Cardiovascular diseases are the leading cause of death and their diagnosis and treatment rely heavily on the variety of clinical data. With the advent of the era of medical big data, artificial intelligence (AI) has been widely applied in many aspects such as imaging, diagnosis and prognosis prediction in cardiovascular medicine, providing a new method for accurate diagnosis and treatment. This paper reviews the application of AI in cardiovascular medicine.
Acute kidney injury (AKI) is a complication with high morbidity and mortality after cardiac surgery. In order to predict the incidence of AKI after cardiac surgery, many risk prediction models have been established worldwide. We made a detailed introduction to the composing features, clinical application and predictive capability of 14 commonly used models. Among the 14 risk prediction models, age, congestive heart failure, hypertension, left ventricular ejection fraction, diabetes, cardiac valve surgery, coronary artery bypass grafting (CABG) combined with cardiac valve surgery, emergency surgery, preoperative creatinine, preoperative estimated glomerular filtration rate (eGFR), preoperative New York Heart Association (NYHA) score>Ⅱ, previous cardiac surgery, cadiopulmonary bypass (CPB) time and low cardiac output syndrome (LCOS) are included in many risks prediction models (>3 times). In comparison to Mehta and SRI models, Cleveland risk prediction model shows the best discrimination for the prediction of renal replacement therapy (RRT)-AKI and AKI in the European. However, in Chinese population, the predictive ability of the above three risk prediction models for RRT-AKI and AKI is poor.
ObjectiveTo explore the influencing factors for Hook-wire precise positioning under CT guidance, determine the best positioning management strategy, and develop Nomogram prediction model. Methods Patients who underwent CT-guided Hook-wire puncture positioning in our hospital from July 2018 to November 2022 were selected. They were randomly divided into a training set and a validation set with a ratio of 7 : 3. Clinical data of the patients were analyzed, and the logistic analysis was used to screen out the risk factors that affected CT-guided Hook-wire precise positioning for the training set. The Nomogram prediction model was constructed according to the risk factors, and the goodness of fit test and clinical decision curve analysis were performed. ResultsA total of 199 patients with CT-guided Hook-wire puncture were included in this study, including 72 males and 127 females, aged 25-83 years. There were 139 patients in the training set and 60 patients in the validation set. In the training set, 70 patients were accurately located, with an incidence of 50.36%. Logistic regression analysis showed that height [OR=3.46, 95%CI (1.44, 8.35), P=0.006], locating needle perpendicular to the horizontal plane [OR=3.40, 95%CI (1.37, 8.43), P=0.008], locating needle perpendicular to the tangent line of skin surface [OR=6.01, 95%CI (2.38, 15.20), P＜0.001], CT scanning times [OR=3.03, 95%CI (1.25, 7.33), P=0.014], occlusion [OR=10.56, 95%CI (1.98, 56.48), P=0.006] were independent risk factors for CT-guided Hook-wire precise localization. The verification results of the Nomogram prediction model based on these independent risk factors showed that the area under the receiver operating characteristic curve (AUC) was 0.843 [95%CI (0.776, 0.910)], and the predicted value of the correction curve was basically consistent with the measured value. The AUC of the model in the validation set was 0.854 [95%CI (0.759, 0.950)]. The decision curves showed that when the threshold probability was within the range of 8%-85% in the training set and 18%-99% in the validation set, there was a high net benefit value. Conclusion Height, the locating needle perpendicular to the horizontal plane, the locating needle perpendicular to the tangent line of skin surface, number of CT scans, and occlusion are independent risk factors for CT-guided Hook-wire accurate localization. The Nomogram model established based on the above risk factors can accurately assess and quantify the risk of CT-guided Hook-wire accurate localization.
Objective To develop and validate a nomogram prediction model for pulmonary infection in patients following cardiac valve replacement surgery, providing a reference for early screening of high-risk populations and implementing targeted preventive measures. Methods Clinical data of patients who underwent cardiac valve replacement surgery at the Second Affiliated Hospital of Anhui Medical University from January 2020 to October 2023 were collected. Patients were randomly assigned to a modeling group and a validation group in a 7 : 3 ratio. Based on the occurrence of pulmonary infection post-surgery, patients were divided into a pulmonary infection group and a non-pulmonary infection group. Risk factors for pulmonary infection after cardiac valve replacement were analyzed using least absolute shrinkage and selection operator (LASSO) regression and logistic regression to establish a risk prediction model, which was subsequently validated. Model evaluation was conducted using the area under the receiver operating characteristic curve (AUC), calibration curves, and decision curve analysis. Results A total of 689 patients were included, comprising 354 males and 335 females, with a median age of 58.0 (50.0, 68.0) years. The incidence of pulmonary infection was 16.0% (110/689). Independent risk factors for pulmonary infection following cardiac valve replacement included emergency admission, smoking history, chronic obstructive pulmonary disease, duration of cardiopulmonary bypass, duration of tracheal intubation, and postoperative renal injury. The AUC for the modeling group was 0.911 [95%CI (0.877, 0.946) ], with a Hosmer-Lemeshow χ2-value of 6.577 (P=0.583) in the modeling group. The AUC value was 0.891 [95%CI (0.840, 0.941) ], with a Hosmer-Lemeshow χ2-value of 5.486 (P=0.705) in the validation group. The model demonstrated good discrimination, calibration, and net benefit. Conclusion The established nomogram prediction model has significant predictive value and can be applied to risk assessment and individualized treatment for pulmonary infection in patients following cardiac valve replacement surgery.
Objective To establish and verify the early prediction model of critical illness patients with influenza. Methods Critical illness patients with influenza who diagnosed with influenza in the emergency departments from West China Hospital of Sichuan University, Shangjin Hospital of West China Hospital of Sichuan University, and Panzhihua Central Hospital between January 1, 2017 and June 30, 2020 were selected. According to K-fold cross validation method, 70% of patients were randomly assigned to the model group, and 30% of patients were assigned to the model verification group. The patients in the model group and the model verification group were divided into the critical illness group and the non-critical illness group, respectively. Based on the modified National Early Warning Score (MEWS) and the Simplified British Thoracic Society Score (confusion, uremia, respiratory, BP, age 65 years, CRB-65 score), a critical illness influenza early prediction model was constructed and its accuracy was evaluated. Results A total of 612 patients were included. Among them, there were 427 cases in the model group and 185 cases in the model verification group. In the model group, there were 304 cases of non-critical illness and 123 cases of critical illness. In the model verification group, there were 152 cases of non-critical illness and 33 cases of critical illness. The results of binary logistic regression analysis showed that age, hypertension, the number of days between the onset of symptoms and presentation at the emergency department, consciousness state, white blood cell count, and lymphocyte count, oxygen saturation of blood were the independent risk factors for critical illness influenza. Based on these 7 risk factors, an early prediction model for critical illness influenza was established. The correct percentages of the model for non-critical illness and critical illness patients were 95.4% and 77.2%, respectively, with an overall correct prediction percentage of 90.2%. The results of the receiver operator characteristic curve showed that the sensitivity and specificity of the early prediction model for critical illness influenza in predicting critical illness patients were 0.909, 0.921, and the area under the curve and its 95% confidence interval were 0.931 (0.860, 0.999). The sensitivity, specificity, and area under the curve (0.935, 0.865, 0.942) of the early prediction model for critical illness influenza were higher than those of MEWS (0.642, 0.595, 0.536) and CRB-65 (0.628, 0.862, 0.703). Conclusions The conclusion is that age, hypertension, the number of days between the onset of symptoms and presentation at the emergency department, consciousness, oxygen saturation, white blood cell count, and absolute lymphocyte count are independent risk factors for predicting severe influenza cases. The early prediction model for critical illness patients with influenza has high accuracy in predicting severe influenza cases, and its predictive value and accuracy are superior to those of the MEWS score and CRB-65 score.
Objective To systematically evaluate the prognostic prediction model for chronic heart failure patients in China, and provide reference for the construction, application, and promotion of related prognostic prediction models. Methods A comprehensive search was conducted on the studies related to prognostic prediction model for Chinese patients with chronic heart failure published in The Cochrane Library, PubMed, EMbase, Web of Science, CNKI, VIP, Wanfang, and the China Biological Medicine databases from inception to March 31, 2023. Two researchers strictly followed the inclusion and exclusion criteria to independently screen literature and extract data, and used the prediction model risk of bias assessment tool (PROBAST) to evaluate the quality of the models. Results A total of 25 studies were enrolled, including 123 prognostic prediction models for chronic heart failure patients. The area under the receiver operating characteristic curve (AUC) of the models ranged from 0.690 to 0.959. Twenty-two studies mostly used random splitting and Bootstrap for internal model validation, with an AUC range of 0.620-0.932. Seven studies conducted external validation of the model, with an AUC range of 0.720-0.874. The overall bias risk of all models was high, and the overall applicability was low. The main predictive factors included in the models were the N-terminal pro-brain natriuretic peptide, age, left ventricular ejection fraction, New York Heart Association heart function grading, and body mass index. Conclusion The quality of modeling methodology for predicting the prognosis of chronic heart failure patients in China is poor, and the predictive performance of different models varies greatly. For developed models, external validation and clinical application research should be vigorously carried out. For model development research, it is necessary to comprehensively consider various predictive factors related to disease prognosis before modeling. During modeling, large sample and prospective studies should be conducted strictly in accordance with the PROBAST standard, and the research results should be comprehensively reported using multivariate prediction model reporting guidelines to develop high-quality predictive models with strong scalability.
ObjectiveTo study the predictability of dynamic prediction model of clinical pancreatic fistula in patients with or without pancreatic duct stent in laparoscopic pancreaticoduodenectomy (LPD).MethodsA total of 66 patients who underwent LPD in West China Hospital of Sichuan University from November 2019 to October 2020 were enrolled in the randomized controlled trial (registration number: ChiCTR1900026653). The perioperative data of the patients were collected in real time. The patients were divided into groups according to whether the pancreatic duct support tube was retained during the operation, and the probability prediction value was output according to the model formula. The specificity, sensitivity, accuracy, discrimination, and stability of the prediction results were analyzed.ResultsFor the group with pancreatic stent tubes, the specificity, sensitivity, and accuracy of the model at the model cut-off points on the postoperative day 2, 3 and 5 were 92.0%, 76.7% and 57.1%, 50.0%, 100% and 66.7%, and 88.8%, 78.8% and 61.3%, respectively. The areas under the ROC curve were 0.870, 0.956 and 0.702, respectively. The kappa values of the prediction result based on model cut-off point and cut-off point of ROC curve were 0.308, 0.582 and 0.744, respectively. Whereas for those who without the stent tube, the specificity, sensitivity, and prediction accuracy of the model on the postoperative day 5 were 66.7%, 100% and 72%, respectively. The area under curve at different time points were 0.304, 0.821, and 0.958, respectively. The kappa values at the last two time points were 0.465 and 0.449, respectively.ConclusionsFor patients with pancreatic duct support during LPD operation, the dynamic model of clinical pancreatic fistula can more accurately screen high-risk groups of clinical pancreatic fistula, and has better stability of prediction results. For patients without supporting tube, in the case of flexible adjustment of the boundary point, the model can also be more accurate screening on the 3rd and 5th days after operation.
Objective To externally validate a prediction model based on clinical and CT imaging features for the preoperative identification of high-grade patterns (HGP), such as micropapillary and solid subtypes, in early-stage lung adenocarcinoma, in order to guide clinical treatment decisions. Methods This study conducted an external validation of a previously developed prediction model using a cohort of patients with clinical stage ⅠA lung adenocarcinoma from the Fourth Hospital of Hebei Medical University. The model, which incorporated factors including tumor size, density, and lobulation, was assessed for its discrimination, calibration performance, and clinical impact. Results A total of 650 patients (293 males, 357 females; age range: 30-82 years) were included. The validation showed that the model demonstrated good performance in discriminating HGP (area under the curve>0.7). After recalibration, the model's calibration performance was improved. Decision curve analysis (DCA) indicated that at a threshold probability>0.6, the number of HGP patients predicted by the model closely approximated the actual number of cases. Conclusion This study confirms the effectiveness of a clinical and imaging feature-based prediction model for identifying HGP in stage ⅠA lung adenocarcinoma in a clinical setting. Successful application of this model may be significant for determining surgical strategies and improving patients' prognosis. Despite certain limitations, these findings provide new directions for future research.
ObjectiveProlonged mechanical ventilation (PMV) is a prognostic marker for short-term adverse outcomes in patients after lung transplantation.The risk of prolonged mechanical ventilation after lung transplantation is still not clear. The study to identify the risk factors of prolonged mechanical ventilation (PMV) after lung transplantation.Methods This retrospective observational study recruited patients who underwent lung transplantation in Wuxi People’s Hospital from January 2020 to December 2022. Relevant information was collected from patients and donors, including recipient data (gender, age, BMI, blood type, comorbidities), donor data (age, BMI, time of endotracheal intubation, oxygenation index, history of smoking, and any comorbidity with multidrug-resistant bacterial infections), and surgical data (surgical mode, incision type, operation time, cold ischemia time of the donor lung, intraoperative bleeding, and ECMO support), and postoperative data (multi-resistant bacterial lung infection, multi-resistant bacterial bloodstream infection, and mean arterial pressure on postoperative admission to the monitoring unit). Patients with a duration of mechanical ventilation ≤72 hours were allocated to the non-prolonged mechanical ventilation group, and patients with a duration of mechanical ventilation>72 hours were allocated to the prolonged mechanical ventilation group. LASSO regression analysis was applied to screen risk factors., and a clinical prediction model for the risk of prolonged mechanical ventilation after lung.ResultsPatients who met the inclusion criteria were divided into the training set and the validation set. There were 307 cases in the training set group and 138 cases in the validation set group. The basic characteristics of the training set and the validation set were compared. There were statistically significant differences in the recipient’s BMI, donor’s gender, CRKP of the donor lung swab, whether the recipient had pulmonary infection before the operation, the type of transplantation, the cold ischemia time of the donor lung, whether ECMO was used during the operation, the duration of ECMO assistance, CRKP of sputum, and the CRE index of the recipient's anal test (P<0.05). 2. The results of the multivariate logistic regression model showed that female recipients, preoperative mechanical ventilation in recipients, preoperative pulmonary infection in recipients, intraoperative application of ECMO, and the detection of multi-drug resistant Acinetobacter baumannii, multi-drug resistant Klebsiella pneumoniae and maltoclomonas aeruginosa in postoperative sputum were independent risk factors for prolonged mechanical ventilation after lung transplantation. The AUC of the clinical prediction model in the training set and the validation set was 0.838 and 0.828 respectively, suggesting that the prediction model has good discrimination. In the decision curves of the training set and the validation set, the threshold probabilities of the curves in the range of 0.05-0.98 and 0.02-0.85 were higher than the two extreme lines, indicating that the model has certain clinical validity.ConclusionsFemale patients, Preoperative pulmonary infection, preoperative mechanical ventilation,blood type B, blood type O, application of ECMO assistance, multi-resistant Acinetobacter baumannii infection, multi-resistant Klebsiella pneumoniae infection, and multi-resistant Stenotrophomonas maltophilia infection are independent risk factors for PMV (prolonged mechanical ventilation) after lung transplantation.
ObjectiveTo construct a model for predicting prognosis risk in patients with pancreatic malignancy (PM).MethodsThe clinicopathological data of 8 763 patients with PM undergone resection between 2010 and 2015 were collected and analyzed by SEER*Stat (v8.3.5) and R software, respectively. The univariate and multivariate Cox proportional hazard regression analysis were used to analyze the factors for predicting prognosis outcome risk and constructed the nomograms of patients with PM, respectively. Kaplan-Meier method was used to evaluate the survival of patients according to relevant factors and the high risk group and low risk group of patients with PM. The discriminative ability and calibration of the nomograms to predict overall survival were tested by using C-index, area under ROC curve (AUC) and calibration plots.ResultsThe multivariate Cox proportional hazard regression analysis showed that age, T staging, N staging, M staging, histological type, the differentiation, number of regional lymph node dissection, chemotherapy, and radiotherapy were independent factors for predicting the prognosis of patients with PM (P<0.05). Based on regression analysis of patients with PM, a nomograms model for predicting the risk of patients with PM was established, including age, T staging, N staging, M staging, histological type, the differentiation, tumor location, type of surgery, number of regional lymph node dissection, chemotherapy, and radiotherapy. The discriminative ability and calibration of the nomograms revealed good predictive ability as indicated by the C-index (0.747 for modeling group and 0.734 for verification group). The 3- and 5-year survival AUC values of the modeling group were 0.766 and 0.781, and the validation group were 0.758 and 0.783, respectively. The calibration plots showed that predictive value of the 3- and 5-year survival were close to the actual values in both modeling group and the verification group. ConclusionsIndependent predictors of survival risk after curative-intent surgery for PM were selected to create nomograms for predicting overall survival. The nomograms provide a basis for judging the prognosis of PM patients.