Dataset Information

Support vector machine-based nomogram predicts postoperative distant metastasis for patients with oesophageal squamous cell carcinoma.

ABSTRACT:

Background

We aim to develop effective models for predicting postoperative distant metastasis for oesophageal squamous cell carcinoma (OSCC) for the purpose of guiding tailored therapy.

Methods

We used data from two centres to establish training (n=319) and validation (n=164) cohorts. All patients underwent curative surgical treatment. The clinicopathological features and 23 immunomarkers detected by immunohistochemistry were involved for variable selection. We constructed eight support vector machine (SVM)-based nomograms (SVM1-SVM4 and SVM1'-SVM4'). The nomogram constructed with the training cohort was tested further with the validation cohort.

Results

The outcome of the SVM1 model in predicting postoperative distant metastasis was as follows: sensitivity, 44.7%; specificity, 90.9%; positive predictive value, 81.0%; negative predictive value, 65.6%; and overall accuracy, 69.5%. The corresponding outcome of the SVM2 model was as follows: 44.7%, 92.1%, 82.9%, 65.9%, and 70.1%, respectively. The corresponding outcome of the SVM3 model was as follows: 55.3%, 93.2%, 87.5%, 70.7%, and 75.6%, respectively. The SVM4 model was the most effective nomogram in prediction, and the corresponding outcome was as follows: 56.6%, 97.7%, 95.6%, 72.3%, and 78.7%, respectively.Similar results were observed in SVM1', SVM2', SVM3', and SVM4', respectively.

Conclusion

The SVM-based models integrating clinicopathological features and molecular markers as variables are helpful in selecting the patients of OSCC with high risk of postoperative distant metastasis.

SUBMITTER: Yang HX

PROVIDER: S-EPMC3778272 | biostudies-literature | 2013 Sep

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Support vector machine-based nomogram predicts postoperative distant metastasis for patients with oesophageal squamous cell carcinoma.

Yang H X HX Feng W W Wei J C JC Zeng T S TS Li Z D ZD Zhang L J LJ Lin P P Luo R Z RZ He J H JH Fu J H JH

British journal of cancer 20130813 5

<h4>Background</h4>We aim to develop effective models for predicting postoperative distant metastasis for oesophageal squamous cell carcinoma (OSCC) for the purpose of guiding tailored therapy.<h4>Methods</h4>We used data from two centres to establish training (n=319) and validation (n=164) cohorts. All patients underwent curative surgical treatment. The clinicopathological features and 23 immunomarkers detected by immunohistochemistry were involved for variable selection. We constructed eight s ...[more]

PMID: 23942069

Similar Datasets

Project description:PurposeThe aim of this study was to investigate if ischemic stroke final infarction volume and location can be used to predict the associated functional outcome using a multi-class support vector machine (SVM).Material and methodsSixty-eight follow-up MR FLAIR datasets of ischemic stroke patients with known modified Rankin Scale (mRS) functional outcome after 30 days were used. The infarct regions were segmented and used to calculate the percentage of lesioned voxels in the predefined MNI, Harvard-Oxford cortical and subcortical atlas regions as well as using four problem-specific VOIs, which were identified from the database using voxel-based lesion symptom mapping. An overall of 12 SVM classification models for predicting the corresponding mRS score were generated using the lesion overlap values from the different brain region definitions, stroke laterality information, and the optional parameters infarct volume, admission NIHSS, and patient age.ResultsLeave-one-out cross validations revealed that including information about the stroke location in terms of lesion overlap measurements led to improved mRS prediction results compared to classification models not utilizing the stroke location information. Furthermore, integration of the optional features led to improved mRS prediction results in all cases tested. The problem-specific brain regions and additional integration of the optional features led to the best mRS predictions with a precise multi-value mRS prediction accuracy of 56%, sliding window multi-value mRS prediction accuracy (mRS±1) of 82%, and binary mRS (0-2 vs. 3-5) prediction accuracy of 85%.ConclusionTherefore, a graded SVM-based functional stroke outcome prediction using the problem-specific brain regions for lesion overlap quantification leads to promising results but needs to be further validated using an independent database to rule out a potential methodical bias and overfitting effects. The prediction of the graded mRS functional outcome could be a valuable tool if combined with voxel-wise tissue outcome predictions based on multi-parametric datasets acquired at the acute phase.

Project description:In this study, gene expression profiles of osteosarcoma (OS) were analyzed to identify critical genes associated with metastasis. Five gene expression datasets were screened and downloaded from Gene Expression Omnibus (GEO). Following assessment by MetaQC, the dataset GSE9508 was excluded for poor quality. Subsequently, differentially expressed genes (DEGs) between metastatic and non-metastatic OS were identified using meta?analysis. A protein-protein interaction (PPI) network was constructed with information from Human Protein Reference Database (HPRD) for the DEGs. Betweenness centrality (BC) was calculated for each node in the network and top featured genes ranked by BC were selected out to construct support vector machine (SVM) classifier using the training set GSE21257, which was then validated using the other three independent datasets. Pathway enrichment analysis was performed for the featured genes using Fisher's exact test. A total of 353 DEGs were identified and a PPI network including 164 nodes and 272 edges was then constructed. The top 64 featured genes ranked by BC were included in the SVM classifier. The SVM classifier exhibited high prediction accuracies in all of the 4 datasets, with accuracies of 100, 100, 92.6 and 100%, respectively. Further analysis of the featured genes revealed that 11 Gene Ontology (GO) biological pathways and 5 Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways were significantly over-represented, including the regulation of cell proliferation, regulation of apoptosis, pathways in cancer, regulation of actin cytoskeleton and the TGF-? signaling pathway. On the whole, an SVM classifier with high prediction accuracy was constructed and validated, in which key genes associated with metastasis in OS were also revealed. These findings may promote the development of genetic diagnostic methods and may enhance our understanding of the molecular mechanisms underlying the metastasis of OS.

Project description:BackgroundThe prognosis of hypopharyngeal squamous cell carcinoma (HPSCC) is poor due to its high incidence of local invasion and distant metastasis (DM). This study aims to explore the DM risk factors of HPSCC and establish a clinical prediction model.MethodsWe downloaded patient data from the Public Surveillance, Epidemiology, and End Results (SEER) database between 2004 and 2018. Univariate and multivariate logistic regression analyses were performed to screen the clinical risk factors for DM of HPSCC. A new nomogram prediction model was then established based on the selected clinical risk factors. We further validated the model's accuracy based on the concordance index (C-index), the area under the receiver operating characteristic (AUC) curve, and the calibration plot. The decision curve analysis (DCA) to test the potential clinical value of the new model was also applied.ResultsA total of 3502 patients were enrolled; the patients with HPSCC were randomly assigned to a training set (n=2463) and a validation set (n=1039). Multivariate Logistic model analysis suggested that sex, T stage, N stage, and the total number of tumors were influence factors for DM of HPSCC. We established and validated a novel nomogram prediction model based on the multivariate logistic model with these influence factors. The C-index was 0.943 and 0.849 in the training and validation sets respectively. The AUC of the training set was 0.705 (95% CI: 0.669-0.741), and the validation set was 0.667 (95% CI: 0.609-0.725). The calibration plot shows that the actual observation value was similar to the predicted value, meaning the model has an excellent discrimination ability. DCA of the nomogram in the training and validation sets suggested that our nomogram has potential application value.ConclusionsWe found that sex, T stage, N stage, and the total number of tumors are independent risk factors for DM of HPSCC. We developed a novel prediction model to predict DM in patients with HPSCC. This nomogram can identify patients with a high risk of DM and has a high clinical application value.

Dataset Information

Support vector machine-based nomogram predicts postoperative distant metastasis for patients with oesophageal squamous cell carcinoma.

Background

Methods

Results

Conclusion

Publications

Support vector machine-based nomogram predicts postoperative distant metastasis for patients with oesophageal squamous cell carcinoma.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets