Area Under The Curve Of Nomogram Research Articles

The radiomics nomogram predicts the prognosis of pancreatic cancer patients with hepatic metastasis after chemoimmunotherapy

ObjectiveTo construct a prognostic model based on MR features and clinical data to evaluate the progression free survival (PFS), overall survival (OS) and objective response rate (ORR) of pancreatic cancer patients with hepatic metastases who received chemoimmunotherapy.Methods105 pancreatic cancer patients with hepatic metastases who received chemoimmunotherapy were assigned to the training set (n = 52), validation set (n = 22), and testing set (n = 31). Multi-lesion volume of interest were delineated, multi-sequence radiomics features were extracted, and the radiomics models for predicting PFS, OS and ORR were constructed, respectively. Clinical variables were extracted, and the clinical models for predicting PFS, OS and ORR were constructed, respectively. The nomogram was jointly constructed by radiomics model and clinical model.ResultThe ORR exhibits no significant correlation with either PFS or OS. The area under the curve (AUC) of nomogram for predicting 6-month PFS reached 0.847 (0.737–0.957), 0.786 (0.566–1.000) and 0.864 (0.735–0.994) in the training set, validation set and testing set, respectively. The AUC of nomogram for predicting 1-year OS reached 0.770 (0.635–0.906), 0.743 (0.479–1.000) and 0.818 (0.630–1.000), respectively. The AUC of nomogram for predicting ORR reached 0.914 (0.828–1.00), 0.938 (0.840–1.00) and 0.846 (0.689–1.00), respectively.ConclusionThe prognostic models based on MR imaging features and clinical data are effective in predicting the PFS, OS and ORR of chemoimmunotherapy in pancreatic cancer patients with hepatic metastasis, and can be used to evaluate the prognosis of patients.

Differentiation of acute and chronic vertebral compression fractures using conventional CT based on deep transfer learning features and hand-crafted radiomics features

BackgroundWe evaluated the diagnostic efficacy of deep learning radiomics (DLR) and hand-crafted radiomics (HCR) features in differentiating acute and chronic vertebral compression fractures (VCFs).MethodsA total of 365 patients with VCFs were retrospectively analysed based on their computed tomography (CT) scan data. All patients completed MRI examination within 2 weeks. There were 315 acute VCFs and 205 chronic VCFs. Deep transfer learning (DTL) features and HCR features were extracted from CT images of patients with VCFs using DLR and traditional radiomics, respectively, and feature fusion was performed to establish the least absolute shrinkage and selection operator. The MRI display of vertebral bone marrow oedema was used as the gold standard for acute VCF, and the model performance was evaluated using the receiver operating characteristic (ROC).To separately evaluate the effectiveness of DLR, traditional radiomics and feature fusion in the differential diagnosis of acute and chronic VCFs, we constructed a nomogram based on the clinical baseline data to visualize the classification evaluation. The predictive power of each model was compared using the Delong test, and the clinical value of the nomogram was evaluated using decision curve analysis (DCA).ResultsFifty DTL features were obtained from DLR, 41 HCR features were obtained from traditional radiomics, and 77 features fusion were obtained after feature screening and fusion of the two. The area under the curve (AUC) of the DLR model in the training cohort and test cohort were 0.992 (95% confidence interval (CI), 0.983-0.999) and 0.871 (95% CI, 0.805-0.938), respectively. While the AUCs of the conventional radiomics model in the training cohort and test cohort were 0.973 (95% CI, 0.955-0.990) and 0.854 (95% CI, 0.773-0.934), respectively. The AUCs of the features fusion model in the training cohort and test cohort were 0.997 (95% CI, 0.994-0.999) and 0.915 (95% CI, 0.855-0.974), respectively. The AUCs of nomogram constructed by the features fusion in combination with clinical baseline data were 0.998 (95% CI, 0.996–0.999) and 0.946 (95% CI, 0.906–0.987) in the training cohort and test cohort, respectively. The Delong test showed that the differences between the features fusion model and the nomogram in the training cohort and the test cohort were not statistically significant (P values were 0.794 and 0.668, respectively), and the differences in the other prediction models in the training cohort and the test cohort were statistically significant (P < 0.05). DCA showed that the nomogram had high clinical value.ConclusionThe features fusion model can be used for the differential diagnosis of acute and chronic VCFs, and its differential diagnosis ability is improved when compared with that when either radiomics is used alone. At the same time, the nomogram has a high predictive value for acute and chronic VCFs and can be a potential decision-making tool to assist clinicians, especially when a patient is unable to undergo spinal MRI examination.

A nomogram for clinical estimation of acute biliary pancreatitis risk among patients with symptomatic gallstones: A retrospective case-control study.

Background/PurposeCurrently, there are no effective tools to accurately assess acute biliary pancreatitis (ABP) risk in patients with gallstones. This study aimed to develop an ABP risk nomogram in patients with symptomatic gallstones.MethodsWe conducted a retrospective nested case-control study and data on 816 conservatively treated patients with symptomatic gallstones admitted to The First Affiliated Hospital of Harbin Medical University between January 6, 2007 and January 22, 2016 were retrospectively collected. We conducted a propensity-score matched (PSM) analysis based on follow-up time in a ratio of 1:4 between ABP group (n=65) and non-ABP group (n=260). These matched patients were randomly divided into study cohort (n=229) and validation cohort (n=96) according to a ratio of 7:3. In the study cohort, independent risk factors for ABP occurrence identified using Cox regression were included in nomogram. Nomogram performance and discrimination were assessed using the concordance index (C-index), area under the curve (AUC), calibration curve, decision curve analysis (DCA) and clinical impact curve (CIC). The model was also validated in the validation cohort.ResultsNomogram was based on 7 independent risk factors: age, diabetes history, gallbladder wall thickness, gallstone diameter, coexisting common bile duct (CBD) stones, direct bilirubin (DBIL), and white blood cell count (WBC). The C-index of nomogram was 0.888, and the 10-year AUCs of nomogram was 0.955. In the validation cohort, nomogram still had good discrimination (C-index, 0.857; 10-year AUC, 0.814). The calibration curve showed good homogeneity between the prediction by nomogram and the actual observation. DCA and CIC demonstrated that nomogram was clinically useful.ConclusionsThe ABP risk nomogram incorporating 7 features is useful to predict ABP risk in symptomatic gallstone patients.

A Nomogram for Predicting Patent Foramen Ovale-Related Stroke Recurrence

BackgroundThe high prevalence of patent foramen ovale (PFO) in cryptogenic stroke suggested a stroke-causing role for PFO. As risk factors for recurrence of such stroke are not recognized, clinicians cannot sufficiently identify, treat, and follow-up high-risk patients. Therefore, this study aimed to establish a prediction model for PFO-related stroke recurrence.MethodsThis study included 392 patients with PFO-related stroke in a training set and 164 patients with PFO-related stroke in an independent validation set. In the training set, independent risk factors for recurrence identified using forward stepwise Cox regression were included in nomogram 1, and those identified using least absolute shrinkage and selection operator(LASSO)regression were included in nomogram 2. Nomogram performance and discrimination were assessed using the concordance index (C-index), area under the curve (AUC), calibration curve, and decision curve analyses (DCA). The results were also validated in the validation set.ResultsNomogram 1 was based on homocysteine (Hcy), high-sensitivity C-reactive protein (hsCRP), and albumin (ALB), and nomogram 2 was based on age, diabetes, hypertension, right-to-left shunt, ALB, prealbumin, hsCRP, and Hcy. The C-index of nomogram 1 was 0.861, which was not significantly different from that of nomogram 2 (0.893). The 2- and 5-year AUCs of nomogram 1 were 0.863 and 0.777, respectively. In the validation set, nomogram 1 still had good discrimination (C-index, 0.862; 2-year AUC, 0.839; 5-year AUC, 0.990). The calibration curve showed good homogeneity between the prediction by nomogram 1 and the actual observation. DCA demonstrated that nomogram 1 was clinically useful. Moreover, patients were successfully divided into two distinct risk groups (low and high risk) for recurrence rate by nomogram 1.ConclusionsNomogram 1, based on Hcy, hsCRP, and ALB levels, provided a more clinically realistic prognostic prediction for patients with PFO-related stroke. This model could help patients with PFO-related stroke to facilitate personalized prognostic evaluations.

Biparametric Magnetic Resonance Imaging-Derived Nomogram to Detect Clinically Significant Prostate Cancer by Targeted Biopsy for Index Lesion.

Currently, it is necessary to investigate how to combine biparametric magnetic resonance imaging (bpMRI) with various clinical parameters for the detection of clinically significant prostate cancer (csPCa). To develop a multivariate prebiopsy nomogram using clinical and bpMRI parameters for estimating the probability of csPCa. Retrospective, single-center study. Two hundred and twenty-six patients who underwent targeted biopsy (TBx) for the MRI-suspected index lesion because of clinical suspicions of PCa. A 3 T MRI including turbo spin-echo T2 -weighted and diffusion-weighted single-shot echo-planar imaging sequences. Prebiopsy clinical and bpMRI parameters were patient age, biopsy history (biopsy-naïve or repeated biopsy status), prostate-specific antigen density (PSAD), Prostate Imaging-Reporting and Data System version 2.1 (PI-RADSv2.1), and apparent diffusion coefficient ratio (ADCR). ADCR was defined as mean ADC of the index lesion divided by mean ADC of the contralateral prostatic region. A multivariate prebiopsy nomogram for csPCa (i.e. Gleason sum ≥7) was developed. Area under the curve (AUC) of each parameter and prebiopsy nomogram was assessed. Five-fold cross-validation was performed for robust estimation of performance of the prebiopsy nomogram. Logistic regression, receiver-operating curve, and 5-fold cross-validation. P-value < 0.05 was considered statistically significant. Proportion of csPCa was 31.9% (72/226). The AUCs of age, biopsy-naïve status, PSAD, PI-RADSv2.1, ADCR, and prebiopsy nomogram were 0.657 (95% confidence interval [CI], 0.580-0.733), 0.593 (95% CI, 0.525-0.660), 0.762 (95% CI, 0.697-0.826), 0.824 (95% CI, 0.770-0.878), 0.829 (95% CI, 0.769-0.888), and 0.906 (95% CI, 0.863-0.948), respectively: AUC of nomogram was significantly different than that of individual parameter. In the 5-fold cross-validation, the mean AUC of the prebiopsy nomogram for csPCa was 0.888 (95% CI, 0.786-0.983). This multivariate prebiopsy nomogram using clinical and bpMRI parameters may help estimate the probability of csPCa in patients undergoing TBx. ADCR seems to enhance the role of bpMRI in detecting csPCa. 3 TECHNICAL EFFICACY: Stage 2.