Leveraging Explainable Artificial Intelligence to Identify Key Features required for Differentiating Clinical Neurocognitive Disorder Diagnoses using Toronto Cognitive Assessment

Background: Acute kidney injury (AKI), with an increase in serum creatinine, is a common adverse drug event. Although various clinical studies have investigated whether a combination of two nephrotoxic drugs has an increased risk of AKI using traditional statistical models such as multivariable logistic regression (MLR), the evaluation metrics have not been evaluated despite the fact that traditional statistical models may over-fit the data. The aim of the present study was to detect drug-drug interactions with an increased risk of AKI by interpreting machine-learning models to avoid overfitting.Methods: We developed six machine-learning models trained using electronic medical records: MLR, logistic least absolute shrinkage and selection operator regression (LLR), random forest, extreme gradient boosting (XGB) tree, and two support vector machine models (kernel = linear function and radial basis function). In order to detect drug-drug interactions, the XGB and LLR models that showed good predictive performance were interpreted by SHapley Additive exPlanations (SHAP) and relative excess risk due to interaction (RERI), respectively.Results: Among approximately 2.5 million patients, 65,667 patients were extracted from the electronic medical records, and assigned to case (N = 5,319) and control (N = 60,348) groups. In the XGB model, a combination of loop diuretic and histamine H2 blocker [mean (|SHAP|) = 0.011] was identified as a relatively important risk factor for AKI. The combination of loop diuretic and H2 blocker showed a significant synergistic interaction on an additive scale (RERI 1.289, 95% confidence interval 0.226–5.591) also in the LLR model.Conclusion: The present population-based case-control study using interpretable machine-learning models suggested that although the relative importance of the individual and combined effects of loop diuretics and H2 blockers is lower than that of well-known risk factors such as older age and sex, concomitant use of a loop diuretic and histamine H2 blocker is associated with increased risk of AKI.

Balance impairments are common in patients with Alzheimer's disease (AD) dementia. We sought to determine the stage along the AD spectrum during which balance impairments appear and identify factors associated with a decline in balance function. Our cross-sectional study included 295 participants; 71 were cognitively normal (CN), 96 reported subjective cognitive decline (SCD), 72 had amnestic mild cognitive impairment (MCI), and 56 had AD dementia. The balance and mobility function was assessed using the Timed Up and Go test (TUG) and the One-Leg Standing Test (OLST). Participants in the MCI and AD dementia groups were older than those in the cognitively normal and SCD groups. TUG and OLST test scores were linearly correlated with Mini-Mental Status Examination-Korean Version score (MMSE-KC). TUG score increased with greater AD spectrum severity (all p < 0.001), whereas OLST score showed a precipitous impairment starting in the SCD group (all p < 0.001), even after adjusting for age, sex, MMSE-KC, Geriatric depression scale, and body mass index. Based on subgroup analyses, in females and apolipoprotein E (APOE) ɛ4 carriers, there was significant balance/mobility impairment in the SCD group when compared to the CN group. Our results suggest that balance/mobility is related to cognitive function and that balance/mobility impairment can be observed beginning in the SCD stage. Furthermore, CN females and APOEɛ4 carriers had better balance and mobility when compared to females and APOEɛ4 carriers along the ADD spectrum/with cognitive impairment respectively.

Flooding causes several threats with outcomes which include peril to human and animal life, damage to property, and adversity to agricultural fields. Therefore, flood prediction is of prime importance for reducing loss of life and devastation to property. To model complex nature of hydrologic processes artificial neural network (ANN) tool is effectively being utilized for modelling different nonlinear relationships, and has proved to be an appropriate method for flood prediction. Present study investigated relative accuracy of radial basis function neural network (RBFNN) and support vector machine (SVM) models combined with Firefly Algorithm (FA) in predicting river flood discharge and contrasted with that of regular ANN, RBFNN and SVM models. Monthly river flow data of Silchar and Dholai stations located in Cachar district of Assam, India are utilized for the present study. For assessing model performance, coefficient of determination (R2), mean square error (MSE) and root mean square error (RMSE) were measured. Evaluation of outcomes shows that both RBF-FA (radial basis function - firefly algorithm) and SVM-FA (support vector machine - firefly algorithm) hybrid models give more precise forecasting results than RBFNN, FFBPNN (feed forward back propagation neural network) and SVM models. Yet, it can be observed that SVM-FA model give better forecasting outputs with R2 value 0.9818 than RBF-FA model. Results also reveal that simple SVM model performs marginally better than simple ANN model.

To assess retinal microvascular alterations in individuals with amnestic mild cognitive impairment (MCI) and nonamnestic MCI. One hundred twelve eyes of 59 amnestic MCI participants, 32 eyes of 17 nonamnestic MCI participants, and 111 eyes of 56 controls with normal cognition were included. Optical coherence tomography angiography vessel density and perfusion density in the Early Treatment Diabetic Retinopathy Study 3-mm circle and ring were assessed. Retinal thickness parameters including retinal nerve fiber layer thickness, ganglion cell-inner plexiform layer thickness, central subfield thickness, and subfoveal choroidal thickness were also analyzed. Multivariable generalized estimating equations were used for statistical analysis. Perfusion density in the 3-mm inner ring was significantly lower in amnestic MCI patients when compared with nonamnestic MCI participants (0.29 ± 0.03 vs. 0.34 ± 0.09, P = 0.025) and controls with normal cognition (0.29 ± 0.03 vs. 0.39 ± 0.02, P < 0.001), after adjustment for age and sex as covariates. Vessel density, retinal nerve fiber layer thickness, ganglion cell-inner plexiform layer thickness, central subfield thickness, and subfoveal choroidal thickness did not differ among or between diagnostic groups. Perfusion density was significantly reduced in individuals with amnestic MCI, compared with those with nonamnestic MCI and controls with normal cognition.

Near infrared spectroscopy was used in discrimination of intact bovine teeth in terms of animal sex, diet, tooth type and place of origin. Discriminant Analysis (DA) models were developed and tested using a stratified random 70:30 data split to calibration and test sets. Of the discriminant techniques of Soft Independent Modelling of Class Analogy (SIMCA), Partial Least Squares – DA, Artificial Neural Networks (ANN) and Support Vector Machine (SVM), SVM and PLS-DA models performed best in most instances, with pretreatment choice impacting technique success. For test set prediction of animal sex, an accuracy of 95% was achieved using a PLS-DA model, and similar performance achieved using a SVM model. SVM, ANN and SIMCA models predicted three categories of tooth type (deciduous, permanent unerupted and erupted) with similar accuracies (89, 79 and 82%, respectively), however, permanent unerupted teeth were conflated with permanent erupted teeth. Prediction accuracy for ANN, SVM and PLS-DA models were also similar for discrimination of teeth from animals on a grain versus grass diet (PLS-DA 82%). There was considerable variation between models in prediction of geographic origin. Of the seven locations, two were discriminated with true positive rate around 70% using SVM and ANN models. Potential applications of this technology are discussed.

To construct machine learning prediction model for sepsis-associated encephalopathy (SAE), and analyze the application value of the model on early identification of SAE risk in elderly septic patients. Patients aged over 60 years with a primary diagnosis of sepsis admitted to intensive care unit (ICU) from 2008 to 2023 were selected from Medical Information Mart for Intensive Care-IV 2.2 (MIMIC-IV 2.2). Demographic variables, disease severity scores, comorbidities, interventions, laboratory indicators, and hospitalization details were collected. Key factors associated with SAE were identified using univariate Logistic regression analysis. The data were randomly divided into training and validation sets in a 7 : 3 ratio. Multivariable Logistic regression analysis was conducted in the training set and visualized using a nomogram model for prediction of SAE. The discrimination of the model was evaluated in the validation set using the receiver operator characteristic curve (ROC curve), and its calibration was assessed using calibration curve. Furthermore, multiple machine learning algorithms, including multi-layer perceptron (MLP), support vector machine (SVM), naive bayes (NB), gradient boosting machine (GBM), random forest (RF), and extreme gradient boosting (XGB), were constructed in the training set. Their predictive performance was subsequently evaluated on the validation set. Taking the XGB model as an example, the interpretability of the model through the SHapley Additive exPlanations (SHAP) algorithm was enhanced to identify the key predictive factors and their contributions. A total of 2 204 septic patients were finally enrolled, of whom 840 developed SAE (38.1%). A total of 21 variables associated with SAE were screened through univariate Logistic regression analysis. Multivariable Logistic regression analysis showed that endotracheal intubation [odds ratio (OR) = 0.40, 95% confidence interval (95%CI) was 0.19-0.88, P < 0.001], oxygen therapy (OR = 0.76, 95%CI was 0.53-0.95, P = 0.023), tracheotomy (OR = 0.20, 95%CI was 0.07-0.53, P < 0.001), continuous renal replacement therapy (CRRT; OR = 0.32, 95%CI was 0.15-0.70, P < 0.001), cerebrovascular disease (OR = 0.31, 95%CI was 0.16-0.60, P < 0.001), rheumatic disease (OR = 0.44, 95%CI was 0.19-0.99, P < 0.001), male (OR = 0.68, 95%CI was 0.54-0.86, P = 0.001), and maximum anion gap (AG; OR = 0.95, 95%CI was 0.93-0.97, P < 0.001) were associated with an decreased probability of SAE, and age (OR = 1.05, 95%CI was 1.03-1.06, P < 0.001), acute physiology score III (APSIII; OR = 1.02, 95%CI was 1.01-1.02, P < 0.001), Oxford acute severity of illness score (OASIS; OR = 1.04, 95%CI was 1.03-1.06, P < 0.001), and length of hospital stay (OR = 1.01, 95%CI was 1.01-1.02, P < 0.001) were associated with an increased probability of SAE. A nomogram model was constructed based on these variables. In the validation set, ROC curve analysis showed that the model achieved an area under the ROC curve (AUC) of 0.723, and the calibration curve showed good consistency between the predicted probability of the model and the observed probability. Among the machine learning algorithms, including MLP, SVM, NB, GBM, RF, and XGB, the SVM model and RF model demonstrated relatively good predictive performance, with AUC of 0.748 and 0.739, respectively, and the sensitivity was both exceeding 85%. The predictive performance of the XGB model was explained through SHAP analysis, and the results indicated that APSIII score (SHAP value was 0.871), age (SHAP value was 0.521), and OASIS score (SHAP value was 0.443) were important factors affecting the predictive performance of the model. The machine learning-based SAE prediction model exhibits good predictive capability and holds significant application value for the early identification of SAE risk in elderly septic patients.

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. We performed this analysis with the aim of developing a simple, clinically useful machine learning (ML) model to predict the probability of POCD at 3months in elderly patients after non-cardiac surgery. 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 3months in elderly patients after non-cardiac surgery. The Shapley Additive exPlanations (SHAP) and methods were introduced to interpret the best model. 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/ . Based on six important perioperative variables, we successfully established a series of ML models for predicting POCD occurrence at 3months 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.

Tremendous advancement has been made in biomarker assessment, molecular neuroimaging and disease modifying therapy in neurodegenerative cognitive disorders. China is lacking clinical cohorts that deeply profile patients from memory clinic that integrate both comprehensive clinical assessment and multi-modal biomarker examinations. The Shanghai Memory Study (SMS) is an ongoing prospective hospital-based cohort study. Participants underwent extensive clinical and neuropsychological assessments, genotyping, multimodal magnetic resonance imaging (MRI), positron emission tomography (PET) imaging, and various biospecimen collection. In this article, the rationale, aims, study design, data collection, and future directions of SMS were summarized. Preliminary analyses including plasma biomarker assay using the Simoa platform, molecular neuroimaging of Aβ PET and tau PET were performed across different diagnostic groups. A total of 2,001 participants with cognitive complaints were enrolled, including 115 with subjective cognitive decline (SCD), 938 with mild cognitive impairment (MCI), and 948 with dementia. The proportions of individuals with A+/T+ in PET scans were 15.8% in SCD cases, 51.2% in MCI cases, and 100% in AD dementia. And the proportions of individuals with A+/T- in PET scans were 5.3% in SCD cases and 3.3% in MCI cases. In the analysis with plasma biomarkers, as the CDR score increased, Aβ40, Aβ42, and Aβ42/Aβ40 ratio exhibited a decreasing trend, whereas t-tau, NfL, and p-tau181 showed an increasing trend. In a subcohort of 251 patients with amnestic MCI (aMCI), during a median follow-up of 4.7 years, 88 (35.1%) progressed to AD dementia, 8 developed non-AD dementia. Individuals with low Aβ42/Aβ40 and high p-tau181 at baseline demonstrated the highest risk of developing AD (hazard ratio=4.83, 95% CI 2.37-9.86). By offering an integrated framework for investigating cognitive disorders, SMS will facilitate the exploration of AD pathogenesis and deepen the understanding of cognitive disorders.

Fatal myocardial infarction (MI) is one of the most common types of cardiovascular diseases that often presents in the emergency department. The prediction of death caused by myocardial infarction within 10-years follow-up is addressed in this study, using comorbidities, daily habits, clinical and laboratory data in binary and continuous data respectively. The used data are included in a cohort of the Ludwigshafen Risk and Cardiovascular Health (LURIC) study. The target feature, namely death caused by MI, contained 106 deceased patients and 2,321 alive patients in the final used dataset. The analysis was based on machine learning models (ML), such as support vector machine (SVM), light gradient-boosting machine (LGBM), Random Forest (RF), Decision Tree (DT). Their performance was estimated by Area Under Receiver Operating Characteristic Curve (AUC), Sensitivity, Specificity, Precision and Accuracy. Results show that LGBM was the most suitable of the aforementioned models to predict death caused by myocardial infarction within 10-years follow-up, achieving area under the curve (AUC) value equal to 77.03 %, accuracy 69.42 %, sensitivity 69.75 %, specificity 69.40% and precision 53.76 %. In addition, explainable artificial intelligence (xAI) was utilized and especially SHapley Additive exPlanations (SHAP) was the selected method. SHAP was utilized in order to shed light on the results, applying the LGBM model as the best predictive model. The provided SHAP plots contribute to the interpretation of how each independent feature aid to the final prediction.

Purpose This study aimed to investigate multivariate regional patterns for schizophrenia (SZ) classification, sex differences, and brain age by utilizing structural MRI, demographics, and explainable artificial intelligence (AI). Methods Various AI models were employed, and the outperforming model was identified for SZ classification, sex differences, and brain age predictions. For the SZ and sex classification tasks, support vector classifier (SVC), k-nearest neighbor (KNN), and deep learning neural network (DL) models were compared. In the case of regression-based brain age prediction, Lasso regression (LR), Ridge regression (RR), support vector regression (SVR), and DL models were compared. For each regression or classification task, the optimal model was further integrated with the Shapley additive explanations (SHAP), and significant multivariate brain regional patterns were identified. Results Our results demonstrated that the DL model outperformed other models in SZ classification, sex differences, and brain age predictions. We then integrated outperforming DL model with SHAP, and this integrated DL-SHAP model was used to identify the individualized multivariate regional patterns associated with each prediction. Using the DL-SHAP approach, we found that individuals with SZ had anatomical changes, particularly in the left pallidum, left posterior insula, left hippocampus, and left putamen regions, and such changes associated with SZ were different between female and male patients. Finally, we further applied the DL-SHAP method to brain age prediction and suggested important brain regions related to aging in health controls (HC) and SZ processes. Conclusion This study systematically utilized predictive modeling and novel explainable AI approaches and identified the complex multivariate brain regions involved with SZ classification, sex differences, and brain aging, thereby building a deeper understanding of neurobiological mechanisms involved in the disease, offering new insights into future SZ diagnosis and treatments, and laying the foundation for the development of precision medicine.

BackgroundVisual rating scales are still the most popular tools in assessing atrophy degrees of whole brain and lobes. However, the false negative rate of the previous cutoff score of visual rating scales was relatively high for detecting dementia of Alzheimer’s type (DAT). This study aimed to evaluate the diagnostic value of new cutoffs of visual rating scales on magnetic resonance imaging for discriminating DAT in a Chinese population.MethodsOut of 585 enrolled subjects, 296 participants were included and diagnosed as normal cognition (NC)(n = 87), 138 diagnosed as amnestic mild cognitive impairment (aMCI), and 71 as dementia of Alzheimer’s type (DAT). Receiver operating characteristic (ROC) curve analyses were used to calculate the diagnostic value of visual rating sales (including medial temporal atrophy (MTA), posterior atrophy rating scale (PA),global cortical atrophy scale (GCA) and medial temporal-lobe atrophy index (MTAi))for detecting NC from DAT .ResultsScores of MTA correlated to age and Mini-mental state examination score. When used to detect DAT from NC, the MTA showed highest diagnostic value than other scales, and when the cutoff score of 1.5 of MTA scale, it obtained an optimal sensitivity (84.5%) and specificity (79.1%) respectively, with a 15.5% of false negative rate. Cutoff scores and diagnostic values were calculated stratified by age. For the age ranges 50–64, 65–74, 75–84 years, the following cut-offs of MTA should be used, ≥1.0(sensitivity and specificity were 92.3 and 68.4%), ≥1.5(sensitivity and specificity were 90.4 and 85.2%), ≥ 2.0(sensitivity and specificity were 70.8 and 82.3%) respectively. All of the scales showed relatively lower diagnostic values for discriminating aMCI from NC.ConclusionsThe new age-based MTA cutoff showed better diagnostic accuracy for detecting DAT than previous standard, the list of practical cut-offs proposed here might be useful in clinical practice.

This study presents a comprehensive methodology for developing a Fault Detection and Diagnosis (FDD) system for an industrial distillation column using advanced machine learning algorithms. Steady state and dynamic simulations in Aspen Plus® generate extensive datasets under normal and faulty conditions. Feature engineering, using the Minimum Redundancy Maximum Relevance (MRMR) algorithm, selects the most relevant features for fault detection. Various machine learning models, including Decision Trees, Support Vector Machines, k-nearest Neighbours, and Neural Networks, were trained and evaluated based on performance metrics such as accuracy, recall, precision, and F1 score. The top models were integrated into a stacked classifier system with a voting mechanism to enhance fault detection reliability. Explainable Artificial Intelligence (XAI) techniques, such as Local Interpretable Model-agnostic Explanations (LIME) and Shapley Additive Explanations (SHAP), were incorporated to improve model interpretability, allowing engineers to understand and validate the FDD system's decision-making process. Simulation results confirm that the proposed methodology accurately identifies and classifies faults. By integrating dynamic simulations, advanced machine learning, and XAI techniques, a robust and scalable solution is achieved for fault detection in distillation columns, improving operational reliability, safety, and reducing downtime. Future work could extend this approach to other industrial processes and explore additional machine learning algorithms to further enhance performance.

The accurate detection and assessment of stress play a pivotal role in enhancing individual well-being and healthcare outcomes. Traditional methods of stress detection often grapple with limitations in accuracy and scalability. With the advent of machine learning (ML), the potential to revolutionize stress detection has emerged. This paper presents a comprehensive study on the application of ML algorithms for stress detection, with a focus on physiological and behavioral data analysis. Central to our approach is the integration of SHapley Additive exPlanations (SHAP), an Explainable Artificial Intelligence (XAI) technique, to interpret ML models. SHAP provides a novel lens to understand the impact of individual features in the complex decision-making processes of ML models, thereby enhancing the transparency and reliability of stress predictions. We demonstrate how SHAP not only aids in elucidating model decisions but also contributes to refining the models for greater accuracy. Our results highlight the effectiveness of ML in detecting stress and the pivotal role of XAI in making these models more interpretable and trustworthy. This study underscores the synergy between advanced ML techniques and XAI, paving the way for more nuanced and reliable stress detection methodologies that are essential in diverse settings, from healthcare to workplace environments.

This study aims to develop and evaluate an obesity classification model using an Artificial Neural Network (ANN) combined with Explainable Artificial Intelligence (XAI) techniques based on SHAP (SHapley Additive exPlanations). The model was trained and tested using two different optimizers, Adaptive Moment Estimation (Adam) and Stochastic Gradient Descent (SGD), across multiple train-test ratios and epoch variations. The experimental results indicate that the Adam optimizer consistently outperformed SGD in terms of accuracy, loss value, and stability of evaluation metrics. The best performance was achieved with a 90:10 train-test ratio at 100 epochs, yielding an accuracy of 94.74%, a loss of 0.1899, precision, recall, and an f1-score of 0.95. To improve interpretability, SHAP was applied to identify the most influential features in the classification process. The analysis revealed that features such as Weight, Height, Gender, and Age significantly contribute to the model's predictions. Based on the SHAP interpretation, feature selection was conducted using the top nine features with the highest SHAP values. Retraining the ANN with these selected features resulted in improved performance, achieving 98.56% accuracy, a loss of 0.0638, and a precision, recall, and F1-score of 0.99 . These findings demonstrate that integrating XAI with ANN not only enhances transparency and interpretability but also boosts classification performance and computational efficiency. This approach shows strong potential for supporting decision-making in healthcare, particularly for early detection and intervention in cases related to obesity.

Objective To identify the differences inneuropsychological characteristics between amnestic(AMCI)and vascular mild cognitive impairment(VMCI). Methods Totally 297 old community residents with mild cognitive impairment(MCI)were divided into amnestic MCI(AMCI)and vascular MCI(VMCI)subgroup from Guangzhou MCI prevalence survey.The elderly with MCI were interviewed and tested with the Chinese version of Montreal Cognitive Assessment(MoCA), the Mini-Mental state examination(MMSE), Auditory Verbal Learning Test(AVLT), the Clinical Dementia Rating scale(CDR), Functional Activity Questionnaire(FAQ), the Modified Hachinski Ischemic Scale(M-HIS), Center for Epidemiologic Studies(CES-DC)to evaluate neuropsychological characteristics. Results AMCI versus VMCI group showed that the total scores of MoCA were(9.63±5.17 vs.9.98±6.02), total scores of MMSE were(16.90±4.84 vs.16.90±6.19), AVLT immediate memory was(2.35±1.39 vs.2.91±1.84), AVLT delayed recall was(2.23±2.09 vs.2.47±2.20), AVLT delayed recognition was(7.33±3.98 vs.6.85±4.02)and total scores of CDR(0.5 vs.0.5), with no differences between the 2 groups(all P>0.05). Based on MoCA survey, AMCI versus VMCI group showed statistically significant differences(all P<0.05)in parameters of visual space and execution(0.71±1.02 vs.0.92±1.26), language function(0.34±0.56 vs.0.50±0.80)and abstract thinking(0.25±0.49 vs.0.15±0.43), but based on MMSE survey, no difference was found in the various cognitive domains between the two groups.The AMCI versus VMCI group showed statistically significant differences(all P<0.05)in parameters of CES-DC scale(1.75±4.27 vs.2.76±6.72), FAQ scale(4.42±4.66 vs.8.71±7.03), M-HIS scale(0.40±0.64 vs.7.59±3.53). Conclusions There is no significant difference in general cognitive impairment between AMCI and VMCI, but the visual space and execution, language function are more impaired in AMCI than VMCI, and the abstract thinking, social function are more impaired with more depressive symptoms in VMCI than in AMCI. Key words: Cognition disorders; Neuropsychological tests; Amnesia; Blood vessels