scholarly journals CT Radiomics Model for Discriminating the Risk Stratification of Gastrointestinal Stromal Tumors: A Multi-Class Classification and Multi-Center Study

2021 ◽  
Vol 11 ◽  
Author(s):  
Zhonghua Chen ◽  
Linyi Xu ◽  
Chuanmin Zhang ◽  
Chencui Huang ◽  
Minhong Wang ◽  
...  
Keyword(s):  
Risk Stratification ◽  
Prediction Model ◽  
Validation Cohort ◽  
External Validation ◽  
Low Risk ◽  
Class Prediction ◽  
Training Cohort ◽  
Stromal Tumors ◽  
Risk Patients ◽  
Multi Class Classification

ObjectiveTo establish and verify a computed tomography (CT)-based multi-class prediction model for discriminating the risk stratification of gastrointestinal stromal tumors (GISTs).Materials and MethodsA total of 381 patients with GISTs were confirmed by surgery and pathology. Information on 213 patients were obtained from one hospital and used as training cohort, whereas the details of 168 patients were collected from two other hospitals and used as independent validation cohort. Regions of interest on CT images of arterial and venous phases were drawn, radiomics features were extracted, and dimensionality reduction processing was performed. Using a one-vs-rest method, a Random Forest-based GISTs risk three-class prediction model was established, and the receiver operating characteristic curve (ROC) was used to evaluate the performance of the multi-class classification model, and the generalization ability was verified using external data.ResultsThe training cohort included 96 very low-risk and low-risk, 60 intermediate-risk and 57 high-risk patients. External validation cohort included 82 very low-risk and low-risk, 48 intermediate-risk and 38 high-risk patients. The GISTs risk three-class radiomics model had a macro/micro average area under the curve (AUC) of 0.84 and an accuracy of 0.78 in the training cohort. It had a stable performance in the external validation cohort, with a macro/micro average AUC of 0.83 and an accuracy of 0.80.ConclusionCT radiomics can discriminate GISTs risk stratification. The performance of the three-class radiomics prediction model is good, and its generalization ability has also been verified in the external validation cohort, indicating its potential to assist stratified and accurate treatment of GISTs in the clinic.

scholarly journals A prediction model of Contrast-associated Acute Kidney Injury in Patients with Hypoalbuminemia Undergoing Coronary Angiography

2020 ◽  
Author(s):  
Liwei Liu ◽  
Jin Liu ◽  
Li Lei ◽  
Bo Wang ◽  
Guoli Sun ◽  
...  
Keyword(s):  
Acute Kidney Injury ◽  
Coronary Angiography ◽  
Risk Stratification ◽  
Prediction Model ◽  
Validation Cohort ◽  
External Validation ◽  
Kidney Injury ◽  
Good Tool ◽  
Lemeshow Test ◽  
Risk Patients

Abstract Background: Risk stratification is recommended as the key step to prevent contrast-associated acute kidney injury (CA-AKI) by allowing for prevention among at-risk patients undergoing coronary angiography (CAG) or percutaneous coronary intervention (PCI). Patients with hypoalbuminemia are prone to CA-AKI and do not have their own risk stratification tool. Therefore, we developed and validated a model for predicting CA-AKI in patients with hypoalbuminemia undergoing CAG/PCI.Methods: A total of 1272 consecutive patients with hypoalbuminemia undergoing CAG/PCI were enrolled and randomly assigned (2:1 ratio) to a development cohort (n = 848) and a validation cohort (n = 424). CA-AKI was defined as a serum creatinine (SCr) increase of ≥ 0.3 mg/dL or 50% from baseline within the first 48 to 72 hours following CAG/PCI. A prediction model was established with independent predictors according to multivariate logistic regression and a stepwise approach, showing as a nomogram. The discrimination of the nomogram was assessed by the area under the receiver operating characteristic (ROC) curve and was compared to the classic Mehran CA-AKI score. Calibration was assessed using the Hosmer–Lemeshow test.Results: Overall, 8.4% (71/848) of patients in the development cohort and 11.2% (48/424) of patients in the validation cohort experienced CA-AKI. The simple nomogram included estimated glomerular filtration rate (eGFR), serum albumin (ALB), age and the use of intra-aortic balloon pump (IABP); showed better predictive ability than the Mehran score (C-index 0.756 vs. 0.693, p = 0.02); and had good calibration (Hosmer–Lemeshow test p = 0.187). Conclusions: Our data suggested that the simple model might be a good tool for predicting CA-AKI in high-risk patients with hypoalbuminemia undergoing CAG/PCI, but our findings require further external validation.Trial registration number NCT01400295

scholarly journals Computed-Tomography-Based Radiomics Model for Predicting the Malignant Potential of Gastrointestinal Stromal Tumors Preoperatively: A Multi-Classifier and Multicenter Study

2021 ◽  
Vol 11 ◽  
Author(s):  
Minhong Wang ◽  
Zhan Feng ◽  
Lixiang Zhou ◽  
Liang Zhang ◽  
Xiaojun Hao ◽  
...  
Keyword(s):  
Logistic Regression ◽  
Random Forest ◽  
Risk Stratification ◽  
Gastrointestinal Stromal Tumors ◽  
Validation Cohort ◽  
External Validation ◽  
Random Forest Model ◽  
Training Cohort ◽  
Stromal Tumors ◽  
Forest Model

Background: Our goal was to establish and verify a radiomics risk grading model for gastrointestinal stromal tumors (GISTs) and to identify the optimal algorithm for risk stratification.Methods: We conducted a retrospective analysis of 324 patients with GISTs, the presence of which was confirmed by surgical pathology. Patients were treated at three different hospitals. A training cohort of 180 patients was collected from the largest center, while an external validation cohort of 144 patients was collected from the other two centers. To extract radiomics features, regions of interest (ROIs) were outlined layer by layer along the edge of the tumor contour on CT images of the arterial and portal venous phases. The dimensionality of radiomic features was reduced, and the top 10 features with importance value above 5 were selected before modeling. The training cohort used three classifiers [logistic regression, support vector machine (SVM), and random forest] to establish three GIST risk stratification prediction models. The receiver operating characteristic curve (ROC) was used to compare model performance, which was validated by external data.Results: In the training cohort, the average area under the curve (AUC) was 0.84 ± 0.07 of the logistic regression, 0.88 ± 0.06 of the random forest, and 0.81 ± 0.08 of the SVM. In the external validation cohort, the AUC was 0.85 of the logistic regression, 0.90 of the random forest, and 0.80 of the SVM. The random forest model performed the best in both the training and the external validation cohorts and could be generalized.Conclusion: Based on CT radiomics, there are multiple machine-learning models that can predict the risk of GISTs. Among them, the random forest algorithm had the highest prediction efficiency and could be readily generalizable. Through external validation data, we assume that the random forest model may be used as an effective tool to guide preoperative clinical decision-making.

scholarly journals Risk Stratification of Cytogenetically Normal Acute Myeloid Leukemia With Biallelic CEBPA Mutations Based on a Multi-Gene Panel and Nomogram Model

2021 ◽  
Vol 11 ◽  
Author(s):  
Li-Xin Wu ◽  
Hao Jiang ◽  
Ying-Jun Chang ◽  
Ya-Lan Zhou ◽  
Jing Wang ◽  
...  
Keyword(s):  
Acute Myeloid Leukemia ◽  
High Risk ◽  
Risk Stratification ◽  
Myeloid Leukemia ◽  
Validation Cohort ◽  
Low Risk ◽  
Training Cohort ◽  
High Risk Patients ◽  
Risk Patients ◽  
Acute Myeloid

BackgroundApproximately 30% of Chinese individuals with cytogenetically normal acute myeloid leukemia (CN-AML) have biallelic CEBPA (biCEBPA) mutations. The prognosis and optimal therapy for these patients are controversial in clinical practice.MethodsIn this study, we performed targeted region sequencing of 236 genes in 158 individuals with this genotype and constructed a nomogram model based on leukemia-free survival (LFS). Patients were randomly assigned to a training cohort (N =111) and a validation cohort (N =47) at a ratio of 7:3. Risk stratification was performed by the prognostic factors to investigate the risk-adapted post-remission therapy by Kaplan–Meier method.ResultsAt least 1 mutated gene other than CEBPA was identified in patients and mutation number was associated with LFS (61.6% vs. 39.0%, P =0.033), survival (85.6% vs. 62.9%, P =0.030) and cumulative incidence of relapse (CIR) (38.4% vs. 59.5%, P =0.0496). White blood cell count, mutations in CFS3R, KMT2A and DNA methylation related genes were weighted to construct a nomogram model and differentiate two risk subgroups. Regarding LFS, low-risk patients were superior to the high-risk (89.3% vs. 33.8%, P <0.001 in training cohort; 87.5% vs. 18.2%, P =0.009 in validation cohort). Compared with chemotherapy, allogenic hematopoietic stem cell transplantation (allo-HSCT) improved 5-year LFS (89.6% vs. 32.6%, P <0.001), survival (96.9% vs. 63.6%, P =0.001) and CIR (7.2% vs. 65.8%, P <0.001) in high-risk patients but not low-risk patients (LFS, 77.4% vs. 88.9%, P =0.424; survival, 83.9% vs. 95.5%, P =0.173; CIR, 11.7% vs. 11.1%, P =0.901).ConclusionsOur study indicated that biCEBPA mutant-positive CN-AML patients could be further classified into two risk subgroups by four factors and allo-HSCT should be recommended for high-risk patients as post-remission therapy. These data will help physicians refine treatment decision-making in biCEBPA mutant-positive CN-AML patients.

P0616REAL-TIME CLINICAL DECISION SUPPORT TO PREDICT AND MANAGE IN-HOSPITAL ACUTE KIDNEY INJURY UTILIZING RECURRENT NEURAL NETWORK

2020 ◽  
Vol 35 (Supplement_3) ◽  
Author(s):  
Hea Eun Kim ◽  
Hyeonsik Yang ◽  
Sejoong Kim ◽  
Kipyo Kim
Keyword(s):  
Prediction Model ◽  
Validation Cohort ◽  
External Validation ◽  
Vital Signs ◽  
Clinical Decision ◽  
Internal Validation ◽  
Training Cohort ◽  
National University ◽  
Seoul National University ◽  
Dynamic Variables

Abstract Background and Aims Rapidly Increasing electronic health record (EHR) data and recent development of machine learning methods offers the possibilities of improvement in quality of care in clinical practice. Machine learning can incorporate huge amount of features into the model, and enable non-linear algorithms with great performance. Previously published AKI prediction models have simple design without real-time assessment. Major risk factors in in-hospital AKI include use of various nephrotoxins, repeatedly measured laboratory findings, and vital signs, which are dynamic variables rather than static. Given that recurrent neural network (RNN) is a powerful tool to handle the sequential data, using RNN method in the prediction model is a promising approach. Therefore, in the present study, we proposed a RNN-based prediction model with external validation for in-hospital AKI and aimed to provide a framework to link the developed model with clinical decision supports. Method Study populations were all patients aged ≥ 18 years and hospitalized more than a week at Seoul National University Bundang Hospital (SNUBH) from 2013 to 2017 (training cohort) and at Seoul National University Hospital (SNUH) in 2017 (validation cohort). All demographics, laboratory values, vital signs, and clinical conditions were obtained from the EHR of each hospital. A total of 102 variables included in the model. Each variable falls into two categories: static and dynamic variable; static variable was time-invariant values during hospitalization, and dynamic variables were daily-updated values. Baseline creatinine was determined by searching the minimum serum Cr level within 2 weeks before admission. We developed two different models (model 1 and model 2) using RNN algorithms. The outcome for model 1 was the occurrence of AKI within 7 days from the present. In model 2, we constructed the prediction model of the trajectory of Cr values after 24 hours, 48 hours, and 72 hours, using available Cr values from 7 days ago to the present. Internal validation was performed by 5-fold cross validation using the training set (SNUBH), and then external validation was done using test set (SNUH). Results A total of 40,552 patients in training cohort and 4,000 patients in external validation cohort (test cohort) were included in the study. The mean age of participants was 62.2 years in training cohort and 58.7 years in test cohort. Baseline eGFR was 93.8 ± 40.4 ml/min/1.73m2 in training cohort and 88.4 ± 23.2 ml/min/1.73m2 in test cohort. In model 1 for the prediction of AKI occurrence within 7 days, the area under the curve was 0.93 (sensitivity 0.90, specificity 0.96) in internal validation, and 0.83 (sensitivity 0.83, specificity 0.82) in external validation. The model 2 predicted the creatinine trajectory within 3 days accurately; root mean square error was 0.1 in training cohort and 0.3 in test cohort. To support the clinical decision for AKI manage, we estimated the predicted trajectories of future creatinine levels after renal insult removal, such as nephrotoxic drugs, based on the established model 2. Conclusion We developed and validated a real-time AKI prediction model using RNN algorithms. This model showed high performance and can accurately visualize future creatinine trajectories. In addition, the model can provide the information about modifiable factors in patients with high risk of AKI.

scholarly journals Development of a preoperative prognostic scoring system to predict benefits of hepatic resection in advanced hepatocellular carcinoma patients

2021 ◽  
Vol 41 (4) ◽  
Author(s):  
Chang-Zhi Chen ◽  
Jian-Hong Zhong ◽  
Ya-Peng Qi ◽  
Jie Zhang ◽  
Tao Huang ◽  
...  
Keyword(s):  
Hepatocellular Carcinoma ◽  
Overall Survival ◽  
Hepatic Resection ◽  
Scoring System ◽  
Risk Group ◽  
External Validation ◽  
Low Risk ◽  
Advanced Hepatocellular Carcinoma ◽  
Training Cohort ◽  
Risk Patients

Abstract Objective: The present study aimed to identify risk factors for overall survival in advanced hepatocellular carcinoma (HCC) patients and establish a scoring system to select patients who would benefit from hepatic resection. Methods: Survival curves were analyzed using the Kaplan–Meier method and log-rank test. The prognostic scoring system was developed from training cohort using a Cox-regression model and validated in a external validation cohort Results: There were 401 patients in the training cohort, 163 patients in the external validation cohorts. The training cohort median survival in all patients was 12 ± 1.07 months, rate of overall survival was 49.6% at 1 year, 25.0% at 3 years, and 18.0% at 5 years. A prognostic scoring system was established based on age, body mass index, alkaline phosphatase, tumor number and tumor capsule. Patients were classified as low- risk group(≤3.5) or high-risk group(>3.5). High-risk patients had a median survival of 9 months, compared with 23 months in low-risk patients. The area under the receiver operating characteristic curve (AUC) of the prognostic scoring system was 0.747 (0.694–0.801), which is significantly better than AFP, Child-Pugh and ALBI. The AUC of validation cohorts was 0.716 (0.63–0.803). Conclusion: A prognostic scoring system for hepatic resection in advanced HCC patients has been developed based entirely on preoperative variables. Patients classified as low risk using this system may experience better prognosis after hepatic resection.

scholarly journals Nomogram for the Prediction of Contrast-associated Acute Kidney Injury in Patients with Hypoalbuminemia Undergoing Coronary Angiography

2020 ◽  
Author(s):  
Liwei Liu ◽  
Jin Liu ◽  
Li Lei ◽  
Bo Wang ◽  
Guoli Sun ◽  
...  
Keyword(s):  
Acute Kidney Injury ◽  
Coronary Angiography ◽  
Risk Stratification ◽  
Validation Cohort ◽  
External Validation ◽  
Kidney Injury ◽  
Coronary Intervention ◽  
Good Tool ◽  
Lemeshow Test ◽  
Risk Patients

Abstract Background: Risk stratification is recommended as the key step to prevent contrast-associated acute kidney injury (CA-AKI) by allowing for prevention among at-risk patients undergoing coronary angiography (CAG) or percutaneous coronary intervention (PCI). Patients with hypoalbuminemia are prone to CA-AKI and do not have their own risk stratification tool. Therefore, we developed and validated a nomogram for predicting CA-AKI in patients with hypoalbuminemia undergoing CAG/PCI.Methods: A total of 1272 consecutive patients with hypoalbuminemia undergoing CAG/PCI were enrolled and randomly assigned (2:1 ratio) to a development cohort (n = 848) and a validation cohort (n = 424). CA-AKI was defined as a serum creatinine (SCr) increase of ≥ 0.3 mg/dL or 50% from baseline within the first 48 to 72 hours following CAG/PCI. A nomogram was established with independent predictors according to multivariate logistic regression and a stepwise approach. The discrimination of the nomogram was assessed by the area under the receiver operating characteristic (ROC) curve and was compared to the classic Mehran CA-AKI score. Calibration was assessed using the Hosmer–Lemeshow test.Results: Overall, 8.4% (71/848) of patients in the development cohort and 11.2% (48/424) of patients in the validation cohort experienced CA-AKI. The simple nomogram included estimated glomerular filtration rate (eGFR), serum albumin (ALB), age and the use of intra-aortic balloon pump (IABP); showed better predictive ability than the Mehran score (C-index 0.756 vs. 0.693, p = 0.02); and had good calibration (Hosmer–Lemeshow test p = 0.187). Decision curve analysis showed that the nomogram was more clinically useful than the Mehran score.Conclusions: Our data suggested that the simple nomogram might be a good tool for predicting CA-AKI in high-risk patients with hypoalbuminemia undergoing CAG/PCI, but our findings require further external validation.Trial registration number NCT01400295

scholarly journals Genetic Risk of Severe Chronic Graft-Versus-Host Disease Defined By Host-Derived CXCR3 Ligands

Blood ◽  
2018 ◽  
Vol 132 (Supplement 1) ◽  
pp. 357-357
Author(s):  
Hao Dai ◽  
Sivaramakrishna P. Rachakonda ◽  
Olaf Penack ◽  
Aleksandar Radujkovic ◽  
Carsten Müller-Tidow ◽  
...  
Keyword(s):  
High Risk ◽  
Graft Versus Host Disease ◽  
Genetic Risk ◽  
Validation Cohort ◽  
Low Risk ◽  
Genetic Score ◽  
Training Cohort ◽  
Graft Versus Host ◽  
Post Transplant ◽  
Risk Patients

Abstract Introduction Severe chronic graft-versus-host disease (cGVHD) is the leading cause of morbidity and mortality in long-term survivors after allogeneic stem cell transplantation (alloSCT). The CXCR3 signalling pathway may be implicated in cGVHD pathophysiology, since CXCR3 and its ligands (CXCL4, CXCL9, CXCL10 and CXCL11) are involved in attracting activated Th1 cells into inflamed tissues. To better understand the role of the CXCR3 axis in cGVHD, we measured serum levels of CXCR3 ligands in allograft recipients pre-transplant and on day 28 post-transplant, and correlated them to the single nucleotide polymorphisms (SNPs) in recipient CXCR3/CXCR3L genes in the context of severe cGVHD. Patients and methods: 287 patients who were allografted at Heidelberg University Hospital, survived more than 6 months after alloSCT, and did not receive statin-based endothelial prophylaxis (SEP) constituted the no-SEP training cohort, whereas 401 patients who received SEP constituted the SEP cohort. DNA for genotyping was available for 545 patients (no-SEP 242, SEP 303) and sera for measuring CXCL9, CXCL10 and CXCL11 by ELISA were collected at pre-transplant for 405 patients (no-SEP 109, SEP 296) and at day +28 for 494 patients (no-SEP 152, SEP 342). The no-SEP validation cohort consisted of 202 patients who had been allografted at Berlin Charité and survived more than 6 months. CGVHD was diagnosed and graded using the National Institutes of Health's 2005 consensus criteria. Eighteen SNPs (7 in CXCL9-11, 7 in CXCL4 and 4 in CXCR3 loci) were selected and analyzed for association with severe cGVHD, treating non-relapse death and relapse without severe cGVHD as competing events. Associations of SNPs with serum chemokine levels were studied by Mann-Whitney U-test. Hazard ratios (HR) with 95% confidence interval (CI) were estimated using (cause-specific) Cox regression. Covariates considered in multivariate analysis were age, diagnosis, donor type, sex of donor and recipient and usage of ATG. Results: Overall, 50 of 287 patients (17.4%) in the no-SEP training cohort, 53 of 401 patients (13.2%) in the SEP cohort and 48 of 202 patients (23.8%) in the no-SEP validation cohort developed at least one episode of severe cGVHD. In the no-SEP training cohort, higher serum CXCL9 levels at day +28 were significantly associated with a higher risk of severe cGVHD in univariate analysis (HR 1.38 for every log2-fold change, 95% CI 1.10-1.75, P=0.01; Figure 1a). No significant association was found for serum CXCL10 and CXCL11 pre- or post-transplant with severe cGVHD. The rs884304 SNP in CXCL9-11 locus showed a significant association with severe cGVHD; patients with AA/AG genotypes carried a HR of 2.32 (95%CI 1.21-4.46, P=0.01) compared to patients with GG genotypes. In addition, 3 other SNPs (rs3733236 and rs4282209 in CXCL9-11, rs655328 in CXCL4 loci) were selected based on the effect on severe cGVHD (P <0.10) to calculate a combined genetic risk score. Patients with any low-risk genotypes (rs884304GG, rs3733236AA/AG, rs4282209AA/AG and rs655328TT) were classified as the low-risk. All others were considered as high-risk. Taken together, high-risk patients were found in 21.4% (52/242) of the no-SEP training cohort, 22.4% (68/303) of the SEP cohort and 19.8% (40/202) of the no-SEP validation cohort. Patients in the high-risk group had significantly higher serum CXCL9 levels at day +28 (Figure 1b) and a significantly higher risk of severe cGVHD (Figure 1c) on both univariate (HR 2.68, 95%CI 1.45-4.95, P=0.001) and multivariate analyses (HR 2.49, 95%CI 1.33-4.66, P=0.004). The effect of the combined score was confirmed in the no-SEP validation cohort (HR 3.02, 95%CI 1.60-5.72, P=0.001). In contrast, in the SEP cohort the adverse effect of high risk genotypes was not observed (HR 1.30, 95% CI 0.60-2.79, P=0.50). In addition, SEP reduced day +28 CXCL9 levels in patients with high-risk genotype but not in low-risk patients. Conclusion: In the absence of SEP, the risk of severe cGVHD could be predicted both by a genetic score of 4 SNPs in recipient CXCR3L genes and by serum CXCL9 levels at day +28. The genetic score influenced serum CXCL9 levels at day +28. Our results suggest that in high-risk patients, host-derived CXCR3 ligands are upregulated early after alloSCT and may promote the development of severe cGVHD. Endothelial prophylaxis may reduce the risk of severe cGVHD by regulating serum CXCL9 levels and, thus, warrants further study. Disclosures No relevant conflicts of interest to declare.

scholarly journals Immune Reconstitution-Based Score for Risk Stratification of Chronic Graft-Versus-Host Disease Patients

2021 ◽  
Vol 11 ◽  
Author(s):  
Fabio Serpenti ◽  
Francesca Lorentino ◽  
Sarah Marktel ◽  
Raffaella Milani ◽  
Carlo Messina ◽  
...  
Keyword(s):  
Overall Survival ◽  
High Risk ◽  
Risk Stratification ◽  
Immune Reconstitution ◽  
Validation Cohort ◽  
Multivariate Model ◽  
Low Risk ◽  
Intermediate Risk ◽  
High Risk Patients ◽  
Risk Patients

IntroductionAllogeneic stem cell transplantation survivors are at a relevant risk of developing chronic GvHD (cGvHD), which importantly affects quality of life and increases morbidity and mortality. Early identification of patients at risk of cGvHD-related morbidity could represent a relevant tool to tailor preventive strategies. The aim of this study was to evaluate the prognostic power of immune reconstitution (IR) at cGvHD onset through an IR-based score.MethodsWe analyzed data from 411 adult patients consecutively transplanted between January 2011 and December 2016 at our Institution: 151 patients developed cGvHD (median follow-up 4 years). A first set of 111 consecutive patients with cGvHD entered the test cohort while an additional consecutive 40 patients represented the validation cohort. A Cox multivariate model for OS (overall survival) in patients with cGvHD of any severity allowed the identification of six variables independently predicting OS and TRM (transplant-related mortality). A formula for a prognostic risk index using the β coefficients derived from the model was designed. Each patient was assigned a score defining three groups of risk (low, intermediate, and high).ResultsOur multivariate model defined the variables independently predicting OS at cGvHD onset: CD4+ &gt;233 cells/mm3, NK &lt;115 cells/mm3, IgA &lt;0.43g/L, IgM &lt;0.45g/L, Karnofsky PS &lt;80%, platelets &lt;100x103/mm3. Low-risk patients were defined as having a score ≤3.09, intermediate-risk patients &gt;3.09 and ≤6.9, and high-risk patients &gt;6.9. By ROC analysis, we identified a cut-off of 6.310 for both TRM and overall mortality.In the training cohort, the 6-year OS and TRM from cGvHD occurrence were 85% (95% CI, 70-92) and 13% (95% CI, 5-25) for low-risk, 64% (95% CI, 44-89) and 30% (95% CI, 15-47) for intermediate-risk, 26% (95% CI, 10-47), and 42% (95% CI, 19-63) for high-risk patients (OS p&lt;0.0001; TRM p = 0.015).The validation cohort confirmed the model with a 6-year OS and TRM of 83% (95% CI, 48-96) and 8% (95% CI, 1-32) for low-risk, 78% (95% CI, 37-94) and 11% (95% CI, 1-41) for intermediate-risk, 37% (95% CI, 17-58), and 63% (95% CI, 36-81) for high-risk patients (OS p = 0.0075; TRM p = 0.0009).ConclusionsIR score at diagnosis of cGvHD predicts GvHD severity and overall survival. IR score may contribute to the risk stratification of patients. If confirmed in a larger and multicenter-based study, IR score could be adopted to identify patients at high risk and modulate cGvHD treatments accordingly in the context of clinical trial.

scholarly journals Development and Validation of a Nomogram for Preoperative Prediction of Lymph Node Metastasis in Lung Adenocarcinoma Based on Radiomics Signature and Deep Learning Signature

2021 ◽  
Vol 11 ◽  
Author(s):  
Jia Ran ◽  
Ran Cao ◽  
Jiumei Cai ◽  
Tao Yu ◽  
Dan Zhao ◽  
...  
Keyword(s):  
Logistic Regression ◽  
Regression Analysis ◽  
Deep Learning ◽  
Prediction Model ◽  
Validation Cohort ◽  
External Validation ◽  
Training Cohort ◽  
Preoperative Prediction ◽  
Lemeshow Test ◽  
Radiomics Signature

Background and PurposeThe preoperative LN (lymph node) status of patients with LUAD (lung adenocarcinoma) is a key factor for determining if systemic nodal dissection is required, which is usually confirmed after surgery. This study aimed to develop and validate a nomogram for preoperative prediction of LN metastasis in LUAD based on a radiomics signature and deep learning signature.Materials and MethodsThis retrospective study included a training cohort of 200 patients, an internal validation cohort of 40 patients, and an external validation cohort of 60 patients. Radiomics features were extracted from conventional CT (computed tomography) images. T-test and Extra-trees were performed for feature selection, and the selected features were combined using logistic regression to build the radiomics signature. The features and weights of the last fully connected layer of a CNN (convolutional neural network) were combined to obtain a deep learning signature. By incorporating clinical risk factors, the prediction model was developed using a multivariable logistic regression analysis, based on which the nomogram was developed. The calibration, discrimination and clinical values of the nomogram were evaluated.ResultsMultivariate logistic regression analysis showed that the radiomics signature, deep learning signature, and CT-reported LN status were independent predictors. The prediction model developed by all the independent predictors showed good discrimination (C-index, 0.820; 95% CI, 0.762 to 0.879) and calibration (Hosmer-Lemeshow test, P=0.193) capabilities for the training cohort. Additionally, the model achieved satisfactory discrimination (C-index, 0.861; 95% CI, 0.769 to 0.954) and calibration (Hosmer-Lemeshow test, P=0.775) when applied to the external validation cohort. An analysis of the decision curve showed that the nomogram had potential for clinical application.ConclusionsThis study presents a prediction model based on radiomics signature, deep learning signature, and CT-reported LN status that can be used to predict preoperative LN metastasis in patients with LUAD.

Sign in / Sign up

Export Citation Format

Share Document