Noninvasive prediction of perineural invasion in intrahepatic cholangiocarcinoma by clinicoradiological features and computed tomography radiomics based on interpretable machine learning: a multicenter cohort study.

BACKGROUND: Perineural invasion (PNI) of intrahepatic cholangiocarcinoma (ICC) is a strong independent risk factor for tumour recurrence and long-term patient survival. However, there is a lack of noninvasive tools for accurately predicting the PNI status. The authors develop and validate a combined model incorporating radiomics signature and clinicoradiological features based on machine learning for predicting PNI in ICC, and used the Shapley Additive explanation (SHAP) to visualize the prediction process for clinical application. METHODS: This retrospective and prospective study included 243 patients with pathologically diagnosed ICC (training, n =136; external validation, n =81; prospective, n =26, respectively) who underwent preoperative contrast-enhanced computed tomography between January 2012 and May 2023 at three institutions (three tertiary referral centres in Guangdong Province, China). The ElasticNet was applied to select radiomics features and construct signature derived from computed tomography images, and univariate and multivariate analyses by logistic regression were used to identify the significant clinical and radiological variables with PNI. A robust combined model incorporating radiomics signature and clinicoradiological features based on machine learning was developed and the SHAP was used to visualize the prediction process. A Kaplan-Meier survival analysis was performed to compare prognostic differences between PNI-positive and PNI-negative groups and was conducted to explore the prognostic information of the combined model. RESULTS: Among 243 patients (mean age, 61.2 years ± 11.0 (SD); 152 men and 91 women), 108 (44.4%) were diagnosed as PNI-positive. The radiomics signature was constructed by seven radiomics features, with areas under the curves of 0.792, 0.748, and 0.729 in the training, external validation, and prospective cohorts, respectively. Three significant clinicoradiological features were selected and combined with radiomics signature to construct a combined model using machine learning. The eXtreme Gradient Boosting exhibited improved accuracy and robustness (areas under the curves of 0.884, 0.831, and 0.831, respectively). Survival analysis showed the construction combined model could be used to stratify relapse-free survival (hazard ratio, 1.933; 95% CI: 1.093-3.418; P =0.021). CONCLUSIONS: We developed and validated a robust combined model incorporating radiomics signature and clinicoradiological features based on machine learning to accurately identify the PNI statuses of ICC, and visualize the prediction process through SHAP for clinical application.

The added value of apparent diffusion coefficient and microcalcifications to the Kaiser score in the evaluation of BI-RADS 4 lesions.

PURPOSE: To explore the added value of combining microcalcifications or apparent diffusion coefficient (ADC) with the Kaiser score (KS) for diagnosing BI-RADS 4 lesions. METHODS: This retrospective study included 194 consecutive patients with 201 histologically verified BI-RADS 4 lesions. Two radiologists assigned the KS value to each lesion. Adding microcalcifications, ADC, or both these criteria to the KS yielded KS1, KS2, and KS3, respectively. The potential of all four scores to avoid unnecessary biopsies was assessed using the sensitivity and specificity. Diagnostic performance was evaluated by the area under the curve (AUC) and compared between KS and KS1. RESULTS: The sensitivity of KS, KS1, KS2, and KS3 ranged from 77.1% to 100.0%.KS1 yielded significantly higher sensitivity than other methods (P < 0.05), except for KS3 (P > 0.05), most of all, when assessing NME lesions. For mass lesions, the sensitivity of these four scores was comparable (p > 0.05). The specificity of KS, KS1, KS2, and KS3 ranged from 56.0% to 69.4%, with no statistically significant differences(P > 0.05), except between KS1 and KS2 (p < 0.05).The AUC of KS1 (0.877) was significantly higher than that of KS (0.837; P = 0.0005), particularly for assessing NME (0.847 vs 0.713; P < 0.0001). CONCLUSION: KS can stratify BI-RADS 4 lesions to avoid unnecessary biopsies. Adding microcalcifications, but not adding ADC, as an adjunct to KS improves diagnostic performance, particularly for NME lesions. ADC provides no additional diagnostic benefit to KS. Thus, only combining microcalcifications with KS is most conducive to clinical practice.

Development and validation of a clinical predictive model for 1-year prognosis in coronary heart disease patients combine with acute heart failure.

Background: The risk factors for acute heart failure (AHF) vary, reducing the accuracy and convenience of AHF prediction. The most common causes of AHF are coronary heart disease (CHD). A short-term clinical predictive model is needed to predict the outcome of AHF, which can help guide early therapeutic intervention. This study aimed to develop a clinical predictive model for 1-year prognosis in CHD patients combined with AHF. Materials and methods: A retrospective analysis was performed on data of 692 patients CHD combined with AHF admitted between January 2020 and December 2020 at a single center. After systemic treatment, patients were discharged and followed up for 1-year for major adverse cardiovascular events (MACE). The clinical characteristics of all patients were collected. Patients were randomly divided into the training (n = 484) and validation cohort (n = 208). Step-wise regression using the Akaike information criterion was performed to select predictors associated with 1-year MACE prognosis. A clinical predictive model was constructed based on the selected predictors. The predictive performance and discriminative ability of the predictive model were determined using the area under the curve, calibration curve, and clinical usefulness. Results: On step-wise regression analysis of the training cohort, predictors for MACE of CHD patients combined with AHF were diabetes, NYHA ≥ 3, HF history, Hcy, Lp-PLA2, and NT-proBNP, which were incorporated into the predictive model. The AUC of the predictive model was 0.847 [95% confidence interval (CI): 0.811-0.882] in the training cohort and 0.839 (95% CI: 0.780-0.893) in the validation cohort. The calibration curve indicated good agreement between prediction by nomogram and actual observation. Decision curve analysis showed that the nomogram was clinically useful. Conclusion: The proposed clinical prediction model we have established is effective, which can accurately predict the occurrence of early MACE in CHD patients combined with AHF.

Predictive Value of Non-high-Density Lipoprotein Cholesterol and Neutrophil-Lymphocyte Ratio for Coronary Artery Vulnerable Plaques in Type 2 Diabetes Mellitus.

Background: Patients with diabetes have an increased risk of developing vulnerable plaques (VPs), in which dyslipidemia and chronic inflammation play important roles. Non-high-density lipoprotein cholesterol (non-HDL-C) and neutrophil-lymphocyte ratio (NLR) have emerged as potential markers of both coronary artery VPs and cardiovascular prognosis. This study aimed to investigate the predictive value of non-HDL-C and NLR for coronary artery VPs in patients with type 2 diabetes mellitus (T2DM). Methods: We retrospectively enrolled 204 patients with T2DM who underwent coronary computed tomography angiography between January 2018 and June 2020. Clinical data including age, sex, hypertension, smoking, total cholesterol, low-density lipoprotein cholesterol, HDL-C, triglyceride, non-HDL-C, glycated hemoglobin, neutrophil count, lymphocyte count, NLR, and platelet count were analyzed. Multivariate logistic regression was used to estimate the association between non-HDL-C, NLR, and coronary artery VPs. Receiver operating curve analysis was performed to evaluate the value of non-HDL-C, NLR, and their combination in predicting coronary artery VPs. Results: In our study, 67 patients (32.84%) were diagnosed with VPs, 75 (36.77%) with non-VP, and 62 (30.39%) with no plaque. Non-HDL-C and NLR were independent risk factors for coronary artery VPs in patients with T2DM. The areas under the ROC curve of non-HDL-C, NLR, and their combination were 0.748 [95% confidence interval (CI): 0.676-0.818], 0.729 (95% CI: 0.650-0.800), and 0.825 (95% CI: 0.757-0.887), respectively. Conclusion: Either non-HDL-C or NLR could be used as a predictor of coronary artery VPs in patients with T2DM, but the predictive efficiency and sensitivity of their combination would be better.