Predicting Delayed Neurocognitive Recovery After Non-cardiac Surgery Using Resting-State Brain Network Patterns Combined With Machine Learning

Brain Structural and Functional Changes Associated With Postoperative Neurocognitive Disorders: Research Update

Postoperative neurocognitive disorders (PNDs) are frequent and serious perioperative complications in the elderly, and are associated with increased morbidity and mortality, length of hospital stay, and need for long-term care. At present, the pathogenesis of PND is not completely clear, and there are various risk factors including surgical trauma and stress mediating systemic inflammation towards neuroinflammation development which causes brain structural and functional changes namely PND. For elderly patients, perioperative neurological monitoring may provide insights into brain function status. Monitoring may also help clinicians identify potential risks which would ultimately allow timely and effective intervention for better perioperative safety and prognosis for elderly patients. In this review, we summarize the risk factors and potential mechanisms of PND, and discuss preliminary evidence regarding application of electroencephalography, functional near-infrared spectroscopy, functional magnetic resonance, and positron emission tomography imaging in monitoring the central nervous system during the postoperative period.

A risk prediction model based on machine learning for postoperative cognitive dysfunction in elderly patients with non-cardiac surgery

BACKGROUND

Early identification of elderly patients undergoing non-cardiac surgery who may be at high risk for postoperative cognitive dysfunction (POCD) can increase the chances of prevention for them, as extra attention and limited resources can be allocated more to these patients.

AIM

We performed this analysis with the aim of developing a simple, clinically useful machine learning (ML) model to predict the probability of POCD at 3 months in elderly patients after non-cardiac surgery.

METHODS

We collected information on patients who received surgical treatment at Nanjing First Hospital from May 2020 to May 2021. We used LASSO regression to select key features and built 5 ML models to assess the risk of POCD at 3 months in elderly patients after non-cardiac surgery. The Shapley Additive exPlanations (SHAP) and methods were introduced to interpret the best model.

RESULTS

A total of 415 patients with non-cardiac surgery were included. The support vector machine (SVM) was the best-performing model of the five ML models. The model showed excellent performance compared to the other four models. The SHAP results showed that VAS score, age, intraoperative hypotension, and preoperative hemoglobin were the four most important features, indicating that the SVM model had good interpretability and reliability. The website of the web-based calculator was https://modricreagan-non-3-pocd-9w2q78.streamlit.app/ .

CONCLUSION

Based on six important perioperative variables, we successfully established a series of ML models for predicting POCD occurrence at 3 months after surgery in elderly non-cardiac patients, with SVM model being the best-performing model. Our models are expected to serve as decision aids for clinicians to monitor screened high-risk patients more closely or to consider further interventions.

Predicative factors and development of a nomogram for postoperative delayed neurocognitive recovery in elderly patients with gastric cancer

BACKGROUND

Delayed neurocognitive recovery (DNR) is a common complication after radical gastrectomy and closely associated with poor outcomes. This study aimed to investigate predictors and develop a nomogram prediction model for DNR.

METHODS

Elderly gastric cancer (GC) patients (≥ 65 years) undergoing elective laparoscopic radical gastrectomy between 2018 and 2022 were prospectively included in this study. DNR was diagnosed according to the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-V, 2013). Independent risk factors for DNR were screened by the multivariate logistic regression analysis. Based on these factors, the nomogram model was established and validated by R.

RESULTS

A total of 312 elderly GC patients were enrolled in the training set, with an incidence of DNR within postoperative 1 month of 23.4% (73/312). Multivariate logistic regression analysis indicated that age (OR: 1.207, 95%CI: 1.113-1.309, P < 0.001), nutritional risk screening 2002 (NRS2002) score (OR: 1.716, 95%CI: 1.211-2.433, P = 0.002), neutrophil-to-lymphocyte ratio (NLR) (OR: 1.976, 95%CI: 1.099-3.552, P = 0.023), albumin-to-fibrinogen ratio (AFR) (OR: 0.774, 95%CI: 0.620-0.966, P = 0.024), and prognostic nutritional index (PNI) (OR: 0.768, 95%CI: 0.706-0.835, P < 0.001) were five independent factors for DNR in elderly GC patients. The constructed nomogram model based on these five factors has a good predictive value for DNR with an area under the curve (AUC) of 0.863.

CONCLUSIONS

In conclusions, the established nomogram model based on age, NRS-2002, NLR, AFR, and PNI has a well predictive value for postoperative DNR in elderly GC patients.

Decreased default mode network functional connectivity with visual processing regions as potential biomarkers for delayed neurocognitive recovery: A resting-state fMRI study and machine-learning analysis

Objectives

The abnormal functional connectivity (FC) pattern of default mode network (DMN) may be key markers for early identification of various cognitive disorders. However, the whole-brain FC changes of DMN in delayed neurocognitive recovery (DNR) are still unclear. Our study was aimed at exploring the whole-brain FC patterns of all regions in DMN and the potential features as biomarkers for the prediction of DNR using machine-learning algorithms.

Methods

Resting-state functional magnetic resonance imaging (fMRI) was conducted before surgery on 74 patients undergoing non-cardiac surgery. Seed-based whole-brain FC with 18 core regions located in the DMN was performed, and FC features that were statistically different between the DNR and non-DNR patients after false discovery correction were extracted. Afterward, based on the extracted FC features, machine-learning algorithms such as support vector machine, logistic regression, decision tree, and random forest were established to recognize DNR. The machine learning experiment procedure mainly included three following steps: feature standardization, parameter adjustment, and performance comparison. Finally, independent testing was conducted to validate the established prediction model. The algorithm performance was evaluated by a permutation test.

Results

We found significantly decreased DMN connectivity with the brain regions involved in visual processing in DNR patients than in non-DNR patients. The best result was obtained from the random forest algorithm based on the 20 decision trees (estimators). The random forest model achieved the accuracy, sensitivity, and specificity of 84.0, 63.1, and 89.5%, respectively. The area under the receiver operating characteristic curve of the classifier reached 86.4%. The feature that contributed the most to the random forest model was the FC between the left retrosplenial cortex/posterior cingulate cortex and left precuneus.

Conclusion

The decreased FC of DMN with regions involved in visual processing might be effective markers for the prediction of DNR and could provide new insights into the neural mechanisms of DNR.

Clinical Trial Registration

: Chinese Clinical Trial Registry, ChiCTR-DCD-15006096.

Association Between the Early Serum Lipid Metabolism Profile and Delayed Neurocognitive Recovery After Cardiopulmonary Bypass in Cardiac Surgical Patients: a Pilot Study

Cardiac surgery with extracorporeal circulation is considered to be one of the surgical types with the highest incidence of delayed neurocognitive recovery (DNR), but the mechanism is unclear. Metabolomics technology can be used to understand the early postoperative metabolic profile and find the relationship between serum metabolites and disease. We performed untargeted analyses of postoperative serum metabolites in all surgical groups, as well as serum metabolites in healthy nonsurgical adults, by using liquid chromatography‒mass spectrometry (LC‒MS). DNR after cardiopulmonary bypass surgery occurred in 35% of surgical patients. Sixty-nine metabolites were found to be associated with DNR. Lipids and lipid-like molecules occupy a total of 55 positions. Lipid metabolism occupies an important position in the serum metabolic profile of DNR patients in the early postoperative period. Phosphatidylinositol (PI), sphingomyelin (SM), and phosphatidylglycerol (PG) appear at the highest frequency. Correlation analysis and receiver operator characteristic curve analysis confirmed PI and SM as potential biomarkers for an increased risk of DNR.

Novel Bedside Dynamic Nomograms to Predict the Probability of Postoperative Cognitive Dysfunction in Elderly Patients Undergoing Noncardiac Surgery: A Retrospective Study

Purpose

Early and accurate prediction of elderly patients at high risk of postoperative cognitive dysfunction (POCD) after non-cardiac surgery will provide favorable evidence for rational perioperative management and long-term postoperative recovery. This study aimed to develop bedside dynamic nomograms to provide accurately an individualized prediction of the risk of POCD at 6-month postoperatively with patients undergoing non-cardiac surgery and to guide clinical decision-making and postoperative management.

Patients and Methods

We retrospectively collected patients undergoing surgical treatment at the Nanjing First Hospital between May 2020 and May 2021. We collected the data on preoperative, intraoperative, and postoperative variables. Clinical and laboratory data on admission and intraoperative variables and postoperative variables were used. We measured the performances of the nomograms using sensitivity, specificity of the receiver operating characteristic (ROC), the area under the ROC curves (AUC), the 10-fold cross-validation, and decision curve analysis (DCA).

Results

POCD was observed in 23 of 415 patients (5.6%) at 6-month postoperatively. The preoperative and postoperative models obtained 91.6% and 94.0% accuracy rates on the data. Compared to the preoperative model, the postoperative model had an area under the receiver characteristic curve (AUC) of 0.973 vs 0.947, corresponding to a specificity of 0.941 vs 0.918 and a sensitivity of 0.913 vs 0.870. The overall performance of the postoperative model was better than the preoperative model.

Conclusion

In this study, we developed novel bedside dynamic nomograms with reasonable clinical utility that can provide individualized prediction of POCD risk at 6-month postoperatively in elderly patients undergoing non-cardiac surgery at different time points based on patient admission and postoperative data. External validations are needed to ensure their value in predicting POCD in elderly patients.