Search

Background: This study aims to assess the efficacy of combining automated machine learning (AutoML) and explainable artificial intelligence (XAI) in identifying metabolomic biomarkers that can differentiate between hepatocellular carcinoma (HCC) and liver cirrhosis in patients with hepatitis C virus (HCV) infection. Methods: We investigated publicly accessible data encompassing HCC patients and cirrhotic controls. The TPOT tool, which is an AutoML tool, was used to optimize the preparation of features and data, as well as to select the most suitable machine learning model. The TreeSHAP approach, which is a type of XAI, was used to interpret the model by assessing each metabolite’s individual contribution to the categorization process. Results: TPOT had superior performance in distinguishing between HCC and cirrhosis compared to other AutoML approaches AutoSKlearn and H2O AutoML, in addition to traditional machine learning models such as random forest, support vector machine, and k-nearest neighbor. The TPOT technique attained an AUC value of 0.81, showcasing superior accuracy, sensitivity, and specificity in comparison to the other models. Key metabolites, including L-valine, glycine, and DL-isoleucine, were identified as essential by TPOT and subsequently verified by TreeSHAP analysis. TreeSHAP provided a comprehensive explanation of the contribution of these metabolites to the model’s predictions, thereby increasing the interpretability and dependability of the results. This thorough assessment highlights the strength and reliability of the AutoML framework in the development of clinical biomarkers. Conclusions: This study shows that AutoML and XAI can be used together to create metabolomic biomarkers that are specific to HCC. The exceptional performance of TPOT in comparison to traditional models highlights its capacity to identify biomarkers. Furthermore, TreeSHAP boosted model transparency by highlighting the relevance of certain metabolites. This comprehensive method has the potential to enhance the identification of biomarkers and generate precise, easily understandable, AI-driven solutions for diagnosing HCC.

Background: The current study’s objective is to evaluate the molecular genetic mechanisms influencing the biological behavior of hepatocellular carcinoma (HCC) by analyzing the transcriptomic and epigenetic signatures of the tumors. Methods: Transcriptomic data were downloaded from the NCBI GEO database. We investigated the expression differences between the GSE46444 (48 cirrhotic tissues versus 88 HCC tissues) and GSE63898 (168 cirrhotic tissues versus 228 HCC tissues) data sets using GEO2R. Differentially expressed genes were evaluated using GO and KEGG metabolic pathway analysis websites. Whole genome bisulfite sequencing (WGBS) and Methylated DNA Immunoprecipitation Sequencing (MeDIP-Seq) data sets (26 HCC tissues versus 26 adjacent non-tumoral tissues) were also downloaded from the NCBI SRA database. These data sets were analyzed using Bismark and QSEA, respectively. The methylation differences between the groups were assessed using functional enrichment analysis. Results: In the GSE46444 data set, 80 genes were upregulated, and 315 genes were downregulated in the tumor tissue (HCC tissue) compared to the non-tumor cirrhotic tissue. In the GSE63898 data set, 1261 genes were upregulated, and 458 genes were downregulated in the cirrhotic tissue compared to the tumor tissues. WGBS revealed that 20 protein-coding loci were hypermethylated. while the hypomethylated regions were non-protein-coding. The methylated residues of the tumor tissue, non-tumorous cirrhotic tissue, and healthy tissue were comparable. MeDIP-Seq, conducted on tumoral and non-tumoral tissues, identified hypermethylated or hypomethylated areas as protein-coding regions. The functional enrichment analysis indicated that these genes were related to pathways including peroxisome, focal adhesion, mTOR, RAP1, Phospholipase D, Ras, and PI3K/AKT signal transduction. Conclusions: The investigation of transcriptomic and epigenetic mechanisms identified several genes significant in the biological behavior of HCC. These genes present potential targets for the development of targeted therapy.

Background: Diabetic retinopathy (DR) is a prevalent microvascular complication of diabetes mellitus, and early detection is crucial for effective management. Metabolomics profiling has emerged as a promising approach for identifying potential biomarkers associated with DR progression. This study aimed to develop a hybrid explainable artificial intelligence (XAI) model for targeted metabolomics analysis of patients with DR, utilizing a focused approach to identify specific metabolites exhibiting varying concentrations among individuals without DR (NDR), those with non-proliferative DR (NPDR), and individuals with proliferative DR (PDR) who have type 2 diabetes mellitus (T2DM). Methods: A total of 317 T2DM patients, including 143 NDR, 123 NPDR, and 51 PDR cases, were included in the study. Serum samples underwent targeted metabolomics analysis using liquid chromatography and mass spectrometry. Several machine learning models, including Support Vector Machines (SVC), Random Forest (RF), Decision Tree (DT), Logistic Regression (LR), and Multilayer Perceptrons (MLP), were implemented as solo models and in a two-stage ensemble hybrid approach. The models were trained and validated using 10-fold cross-validation. SHapley Additive exPlanations (SHAP) were employed to interpret the contributions of each feature to the model predictions. Statistical analyses were conducted using the Shapiro–Wilk test for normality, the Kruskal–Wallis H test for group differences, and the Mann–Whitney U test with Bonferroni correction for post-hoc comparisons. Results: The hybrid SVC + MLP model achieved the highest performance, with an accuracy of 89.58%, a precision of 87.18%, an F1-score of 88.20%, and an F-beta score of 87.55%. SHAP analysis revealed that glucose, glycine, and age were consistently important features across all DR classes, while creatinine and various phosphatidylcholines exhibited higher importance in the PDR class, suggesting their potential as biomarkers for severe DR. Conclusion: The hybrid XAI models, particularly the SVC + MLP ensemble, demonstrated superior performance in predicting DR progression compared to solo models. The application of SHAP facilitates the interpretation of feature importance, providing valuable insights into the metabolic and physiological markers associated with different stages of DR. These findings highlight the potential of hybrid XAI models combined with explainable techniques for early detection, targeted interventions, and personalized treatment strategies in DR management.

The quick and large development in the accumulation of medical data provides broad potential for the application of artificial intelligence technologies [...]

This study aims to develop an interpretable prediction model based on explainable artificial intelligence to predict bacterial sepsis and discover important biomarkers. A total of 1572 adult patients, 560 of whom were sepsis positive and 1012 of whom were negative, who were admitted to the emergency department with suspicion of sepsis, were examined. We investigated the performance characteristics of sepsis biomarkers alone and in combination for confirmed sepsis diagnosis using Sepsis-3 criteria. Three different tree-based algorithms—Extreme Gradient Boosting (XGBoost), Light Gradient Boosting Machine (LightGBM), Adaptive Boosting (AdaBoost)—were used for sepsis prediction, and after examining comprehensive performance metrics, descriptions of the optimal model were obtained with the SHAP method. The XGBoost model achieved accuracy of 0.898 (0.868–0.929) and area under the ROC curve (AUC) of 0.940 (0.898–0.980) with a 95% confidence interval. The five biomarkers for predicting sepsis were age, respiratory rate, oxygen saturation, procalcitonin, and positive blood culture. SHAP results revealed that older age, higher respiratory rate, procalcitonin, neutrophil–lymphocyte count ratio, C-reactive protein, plaque, leukocyte particle concentration, as well as lower oxygen saturation, systolic blood pressure, and hemoglobin levels increased the risk of sepsis. As a result, the Explainable Artificial Intelligence (XAI)-based prediction model can guide clinicians in the early diagnosis and treatment of sepsis, providing more effective sepsis management and potentially reducing mortality rates and medical costs.

Background: Myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) is a complex and debilitating illness with a significant global prevalence, affecting over 65 million individuals. It affects various systems, including the immune, neurological, gastrointestinal, and circulatory systems. Studies have shown abnormalities in immune cell types, increased inflammatory cytokines, and brain abnormalities. Further research is needed to identify consistent biomarkers and develop targeted therapies. This study uses explainable artificial intelligence and machine learning techniques to identify discriminative metabolites for ME/CFS. Material and Methods: The model investigates a metabolomics dataset of CFS patients and healthy controls, including 26 healthy controls and 26 ME/CFS patients aged 22–72. The dataset encapsulated 768 metabolites into nine metabolic super-pathways: amino acids, carbohydrates, cofactors, vitamins, energy, lipids, nucleotides, peptides, and xenobiotics. Random forest methods together with other classifiers were applied to the data to classify individuals as ME/CFS patients and healthy individuals. The classification learning algorithms’ performance in the validation step was evaluated using a variety of methods, including the traditional hold-out validation method, as well as the more modern cross-validation and bootstrap methods. Explainable artificial intelligence approaches were applied to clinically explain the optimum model’s prediction decisions. Results: The metabolomics of C-glycosyltryptophan, oleoylcholine, cortisone, and 3-hydroxydecanoate were determined to be crucial for ME/CFS diagnosis. The random forest model outperformed the other classifiers in ME/CFS prediction using the 1000-iteration bootstrapping method, achieving 98% accuracy, precision, recall, F1 score, 0.01 Brier score, and 99% AUC. According to the obtained results, the bootstrap validation approach demonstrated the highest classification outcomes. Conclusion: The proposed model accurately classifies ME/CFS patients based on the selected biomarker candidate metabolites. It offers a clear interpretation of risk estimation for ME/CFS, aiding physicians in comprehending the significance of key metabolomic features within the model.

Aim: Method: This research presents a model combining machine learning (ML) techniques and eXplainable artificial intelligence (XAI) to predict breast cancer (BC) metastasis and reveal important genomic biomarkers in metastasis patients. Method: A total of 98 primary BC samples was analyzed, comprising 34 samples from patients who developed distant metastases within a 5-year follow-up period and 44 samples from patients who remained disease-free for at least 5 years after diagnosis. Genomic data were then subjected to biostatistical analysis, followed by the application of the elastic net feature selection method. This technique identified a restricted number of genomic biomarkers associated with BC metastasis. A light gradient boosting machine (LightGBM), categorical boosting (CatBoost), Extreme Gradient Boosting (XGBoost), Gradient Boosting Trees (GBT), and Ada boosting (AdaBoost) algorithms were utilized for prediction. To assess the models’ predictive abilities, the accuracy, F1 score, precision, recall, area under the ROC curve (AUC), and Brier score were calculated as performance evaluation metrics. To promote interpretability and overcome the “black box” problem of ML models, a SHapley Additive exPlanations (SHAP) method was employed. Results: The LightGBM model outperformed other models, yielding remarkable accuracy of 96% and an AUC of 99.3%. In addition to biostatistical evaluation, in XAI-based SHAP results, increased expression levels of TSPYL5, ATP5E, CA9, NUP210, SLC37A1, ARIH1, PSMD7, UBQLN1, PRAME, and UBE2T (p ≤ 0.05) were found to be associated with an increased incidence of BC metastasis. Finally, decreased levels of expression of CACTIN, TGFB3, SCUBE2, ARL4D, OR1F1, ALDH4A1, PHF1, and CROCC (p ≤ 0.05) genes were also determined to increase the risk of metastasis in BC. Conclusion: The findings of this study may prevent disease progression and metastases and potentially improve clinical outcomes by recommending customized treatment approaches for BC patients.

Obesity is the excessive accumulation of adipose tissue in the body that leads to health risks. The study aimed to classify obesity levels using a tree-based machine-learning approach considering physical activity and nutritional habits. Methods: The current study employed an observational design, collecting data from a public dataset via a web-based survey to assess eating habits and physical activity levels. The data included gender, age, height, weight, family history of being overweight, dietary patterns, physical activity frequency, and more. Data preprocessing involved addressing class imbalance using Synthetic Minority Over-sampling TEchnique-Nominal Continuous (SMOTE-NC) and feature selection using Recursive Feature Elimination (RFE). Three classification algorithms (logistic regression (LR), random forest (RF), and Extreme Gradient Boosting (XGBoost)) were used for obesity level prediction, and Bayesian optimization was employed for hyperparameter tuning. The performance of different models was evaluated using metrics such as accuracy, recall, precision, F1-score, area under the curve (AUC), and precision–recall curve. The LR model showed the best performance across most metrics, followed by RF and XGBoost. Feature selection improved the performance of LR and RF models, while XGBoost’s performance was mixed. The study contributes to the understanding of obesity classification using machine-learning techniques based on physical activity and nutritional habits. The LR model demonstrated the most robust performance, and feature selection was shown to enhance model efficiency. The findings underscore the importance of considering both physical activity and nutritional habits in addressing the obesity epidemic.

Background: The first aim of this study is to perform bioinformatic analysis of lncRNAs obtained from liver tissue samples from rats treated with cisplatin hepatotoxicity and without pathology. Another aim is to identify possible biomarkers for the diagnosis/early diagnosis of hepatotoxicity by modeling the data obtained from bioinformatics analysis with ensemble learning methods. Methods: In the study, 20 female Sprague-Dawley rats were divided into a control group and a hepatotoxicity group. Liver samples were taken from rats, and transcriptomic and histopathological analyses were performed. The dataset achieved from the transcriptomic analysis was modeled with ensemble learning methods (stacking, bagging, and boosting). Modeling results were evaluated with accuracy (Acc), balanced accuracy (B-Acc), sensitivity (Se), specificity (Sp), positive predictive value (Ppv), negative predictive value (Npv), and F1 score performance metrics. As a result of the modeling, lncRNAs that could be biomarkers were evaluated with variable importance values. Results: According to histopathological and immunohistochemical analyses, a significant increase was observed in the sinusoidal dilatation and Hsp60 immunoreactivity values in the hepatotoxicity group compared to the control group (p < 0.0001). According to the results of the bioinformatics analysis, 589 lncRNAs showed different expressions in the groups. The stacking model had the best classification performance among the applied ensemble learning models. The Acc, B-Acc, Se, Sp, Ppv, Npv, and F1-score values obtained from this model were 90%, 90%, 80%, 100%, 100%, 83.3%, and 88.9%, respectively. lncRNAs with id rna-XR_005492522.1, rna-XR_005492536.1, and rna-XR_005505831.1 with the highest three values according to the variable importance obtained as a result of stacking modeling can be used as predictive biomarker candidates for hepatotoxicity. Conclusions: Among the ensemble algorithms, the stacking technique yielded higher performance results as compared to the bagging and boosting methods on the transcriptomic data. More comprehensive studies can support the possible biomarkers determined due to the research and the decisive results for the diagnosis of drug-induced hepatotoxicity.

Background: The primary aim of this study was to compare liver transplant (LT) recipients with and without hepatocellular carcinoma (HCC) in terms of COVID-19-related depression, anxiety, and stress. Method: A total of 504 LT recipients with (HCC group; n = 252) and without HCC (non-HCC group; n = 252) were included in the present case–control study. Depression Anxiety Stress Scales (DASS-21) and Coronavirus Anxiety Scale (CAS) were used to evaluate the depression, stress, and anxiety levels of LT patients. DASS-21 total and CAS-SF scores were determined as the primary outcomes of the study. Poisson regression and negative binomial regression models were used to predict the DASS and CAS scores. The incidence rate ratio (IRR) was used as a coefficient. Both groups were also compared in terms of awareness of the COVID-19 vaccine. Results: Poisson regression and negative binomial regression analyses for DASS-21 total and CAS-SF scales showed that the negative binomial regression method was the appropriate model for both scales. According to this model, it was determined that the following independent variables increased the DASS-21 total score: non-HCC (IRR: 1.26; p = 0.031), female gender (IRR: 1.29; p = 0.036), presence of chronic disease (IRR: 1.65; p < 0.001), exposure to COVID-19 (IRR: 1.63; p < 0.001), and nonvaccination (IRR: 1.50; p = 0.002). On the other hand, it was determined that the following independent variables increased the CAS score: female gender (IRR:1.75; p = 0.014) and exposure to COVID-19 (IRR: 1.51; p = 0.048). Significant differences were found between the HCC and non-HCC groups in terms of median DASS-21 total (p < 0.001) and CAS-SF (p = 0.002) scores. Cronbach’s alpha internal consistency coefficients of DASS-21 total and CAS-SF scales were calculated to be 0.823 and 0.783, respectively. Conclusion: This study showed that the variables including patients without HCC, female gender, having a chronic disease, being exposed to COVID-19, and not being vaccinated against COVID-19 increased anxiety, depression, and stress. High internal consistency coefficients obtained from both scales indicate that these results are reliable.

Background: The primary aim of this study was to create a machine learning (ML) model that can predict perforated and nonperforated acute appendicitis (AAp) with high accuracy and to demonstrate the clinical interpretability of the model with explainable artificial intelligence (XAI). Method: A total of 1797 patients who underwent appendectomy with a preliminary diagnosis of AAp between May 2009 and March 2022 were included in the study. Considering the histopathological examination, the patients were divided into two groups as AAp (n = 1465) and non-AAp (NA; n = 332); the non-AAp group is also referred to as negative appendectomy. Subsequently, patients confirmed to have AAp were divided into two subgroups: nonperforated (n = 1161) and perforated AAp (n = 304). The missing values in the data set were assigned using the Random Forest method. The Boruta variable selection method was used to identify the most important variables associated with AAp and perforated AAp. The class imbalance problem in the data set was resolved by the SMOTE method. The CatBoost model was used to classify AAp and non-AAp patients and perforated and nonperforated AAp patients. The performance of the model in the holdout test set was evaluated with accuracy, F1- score, sensitivity, specificity, and area under the receiver operator curve (AUC). The SHAP method, which is one of the XAI methods, was used to interpret the model results. Results: The CatBoost model could distinguish AAp patients from non-AAp individuals with an accuracy of 88.2% (85.6–90.8%), while distinguishing perforated AAp patients from nonperforated AAp individuals with an accuracy of 92% (89.6–94.5%). According to the results of the SHAP method applied to the CatBoost model, it was observed that high total bilirubin, WBC, Netrophil, WLR, NLR, CRP, and WNR values, and low PNR, PDW, and MCV values increased the prediction of AAp biochemically. On the other hand, high CRP, Age, Total Bilirubin, PLT, RDW, WBC, MCV, WLR, NLR, and Neutrophil values, and low Lymphocyte, PDW, MPV, and PNR values were observed to increase the prediction of perforated AAp. Conclusion: For the first time in the literature, a new approach combining ML and XAI methods was tried to predict AAp and perforated AAp, and both clinical conditions were predicted with high accuracy. This new approach proved successful in showing how well which demographic and biochemical parameters could explain the current clinical situation in predicting AAp and perforated AAp.