characteristic wavelengths, is all higher than 98.00%. Random forest and KNN have the same excellent identification results, and the accuracy of calibration and prediction sets of kudzu powder from different producing areas reached 100%. Compared with the full spectrum discriminant analysis model, the discriminant analysis model based on the characteristic wavelength had almost the same discriminant effects, and the input variables were reduced by 99.92%. The results of this research show that the characteristic wavelength can be used instead of the LIBS full spectrum to quickly identify kudzu powder from different producing areas, which had the advantages of reducing input, simplifying the model, increasing the speed and improving the model effect. Therefore, LIBS technology is an effective method for rapid identification of kudzu powder from different habitats. This study provides a basis for LIBS to be applied in the genuineness and authenticity identification of Chinese medicine.