Construction and validation of a nomogram prediction model for early recurrence of patients undergoing radical pancreaticoduodenectomy for pancreatic ductal adenocarcinoma / 中华肝胆外科杂志

Objective:To study the risk factors for early recurrence of patients undergoing radical pancreaticoduodenectomy (PD) for pancreatic ductal adenocarcinoma (PDAC) and construct a normogram model.Methods:Patients undergoing open radical PD for PDAC at Faculty of Hepato-Pancreato-Biliary Surgery, Chinese PLA General Hospital from January 2014 to December 2021 were retrospectively screened. A total of 213 patients were enrolled, including 145 males and 68 females, aged (58.4±9.8) years. Patients were divided into the early recurrence group ( n=59, recurrence within 6 months after surgery) and a control group ( n=154, no recurrence within 6 months after surgery). Using minimum absolute value convergence and selection operator regression (LASSO) and multi-factor logistic regression analysis, we screened out the best predictor of early recurrence after PD for PDAC, and then established a nomogram model. The effectiveness of the model was validated by receiver operating characteristic (ROC) curve, calibration curves, and decision analysis curves. Results:Multivariate logistic regression analysis showed that patients with obstructive jaundice, vascular invasion, massive intraoperative bleeding, high-risk tumors (poorly differentiated or undifferentiated), high carbohydrate antigen 19-9 to total bilirubin ratio, and high fibrinogen and neutrophil to lymphocyte ratio scores had a higher risk of early postoperative recurrence. Based on the indexes above, a nomogram prediction model was constructed. The area under the ROC curve was 0.797 (95% CI: 0.726-0.854). Validation of the calibration curve exhibited good concordance between the predicted probability and ideal probability, decision curve analysis showed that the net benefits of the groupings established according to the model were all greater than 0 within the high risk threshold of 0.08 to 1.00. Conclusion:The nomogram for predicting early recurrence after PD for PDAC has a good efficiency, which could be helpful to screen out the high-risk patients for adjuvant or neoadjuvant therapy.

Validación prospectiva multicéntrica del EuroSCORE II en Argentina / Multicenter prospective validation of the EuroSCORE II in Argentina

Resumen Objetivo: Validar, en forma prospectiva y en múltiples centros, la precisión y utilidad clínica del European System for Cardiac Operative Risk Evaluation II (EuroSCORE II) para predecir la mortalidad operatoria de la cirugía cardíaca en centros de Argentina Método: Entre enero de 2012 y febrero de 2018 se incluyeron en forma prospectiva 2,000 pacientes consecutivos que fueron sometidos a cirugía cardíaca en diferentes centros de Argentina. El punto final fue mortalidad hospitalaria por cualquier causa. La discriminación, calibración, precisión y utilidad clínica del EuroSCORE II se evaluaron en la cohorte global y en los diferentes tipos de cirugías, basándose en las curvas Receiver Operating Characteristics (ROC), bondad de ajuste de Hosmer-Lemeshow, razón de mortalidad observada/esperada, índice de Shannon y curvas de decisión. Resultados: El área ROC del EuroSCORE II estuvo entre 0.73 y 0.80 para todo tipo de cirugía, y el valor más bajo fue para la cirugía coronaria. La mortalidad observada y esperada fue 4.3 y 3.0%, respectivamente (p = 0.034). El análisis de la curva de decisión demostró un beneficio neto positivo para los umbrales por debajo de 0.24 para todo tipo de cirugía. Conclusiones: El EuroSCORE II tuvo un desempeño adecuado en términos de discriminación y calibración para todos los tipos de cirugía, aunque algo inferior para la cirugía coronaria. Si bien en términos generales subestimó el riesgo en los grupos de riesgo intermedio, el comportamiento global fue aceptable. El EuroSCORE II podría considerarse una opción de modelo genérico y actualizado de estratificación del riesgo operatorio para predecir la mortalidad hospitalaria de la cirugía cardíaca en nuestro contexto.

Abstract Objective: To validate prospectively in multiple centers, the accuracy and clinical utility of the European System for Cardiac Operative Risk Evaluation (EuroSCORE II) to predict the operative mortality of cardiac surgery in Argentina. Methods: Between January 2012 and February 2018, 2,000 consecutive adult patients who underwent cardiac surgery in different centers in Argentina were prospectively included. The end-point was in-hospital all-cause mortality. Discrimination, calibration, precision and clinical utility of the EuroSCORE II were evaluated in the global cohort and in the different types of surgeries, based on ROC (Receiver Operating Characteristics) curves, Hosmer-Lemeshow goodness-of-fit test, observed/expected mortality ratio, Shannon index and decision curves analysis. Results: ROC area of the EuroSCORE II was between 0.73 and 0.80 for all types of surgery, being the lowest value for coronary surgery. The observed and expected mortality was 4.3% and 3.0%, respectively (p = 0.034). The decision curve analysis showed a positive net benefit for all thresholds below 0.24, considering all type of surgeries. Conclusion: The EuroSCORE II showed an adequate performance in terms of discrimination and calibration for all types of surgery, although somewhat inferior for coronary surgery. Though in general terms this model underestimated the risk in intermediate risk groups, its overall performance was acceptable. The EuroSCORE II could be considered an optional updated generic model of operative risk stratification to predict in-hospital mortality after cardiac surgery in our context.

Feasibility of constructing a unilateral sudden sensorineural hearing loss machine learning classification model based on diffusion tensor imaging / 中华放射学杂志

Objective@#To explore the feasibility of constructing a machine learning classification model for unilateral sudden sensorineural hearing loss (SSHL) patients and normal controls based on diffusion tensor imaging.@*Methods@#Prospective collection of 84 patients with untreated SSHL were recruited from the otolaryngology department of the Union Hospital of Tongji Medical College of Huazhong University of Science and Technology between June 2013 to May 2015 as the SSHL group. Meanwhile, a total of 63 healthy volunteers who were no any ear disease history, and the hearing function were confirmed with pure tone audiometry, were collected as the control group. All subjects underwent a brain DTI scan. The data were divided into the training set and validation set according to the ratio of 7 to 3, that was, the training set contained 58 cases of SSHL patients and 44 control groups, and the validation set included 26 cases of SSHL patients and 19 control groups. A vector which included the DTI parameters such as fractional anisotropy, mean diffusivity, axial diffusivity and radial diffusivity was constructed with the software R. The LASSO regression of machine learning method was used to perform feature dimensionality reduction and construct a classification model. The training set samples were used to map the nomogram based on the multivariate logistic analysis method, the validation set and the AUC were used to evaluate the prediction ability of the nomogram, and the calibration curve was used to evaluate the model.@*Results@#From the 200 feature vectors including the fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD) values of each brain region, after each dimension reduction process, a total of six features were retained, which were the MD of left superior corona radiate and right superior fronto-occipital fasciculus, the AD of the body of corpus callosum, and the RD of left inferior cerebellar peduncle, left superior corona radiate and right posterior limb of internal capsule. The six features of patients with unilateral SSHL were higher than the control group, and the difference was statistically significant (P<0.05). Based on this, a two-class model is constructed and a nomogram is drawn. The sensitivity, specificity, accuracy and AUC of the training set were 93.1% (54/58), 72.7% (32/44), 84.3% (86/102) and 0.854, respectively; the sensitivity, specificity, accuracy and AUC of validation set were 80.8% (21/26), 84.2% (16/19), 82.2% (37/45), 0.870, respectively. Nomogram could significantly improve the classification efficiency of the control group and patients, and the model with the LASSO method showed a higher prediction curve than other models.@*Conclusions@#The machine learning classification model based on DTI metrics can effectively distinguish patients with unilateral sudden sensorineural deafness from healthy control people.

Feasibility of constructing a unilateral sudden sensorineural hearing loss machine learning classification model based on diffusion tensor imaging / 中华放射学杂志

Objective To explore the feasibility of constructing a machine learning classification model for unilateral sudden sensorineural hearing loss (SSHL) patients and normal controls based on diffusion tensor imaging. Methods Prospective collection of 84 patients with untreated SSHL were recruited from the otolaryngology department of the Union Hospital of Tongji Medical College of Huazhong University of Science and Technology between June 2013 to May 2015 as the SSHL group. Meanwhile, a total of 63 healthy volunteers who were no any ear disease history, and the hearing function were confirmed with pure tone audiometry, were collected as the control group. All subjects underwent a brain DTI scan. The data were divided into the training set and validation set according to the ratio of 7 to 3, that was, the training set contained 58 cases of SSHL patients and 44 control groups, and the validation set included 26 cases of SSHL patients and 19 control groups. A vector which included the DTI parameters such as fractional anisotropy, mean diffusivity, axial diffusivity and radial diffusivity was constructed with the software R. The LASSO regression of machine learning method was used to perform feature dimensionality reduction and construct a classification model. The training set samples were used to map the nomogram based on the multivariate logistic analysis method, the validation set and the AUC were used to evaluate the prediction ability of the nomogram, and the calibration curve was used to evaluate the model. Results From the 200 feature vectors including the fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD) values of each brain region, after each dimension reduction process, a total of six features were retained, which were the MD of left superior corona radiate and right superior fronto-occipital fasciculus, the AD of the body of corpus callosum, and the RD of left inferior cerebellar peduncle, left superior corona radiate and right posterior limb of internal capsule. The six features of patients with unilateral SSHL were higher than the control group, and the difference was statistically significant (P<0.05). Based on this, a two-class model is constructed and a nomogram is drawn. The sensitivity, specificity, accuracy and AUC of the training set were 93.1% (54/58), 72.7% (32/44), 84.3% (86/102) and 0.854, respectively; the sensitivity, specificity, accuracy and AUC of validation set were 80.8% (21/26), 84.2% (16/19), 82.2% (37/45), 0.870, respectively. Nomogram could significantly improve the classification efficiency of the control group and patients, and the model with the LASSO method showed a higher prediction curve than other models. Conclusions The machine learning classification model based on DTI metrics can effectively distinguish patients with unilateral sudden sensorineural deafness from healthy control people.