Dataset Information

Machine learning predicts lymph node metastasis of poorly differentiated-type intramucosal gastric cancer.

ABSTRACT: To construct a machine learning algorithm model of lymph node metastasis (LNM) in patients with poorly differentiated-type intramucosal gastric cancer. 1169 patients with postoperative gastric cancer were divided into a training group and a test group at a ratio of 7:3. The model for lymph node metastasis was established with python machine learning. The Gbdt algorithm in the machine learning results finds that number of resected nodes, lymphovascular invasion and tumor size are the primary 3 factors that account for the weight of LNM. Effect of the LNM model of PDC gastric cancer patients in the training group: Among the 7 algorithm models, the highest accuracy rate was that of GBDT (0.955); The AUC values for the 7 algorithms were, from high to low, XGB (0.881), RF (0.802), GBDT (0.798), LR (0.778), XGB?+?LR (0.739), RF?+?LR (0.691) and GBDT?+?LR (0.626). Results of the LNM model of PDC gastric cancer patients in test group : Among the 7 algorithmic models, XGB had the highest accuracy rate (0.952); Among the 7 algorithms, the AUC values, from high to low, were GBDT (0.788), RF (0.765), XGB (0.762), LR (0.750), RF?+?LR (0.678), GBDT?+?LR (0.650) and XGB?+?LR (0.619). Single machine learning algorithm can predict LNM in poorly differentiated-type intramucosal gastric cancer, but fusion algorithm can not improve the effect of machine learning in predicting LNM.

SUBMITTER: Zhou CM

PROVIDER: S-EPMC7809018 | biostudies-literature | 2021 Jan

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Machine learning predicts lymph node metastasis of poorly differentiated-type intramucosal gastric cancer.

Zhou Cheng-Mao CM Wang Ying Y Ye Hao-Tian HT Yan Shuping S Ji Muhuo M Liu Panmiao P Yang Jian-Jun JJ

Scientific reports 20210114 1

To construct a machine learning algorithm model of lymph node metastasis (LNM) in patients with poorly differentiated-type intramucosal gastric cancer. 1169 patients with postoperative gastric cancer were divided into a training group and a test group at a ratio of 7:3. The model for lymph node metastasis was established with python machine learning. The Gbdt algorithm in the machine learning results finds that number of resected nodes, lymphovascular invasion and tumor size are the primary 3 fa ...[more]

PMID: 33446730

Similar Datasets

Project description:BackgroundIntramucosal, histologically mixed-type, low-grade (LG), well-differentiated gastric tubular adenocarcinomas (tub1s; LG-tub1s) have larger mean diameters and exhibit a higher frequency of the gastric mucin phenotype (G-phenotype) than pure LG-tub1s. In proportion to their increases in diameter, G-phenotype differentiated-type early gastric cancer (EGC) tumours reportedly grow to eventually contain (an) undifferentiated-type component(s) and LG-tub1s, which are included in differentiated-type EGCs, reportedly exhibit changes in their glandular architectural and cytological atypia grades from LG to high-grade (HG) and can grow to contain a moderately differentiated tubular adenocarcinoma (tub2) component and undifferentiated components. Because they generally show a higher frequency of malignancy relative to tumours with a higher atypia grade and lower differentiation degree, it is suggested that, among mixed-type LG-tub1s, G-phenotype LG-tub1s containing an HG-tub2 component (LG-tub1s?>?HG-tub2) with undifferentiated components might lead to late-onset metastasis to lymph nodes even after a successful endoscopic submucosal dissection (ESD). We aimed to clarify the endoscopic and clinicopathological features of these G-phenotype LG-tub1s?>?HG-tub2.MethodsOf the 13,217 oesophagogastroduodenoscopies performed at our institutions between September 2008 and March 2016, 185 EGC lesions were evaluated in this retrospective observational study. Among these EGC lesions, 60 intramucosal LG-tub1s were divided into 53 tub1 (44 pure LG-tub1s and nine LG-tub1s containing HG-tub1) lesions and seven LG-tub1?>?tub2 (LG-tub1 containing LG- and HG-tub2) lesions.ResultsThe frequencies of the superficial depressed type (P?=?0.026), reddish colour (P?=?0.006), HG of contained tub2s (P?=?0.006), and G-phenotype (P?=?0.028) were significantly higher in the LG-tub1?>?tub2 group than those in the tub1 group. However, the largest lesion of the LG-tub1?>?tub2 group had a superficial flat appearance, an isochromatic colour, an HG-tub2 and an undifferentiated component, and a large diameter greater than 30?mm, and it exhibited a G-phenotype.ConclusionsIntramucosal G-phenotype LG-tub1s?>?HG-tub2 are potential premalignant stomach neoplasms that may have specific endoscopic and clinicopathological features. However, G-phenotype LG-tub1s?>?HG-tub2 with undifferentiated component, which potentially show higher malignancy than those without undifferentiated components might change from a reddish to isochromatic colour. Accurately diagnosing, treating, and following-up G-phenotype LG-tub1s?>?HG-tub2 might decrease the number of patients who experience late-onset metastasis after ESD.

Dataset Information

Machine learning predicts lymph node metastasis of poorly differentiated-type intramucosal gastric cancer.

Publications

Machine learning predicts lymph node metastasis of poorly differentiated-type intramucosal gastric cancer.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets