Dataset Information

Development and Validation of the Random Forest Model via Combining CT-PET Image Features and Demographic Data for Distant Metastases among Lung Cancer Patients.

ABSTRACT: The work aimed at developing and validating a random forest model of CT-PET image features combined with demographic data to diagnose distant metastases among lung cancer patients. This study involved lung cancer patients from The Cancer Genome Atlas lung adenocarcinoma (TCGA-LUAD) dataset, the lung PET-CT dataset, the lung squamous cell carcinoma (LSCC) dataset, and the National Cancer Institute's Clinical Proteomic Tumor Analysis Consortium lung adenocarcinoma (CPTAC-LUAD) dataset and collected the information on 178 CT, 178 PET, and the patients' age, history of smoking, and gender. We conducted image processing and feature extraction. Finally, 4 computed tomography (CT) image features and 2 positron emission tomography (PET) image features were extracted. Four prediction models based on CT image features, PET image features, and demographic data were developed, and the area under the receiver operating characteristic (ROC) curve was used to evaluate the performance of prediction models. A total of 178 eligible samples were randomly divided into a training set (n = 134) and a testing set (n = 44) at a ratio of 3 : 1, with 2021 as a random number. ROC analyses illustrated that the predictive performance for distant metastases of combining CT-PET image features and demographic data for training and testing were 0.923 (95% confidence interval (CI): 0.873-0.973) and 0.873 (95% CI: 0.757-0.990). In addition, the predictive performance of the combined model in the testing set was significantly better than that of the CT-demographic data model (0.716, 95% CI: 0.531-0.902), PET-demographic data model (0.802, 95% CI: 0.633-0.970), and CT-PET model (0.797, 95% CI: 0.666-0.928). The random forest model via combining CT-PET image features and demographic data could have great performance in predicting distant metastases among lung cancer patients.

SUBMITTER: Bi L

PROVIDER: S-EPMC9767733 | biostudies-literature | 2022

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Development and Validation of the Random Forest Model via Combining CT-PET Image Features and Demographic Data for Distant Metastases among Lung Cancer Patients.

Bi Lijun L Guo Yi Y

Journal of healthcare engineering 20221213

The work aimed at developing and validating a random forest model of CT-PET image features combined with demographic data to diagnose distant metastases among lung cancer patients. This study involved lung cancer patients from The Cancer Genome Atlas lung adenocarcinoma (TCGA-LUAD) dataset, the lung PET-CT dataset, the lung squamous cell carcinoma (LSCC) dataset, and the National Cancer Institute's Clinical Proteomic Tumor Analysis Consortium lung adenocarcinoma (CPTAC-LUAD) dataset and collecte ...[more]

PMID: 36561373

Similar Datasets

Project description:Pancreatic ductal adenocarcinoma (PDAC) is one of the most aggressive and treatment-refractory malignancies. The lack of an effective screening tool results in the majority of patients being diagnosed at late stages, which underscores the urgent need to develop more sensitive and specific imaging modalities, particularly in detecting occult metastases, to aid clinical decision-making. The tumor microenvironment of PDAC is heavily infiltrated with myeloid-derived suppressor cells (MDSCs) that express C-C chemokine receptor type 2 (CCR2). These CCR2-expressing MDSCs accumulate at a very early stage of metastasis and greatly outnumber PDAC cells, making CCR2 a promising target for detecting early, small metastatic lesions that have scant PDAC cells. Herein, we evaluated a CCR2 targeting PET tracer (68Ga-DOTA-ECL1i) for PET imaging on PDAC metastasis in two mouse models. Positron emission tomography/computed tomography (PET/CT) imaging of 68Ga-DOTA-ECL1i was performed in a hemisplenic injection metastasis model (KI) and a genetically engineered orthotopic PDAC model (KPC), which were compared with 18F-FDG PET concurrently. Autoradiography, hematoxylin and eosin (H&E), and CCR2 immunohistochemical staining were performed to characterize the metastatic lesions. PET/CT images visualized the PDAC metastases in the liver/lung of KI mice and in the liver of KPC mice. Quantitative uptake analysis revealed increased metastasis uptake during disease progression in both models. In comparison, 18F-FDG PET failed to detect any metastases during the time course studies. H&E staining showed metastases in the liver and lung of KI mice, within which immunostaining clearly demonstrated the overexpression of CCR2 as well as CCR2+ cell infiltration into the normal liver. H&E staining, CCR2 staining, and autoradiography also confirmed the expression of CCR2 and the uptake of 68Ga-DOTA-ECL1i in the metastatic foci in KPC mice. Using our novel CCR2 targeted radiotracer 68Ga-DOTA-ECL1i and PET/CT, we demonstrated the sensitive and specific detection of CCR2 in the early PDAC metastases in two mouse models, indicating its potential in future clinical translation.

Project description:ObjectiveTo evaluate the prediction performance of 18F-PSMA-1007 PET/CT and clinicopathologic characteristics on prostate cancer (PCa) risk stratification and distant metastatic prediction.Materials and methodsA retrospective analysis was performed on 101 consecutively patients with biopsy or radical prostatectomy proved PCa who underwent 18F-PSMA-1007 PET/CT. The semi-quantitative analysis provided minimum, maximum and mean standardized uptake (SUVmin, SUVmax and SUVmean) of PCa. Association between clinicopathologic characteristics (total prostate-specific antigen, tPSA and Gleason Score, GS) and PET/CT indexes were analyzed. The diagnostic performance of distant metastatic on PET/CT parameters, tPSA and GS was evaluated using logistic regression analyses. A path analysis was conducted to evaluate the mediating effect of tPSA level on the relation between semi-quantitative parameters of primary tumors and metastatic lesions.ResultsThe PET/CT parameters were all higher in high risk stratification subgroups (tPSA>20 ng/mL, GS ≥ 8, and tPSA>20 ng/mL and/or GS ≥ 8, respectively) with high sensitivity (86.89%, 90.16% and 83.61%, respectively). The SUVmax, tPSA and GS could effectively predict distant metastatic with high sensitivity of SUVmax (90.50%) compared with tPSA (57.14%) and GS (55.61%). With a cutoff value of 29.01ng/mL for tPSA, the detection rate of distant metastasis between low and high prediction tPSA group had statistical differences (50.00% vs. 76.60%, respectively; P = 0.006) which was not found on guideline tPSA level (P>0.05). 6/15 (40%) patients tPSA between 20ng/mL to 29.01ng/mL without distant metastases may change the risk stratification. Finally, tPSA had a partial mediating effect on SUVmax of primary tumors and metastases lesions.ConclusionThe 18F-PSMA-1007 PET/CT SUVmax has a higher sensitivity and can be an "imaging biomarker" for primary PCa risk stratification. The prediction tPSA level (29.01 ng/mL) is more conducive to the assessment of distant metastasis and avoid unnecessary biopsy.

Project description:We present a probabilistic approach to identify patients with primary and secondary hepatic malignancies as responders or nonresponders to yttrium-90 radioembolization therapy. Recent advances in computer-aided detection have decreased false-negative and false-positive rates of perceived abnormalities; however, there is limited research in using similar concepts to predict treatment response. Our approach is driven by the goal of precision medicine to determine pretherapy fluorine-18-2-fluoro-2-deoxy-d-glucose positron emission tomography and computed tomography imaging parameters to facilitate the identification of patients who would benefit most from yttrium-90 radioembolization therapy, while avoiding complex and costly procedures for those who would not. Our algorithm seeks to predict a patient's response by discovering common co-occurring image patterns in the lesions of baseline fluorine-18-2-fluoro-2-deoxy-d-glucose positron emission tomography and computed tomography scans by extracting invariant shape and texture features. The extracted imaging features were represented as a distribution of each subject based on the bag-of-feature paradigm. The distribution was applied in a multinomial naive Bayes classifier to predict whether a patient would be a responder or nonresponder to yttrium-90 radioembolization therapy based on the imaging features of a pretherapy fluorine-18-2-fluoro-2-deoxy-d-glucose positron emission tomography and computed tomography scan. Comprehensive published criteria were used to determine lesion-based clinical treatment response based on fluorine-18-2-fluoro-2-deoxy-d-glucose positron emission tomography and computed tomography imaging findings. Our results show that the model is able to predict a patient with liver cancer as a responder or nonresponder to yttrium-90 radioembolization therapy with a sensitivity of 0.791 using extracted invariant imaging features from the pretherapy fluorine-18-2-fluoro-2-deoxy-d-glucose positron emission tomography and computed tomography test. The sensitivity increased to 0.821 when combining extracted invariant image features with variable features of tumor volume.

Project description:We extracted image features from serial 18F-labeled fluorodeoxyglucose (FDG) positron emission tomography (PET) / computed tomography (CT) scans of anal cancer patients for the prediction of tumor recurrence after chemoradiation therapy (CRT). Seventeen patients (4 recurrent and 13 non-recurrent) underwent three PET/CT scans at baseline (Pre-CRT), in the middle of the treatment (Mid-CRT) and post-treatment (Post-CRT) were included. For each patient, Mid-CRT and Post-CRT scans were aligned to Pre-CRT scan. Comprehensive image features were extracted from CT and PET (SUV) images within manually delineated gross tumor volume, including geometry features, intensity features and texture features. The difference of feature values between two time points were also computed and analyzed. We employed univariate logistic regression model, multivariate model, and naïve Bayesian classifier to analyze the image features and identify useful tumor recurrent predictors. The area under the receiver operating characteristic (ROC) curve (AUC) was used to evaluate the accuracy of the prediction. In univariate analysis, six geometry, three intensity, and six texture features were identified as significant predictors of tumor recurrence. A geometry feature of Roundness between Post-CRT and Pre-CRT CTs was identified as the most important predictor with an AUC value of 1.00 by multivariate logistic regression model. The difference of Number of Pixels on Border (geometry feature) between Post-CRT and Pre-CRT SUVs and Elongation (geometry feature) of Post-CRT CT were identified as the most useful feature set (AUC = 1.00) by naïve Bayesian classifier. To investigate the early prediction ability, we used features only from Pre-CRT and Mid-CRT scans. Orientation (geometry feature) of Pre-CRT SUV, Mean (intensity feature) of Pre-CRT CT, and Mean of Long Run High Gray Level Emphasis (LRHGLE) (texture feature) of Pre-CRT CT were identified as the most important feature set (AUC = 1.00) by multivariate logistic regression model. Standard deviation (intensity feature) of Mid-CRT SUV and difference of Mean of LRHGLE (texture feature) between Mid-CRT and Pre-CRT SUVs were identified as the most important feature set (AUC = 0.86) by naïve Bayesian classifier. The experimental results demonstrated the potential of serial PET/CT scans in early prediction of anal tumor recurrence.

Project description:Early detection of gastrointestinal stromal tumor (GIST) liver metastases is crucial for the management and prognosis. In our experience, GIST liver metastases can display hypermetabolism on 18F-fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) and marked enhancement on magnetic resonance imaging (MRI), which are uncommon in other tumors before treatment. Most literature focus on the imaging evaluation, prognosis after treatment and less is known about imaging features on both imaging methods before treatment. This study analyzes the imaging features of newly diagnosed GIST liver metastases on 18F-FDG PET/CT and MRI, with goal of improving diagnostic accuracy. This retrospective study included 55 patients with pathological or radiographical confirmed GIST liver metastases who underwent PET/CT (n=29), MRI (n=22), or both methods (n=4). PET/CT and MRI interpretation including lesion's morphologic features, number, density or signal intensity, hemorrhage, cystic changes or necrosis, maximum standardized uptake value (SUVmax) of liver metastases and liver background on PET imaging, degree and pattern of enhancement on MRI were obtained by two experienced nuclear medicine physicians and two radiologists respectively. Data are presented as numbers, percentages, means ± standard deviations or median (interquartile range). The correlation between diameter and SUVmax of metastases, and primary tumor SUVmax and synchronous liver metastases SUVmax were analyzed by Spearman's rank test. On PET/CT visual analysis, 38.9%, 23.9%, and 37.2% of lesions showed significant hypermetabolism, slightly higher metabolism, and equal or lower metabolism than liver, respectively. There was a weak correlation between the diameter and SUVmax of liver metastases (rs =0.370, P<0.001), and a moderate correlation between SUVmax of synchronous liver metastases and the primary tumors (rs =0.492, P<0.001). On contrast-enhanced MRI, 90.8% of lesions showed heterogeneous enhancement in the arterial phase with the variable presentation, and 74.3% had different enhancement patterns between margins and intratumoral parenchyma. Liver lesions in GIST displaying significant, slight hypermetabolism on 18F-FDG PET/CT, marked or heterogeneous gradual enhancement within the intratumoral parenchyma with ring-like enhancement on MRI may denote the diagnosis of liver metastasis. However, GIST liver metastases may also display equal or lower metabolism than liver parenchyma on PET, making small lesions more difficult to diagnose.

Dataset Information

Development and Validation of the Random Forest Model via Combining CT-PET Image Features and Demographic Data for Distant Metastases among Lung Cancer Patients.

Publications

Development and Validation of the Random Forest Model via Combining CT-PET Image Features and Demographic Data for Distant Metastases among Lung Cancer Patients.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets