Dataset Information

A decision tree model for predicting mediastinal lymph node metastasis in non-small cell lung cancer with F-18 FDG PET/CT.

ABSTRACT: We aimed to develop a decision tree model to improve diagnostic performance of positron emission tomography/computed tomography (PET/CT) to detect metastatic lymph nodes (LN) in non-small cell lung cancer (NSCLC). 115 patients with NSCLC were included in this study. The training dataset included 66 patients. A decision tree model was developed with 9 variables, and validated with 49 patients: short and long diameters of LNs, ratio of short and long diameters, maximum standardized uptake value (SUVmax) of LN, mean hounsfield unit, ratio of LN SUVmax and ascending aorta SUVmax (LN/AA), and ratio of LN SUVmax and superior vena cava SUVmax. A total of 301 LNs of 115 patients were evaluated in this study. Nodular calcification was applied as the initial imaging parameter, and LN SUVmax (?3.95) was assessed as the second. LN/AA (?2.92) was required to high LN SUVmax. Sensitivity was 50% for training dataset, and 40% for validation dataset. However, specificity was 99.28% for training dataset, and 96.23% for validation dataset. In conclusion, we have developed a new decision tree model for interpreting mediastinal LNs. All LNs with nodular calcification were benign, and LNs with high LN SUVmax and high LN/AA were metastatic Further studies are needed to incorporate subjective parameters and pathologic evaluations into a decision tree model to improve the test performance of PET/CT.

SUBMITTER: Pak K

PROVIDER: S-EPMC5828356 | biostudies-literature | 2018

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

A decision tree model for predicting mediastinal lymph node metastasis in non-small cell lung cancer with F-18 FDG PET/CT.

Pak Kyoungjune K Kim Keunyoung K Kim Mi-Hyun MH Eom Jung Seop JS Lee Min Ki MK Cho Jeong Su JS Kim Yun Seong YS Kim Bum Soo BS Kim Seong Jang SJ Kim In Joo IJ

PloS one 20180227 2

We aimed to develop a decision tree model to improve diagnostic performance of positron emission tomography/computed tomography (PET/CT) to detect metastatic lymph nodes (LN) in non-small cell lung cancer (NSCLC). 115 patients with NSCLC were included in this study. The training dataset included 66 patients. A decision tree model was developed with 9 variables, and validated with 49 patients: short and long diameters of LNs, ratio of short and long diameters, maximum standardized uptake value (S ...[more]

PMID: 29486012

Similar Datasets

Project description:ObjectivesThe accurate assessment of lymph node metastases (LNMs) and the preoperative nodal (N) stage are critical for the precise treatment of patients with gastric cancer (GC). The diagnostic performance, however, of current imaging procedures used for this assessment is sub-optimal. Our aim was to investigate the value of preoperative 18F-FDG PET/CT radiomic features to predict LNMs and the N stage.MethodsWe retrospectively collected clinical and 18F-FDG PET/CT imaging data of 185 patients with GC who underwent total or partial radical gastrectomy. Patients were allocated to training and validation sets using the stratified method at a fixed ratio (8:2). There were 2,100 radiomic features extracted from the 18F-FDG PET/CT scans. After selecting radiomic features by the random forest, relevancy-based, and sequential forward selection methods, the BalancedBagging ensemble classifier was established for the preoperative prediction of LNMs, and the OneVsRest classifier for the N stage. The performance of the models was primarily evaluated by the AUC and accuracy, and validated by the independent validation methods. Analysis of the feature importance and the correlation were also conducted. We also compared the predictive performance of our radiomic models to that with the contrast-enhanced CT (CECT) and 18F-FDG PET/CT.ResultsThere were 185 patients-127 men, 58 women, with the median age of 62, and an age range of 22-86 years. One CT feature and one PET feature were selected to predict LNMs and achieved the best performance (AUC: 82.2%, accuracy: 85.2%). This radiomic model also detected some LNMs that were missed in CECT (19.6%) and 18F-FDG PET/CT (35.7%). For predicting the N stage, four CT features and one PET feature were selected (AUC: 73.7%, accuracy: 62.3%). Of note, a proportion of patients in the validation set whose LNMs were incorrectly staged by CECT (57.4%) and 18F-FDG PET/CT (55%) were diagnosed correctly by our radiomic model.ConclusionWe developed and validated two machine learning models based on the preoperative 18F-FDG PET/CT images that have a predictive value for LNMs and the N stage in GC. These predictive models show a promise to offer a potentially useful adjunct to current staging approaches for patients with GC.

Project description:BackgroundThe aim of this study was to assess the diagnostic utility of metabolic parameters on fluorine-18-fluoro-2-deoxy-D-glucose-positron emission tomography (FDG-PET)/computed tomography (CT) for predicting lymph node (LN) metastasis in patients with cN2 non-small cell lung cancer (NSCLC).MethodsWe retrospectively reviewed patients who underwent surgery for cN2 NSCLC between 2007 and 2020. Those who had clinically diagnosed positive hilar and mediastinal LNs by routine CT and PET/CT imaging were investigated. To measure the metabolic parameters of LNs, the data according to maximum standardized uptake value (SUVmax), metabolic tumor volume (MTV), total lesion glycolysis (TLG), and LN-to-primary tumor ratio of SUVmax (LPR) were examined. The diagnosis of each retrieved LN was confirmed based on histopathological examination of surgical tissue specimens. Receiver operating characteristics (ROC) curves with area under the curve (AUC) calculations and multivariate analysis by logistic regression were performed.ResultsForty-five patients with 84 clinically diagnosed positive hilar or mediastinal LNs were enrolled in the present study. Of the 84 LNs, 63 LNs were pathologically proven as positive (75%). The SUVmax, MTV, TLG, and LPR of LN metastasis were significantly higher than those of benign nodes. In the ROC analysis, the AUC value of LPR [AUC, 0.776; 95% confidence interval (CI), 0.640-0.913] was higher than that of LN SUVmax (AUC, 0.753; 95% CI, 0.626-0.880) or LN TLG3.5 (AUC, 0.746; 95% CI, 0.607-0.885). Using the optimal LPR cutoff value of 0.47, the sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were 84.1, 66.7, 88.3, 58.3, and 79.8%, respectively. Multivariate analysis by logistic regression showed that LPR was an independent predictor for LN metastasis (odds ratio, 6.45; 95% CI, 1.785-23.301; P = 0.004). In the subgroup analysis of adenocarcinoma patients (n = 18; 32 LNs), TLG3.5 was a better predictor (AUC, 0.816; 95% CI, 0.639-0.985) than LPR (AUC, 0.792; 95% CI, 0.599-0.986) or LN SUVmax (AUC, 0.792; 95% CI, 0.625-0.959).ConclusionsOur findings suggest that LPR on FDG-PET is a useful predictor for LN metastasis in patients with cN2 NSCLC. TLG can be a good predictor for LN metastasis in patients with adenocarcinoma.

Dataset Information

A decision tree model for predicting mediastinal lymph node metastasis in non-small cell lung cancer with F-18 FDG PET/CT.

Publications

A decision tree model for predicting mediastinal lymph node metastasis in non-small cell lung cancer with F-18 FDG PET/CT.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets