A Prediction Model for Cognitive Impairment Risk in Colorectal Cancer after Chemotherapy Treatment

Predicting adverse drug event using machine learning based on electronic health records: a systematic review and meta-analysis

Introduction

Adverse drug events (ADEs) pose a significant challenge in current clinical practice. Machine learning (ML) has been increasingly used to predict specific ADEs using electronic health record (EHR) data. This systematic review provides a comprehensive overview of the application of ML in predicting specific ADEs based on EHR data.

Methods

A systematic search of PubMed, Web of Science, Embase, and IEEE Xplore was conducted to identify relevant articles published from the inception to 20 May 2024. Studies that developed ML models for predicting specific ADEs or ADEs associated with particular drugs were included using EHR data.

Results

A total of 59 studies met the inclusion criteria, covering 15 drugs and 15 ADEs. In total, 38 machine learning algorithms were reported, with random forest (RF) being the most frequently used, followed by support vector machine (SVM), eXtreme gradient boosting (XGBoost), decision tree (DT), and light gradient boosting machine (LightGBM). The performance of the ML models was generally strong, with an average area under the curve (AUC) of 76.68% ± 10.73, accuracy of 76.00% ± 11.26, precision of 60.13% ± 24.81, sensitivity of 62.35% ± 20.19, specificity of 75.13% ± 16.60, and an F1 score of 52.60% ± 21.10. The combined sensitivity, specificity, diagnostic odds ratio (DOR), and AUC from the summary receiver operating characteristic (SROC) curve using a random effects model were 0.65 (95% CI: 0.65-0.66), 0.89 (95% CI: 0.89-0.90), 12.11 (95% CI: 8.17-17.95), and 0.8069, respectively. The risk factors associated with different drugs and ADEs varied.

Discussion

Future research should focus on improving standardization, conducting multicenter studies that incorporate diverse data types, and evaluating the impact of artificial intelligence predictive models in real-world clinical settings.

Systematic Review Registration

https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42024565842, identifier CRD42024565842.

Machine Learning Approaches to Predict Symptoms in People With Cancer: Systematic Review

BACKGROUND

People with cancer frequently experience severe and distressing symptoms associated with cancer and its treatments. Predicting symptoms in patients with cancer continues to be a significant challenge for both clinicians and researchers. The rapid evolution of machine learning (ML) highlights the need for a current systematic review to improve cancer symptom prediction.

OBJECTIVE

This systematic review aims to synthesize the literature that has used ML algorithms to predict the development of cancer symptoms and to identify the predictors of these symptoms. This is essential for integrating new developments and identifying gaps in existing literature.

METHODS

We conducted this systematic review in accordance with the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) checklist. We conducted a systematic search of CINAHL, Embase, and PubMed for English records published from 1984 to August 11, 2023, using the following search terms: cancer, neoplasm, specific symptoms, neural networks, machine learning, specific algorithm names, and deep learning. All records that met the eligibility criteria were individually reviewed by 2 coauthors, and key findings were extracted and synthesized. We focused on studies using ML algorithms to predict cancer symptoms, excluding nonhuman research, technical reports, reviews, book chapters, conference proceedings, and inaccessible full texts.

RESULTS

A total of 42 studies were included, the majority of which were published after 2017. Most studies were conducted in North America (18/42, 43%) and Asia (16/42, 38%). The sample sizes in most studies (27/42, 64%) typically ranged from 100 to 1000 participants. The most prevalent category of algorithms was supervised ML, accounting for 39 (93%) of the 42 studies. Each of the methods-deep learning, ensemble classifiers, and unsupervised ML-constituted 3 (3%) of the 42 studies. The ML algorithms with the best performance were logistic regression (9/42, 17%), random forest (7/42, 13%), artificial neural networks (5/42, 9%), and decision trees (5/42, 9%). The most commonly included primary cancer sites were the head and neck (9/42, 22%) and breast (8/42, 19%), with 17 (41%) of the 42 studies not specifying the site. The most frequently studied symptoms were xerostomia (9/42, 14%), depression (8/42, 13%), pain (8/42, 13%), and fatigue (6/42, 10%). The significant predictors were age, gender, treatment type, treatment number, cancer site, cancer stage, chemotherapy, radiotherapy, chronic diseases, comorbidities, physical factors, and psychological factors.

CONCLUSIONS

This review outlines the algorithms used for predicting symptoms in individuals with cancer. Given the diversity of symptoms people with cancer experience, analytic approaches that can handle complex and nonlinear relationships are critical. This knowledge can pave the way for crafting algorithms tailored to a specific symptom. In addition, to improve prediction precision, future research should compare cutting-edge ML strategies such as deep learning and ensemble methods with traditional statistical models.

Developing and validating a nomogram for cognitive impairment in the older people based on the NHANES

Objective

To use the United States National Health and Nutrition Examination Study (NHANES) to develop and validate a risk-prediction nomogram for cognitive impairment in people aged over 60 years.

Methods

A total of 2,802 participants (aged ≥ 60 years) from NHANES were analyzed. The least absolute shrinkage and selection operator (LASSO) regression model and multivariable logistic regression analysis were used for variable selection and model development. ROC-AUC, calibration curve, and decision curve analysis (DCA) were used to evaluate the nomogram's performance.

Results

The nomogram included five predictors, namely sex, moderate activity, taste problem, age, and education. It demonstrated satisfying discrimination with a AUC of 0.744 (95% confidence interval, 0.696-0.791). The nomogram was well-calibrated according to the calibration curve. The DCA demonstrated that the nomogram was clinically useful.

Conclusion

The risk-prediction nomogram for cognitive impairment in people aged over 60 years was effective. All predictors included in this nomogram can be easily accessed from its' user.

Predictors of cognitive impairment in patients undergoing ileostomy for colorectal cancer: a retrospective analysis

Background

Early detection of cognitive impairment in patients undergoing ileostomy for colorectal cancer may help improve patient outcomes and quality of life. Identifying risk factors and clinically accessible factors is crucial for prevention and treatment.

Objective

This retrospective study aimed to identify risk factors for post-operative cognitive impairment in patients undergoing ileostomy for colorectal cancer and to explore potential factors for its prevention and treatment.

Methods

A total of 108 cases were selected and included in the study. Patient data including general characteristics, disease stage, complications, and chemotherapy status were collected, and sleep quality and cognitive function were assessed using questionnaires and follow-up. Patients were randomly divided into training and validation groups. A random forest model was used to rank clinical features based on their contribution to predicting the prognosis of cancer-related cognitive impairment (CRCI). Nomograms were constructed using the support vector machine-recursive feature elimination (SVM-RFE) method, and the minimal root-mean-square error (RMSE) values were compared to select the best model. Regression analysis was performed to determine independent predictors.

Results

Significant differences were observed in age, body mass index (BMI), alcohol consumption, frequency of physical activity, comorbidity, and cancer-related anemia (CRA) between the CRCI and non-CRCI groups. Random forest analysis revealed that age, BMI, exercise intensity, PSQI scores, and history of hypertension were the most significant predictors of outcome. Univariate logistic regression analysis of 18 variables revealed that age, alcohol consumption, exercise intensity, BMI, and comorbidity were significantly associated with the outcome of CRCI (p < 0.05). Univariate and multivariate models with P-values less than 0.1 and 0.2, respectively, showed better predictive performance for CRCI. The results of univariate analysis were plotted on a nomogram to evaluate the risk of developing CRCI after colorectal cancer surgery. The nomogram was found to have good predictive performance. Finally, regression analysis revealed that age, exercise intensity, BMI, comorbidity, and CRA were independent predictors of CRCI.

Conclusions

This retrospective cohort study revealed that age, exercise intensity, BMI, comorbidity, CRA, and mobility are independent predictors of cognitive impairment in patients undergoing ileostomy for colorectal cancer. Identifying these factors and potential factors may have clinical implications in predicting and managing post-operative cognitive impairment in this patient population.

Methodological and Quality Flaws in the Use of Artificial Intelligence in Mental Health Research: Systematic Review

BACKGROUND

Artificial intelligence (AI) is giving rise to a revolution in medicine and health care. Mental health conditions are highly prevalent in many countries, and the COVID-19 pandemic has increased the risk of further erosion of the mental well-being in the population. Therefore, it is relevant to assess the current status of the application of AI toward mental health research to inform about trends, gaps, opportunities, and challenges.

OBJECTIVE

This study aims to perform a systematic overview of AI applications in mental health in terms of methodologies, data, outcomes, performance, and quality.

METHODS

A systematic search in PubMed, Scopus, IEEE Xplore, and Cochrane databases was conducted to collect records of use cases of AI for mental health disorder studies from January 2016 to November 2021. Records were screened for eligibility if they were a practical implementation of AI in clinical trials involving mental health conditions. Records of AI study cases were evaluated and categorized by the International Classification of Diseases 11th Revision (ICD-11). Data related to trial settings, collection methodology, features, outcomes, and model development and evaluation were extracted following the CHARMS (Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modelling Studies) guideline. Further, evaluation of risk of bias is provided.

RESULTS

A total of 429 nonduplicated records were retrieved from the databases and 129 were included for a full assessment-18 of which were manually added. The distribution of AI applications in mental health was found unbalanced between ICD-11 mental health categories. Predominant categories were Depressive disorders (n=70) and Schizophrenia or other primary psychotic disorders (n=26). Most interventions were based on randomized controlled trials (n=62), followed by prospective cohorts (n=24) among observational studies. AI was typically applied to evaluate quality of treatments (n=44) or stratify patients into subgroups and clusters (n=31). Models usually applied a combination of questionnaires and scales to assess symptom severity using electronic health records (n=49) as well as medical images (n=33). Quality assessment revealed important flaws in the process of AI application and data preprocessing pipelines. One-third of the studies (n=56) did not report any preprocessing or data preparation. One-fifth of the models were developed by comparing several methods (n=35) without assessing their suitability in advance and a small proportion reported external validation (n=21). Only 1 paper reported a second assessment of a previous AI model. Risk of bias and transparent reporting yielded low scores due to a poor reporting of the strategy for adjusting hyperparameters, coefficients, and the explainability of the models. International collaboration was anecdotal (n=17) and data and developed models mostly remained private (n=126).

CONCLUSIONS

These significant shortcomings, alongside the lack of information to ensure reproducibility and transparency, are indicative of the challenges that AI in mental health needs to face before contributing to a solid base for knowledge generation and for being a support tool in mental health management.

High-fiber-diet-related metabolites improve neurodegenerative symptoms in patients with obesity with diabetes mellitus by modulating the hippocampal-hypothalamic endocrine axis

Objective

Through transcriptomic and metabolomic analyses, this study examined the role of high-fiber diet in obesity complicated by diabetes and neurodegenerative symptoms.

Method

The expression matrix of high-fiber-diet-related metabolites, blood methylation profile associated with pre-symptomatic dementia in elderly patients with type 2 diabetes mellitus (T2DM), and high-throughput single-cell sequencing data of hippocampal samples from patients with Alzheimer's disease (AD) were retrieved from the Gene Expression Omnibus (GEO) database and through a literature search. Data were analyzed using principal component analysis (PCA) after quality control and data filtering to identify different cell clusters and candidate markers. A protein-protein interaction network was mapped using the STRING database. To further investigate the interaction among high-fiber-diet-related metabolites, methylation-related DEGs related to T2DM, and single-cell marker genes related to AD, AutoDock was used for semi-flexible molecular docking.

Result

Based on GEO database data and previous studies, 24 marker genes associated with high-fiber diet, T2DM, and AD were identified. Top 10 core genes include SYNE1, ANK2, SPEG, PDZD2, KALRN, PTPRM, PTPRK, BIN1, DOCK9, and NPNT, and their functions are primarily related to autophagy. According to molecular docking analysis, acetamidobenzoic acid, the most substantially altered metabolic marker associated with a high-fiber diet, had the strongest binding affinity for SPEG.

Conclusion

By targeting the SPEG protein in the hippocampus, acetamidobenzoic acid, a metabolite associated with high-fiber diet, may improve diabetic and neurodegenerative diseases in obese people.

Impact of sarcopenia on intertrochanteric femoral fracture in the elderly

Objective

The aim of this study was to investigate the effect of skeletal sarcopenia on the prognosis of intertrochanteric fracture in the elderly.

Methods

We collected information on 144 patients with femoral intertrochanteric fracture (FIF). The influence of sarcopenia on the chance of death was determined using binary Probit regression analysis. For additional analysis, the Chow test was utilized to select the best distinguishing node in the instrumental activities of daily living (IADL) score. We looked for characteristics that were linked to a higher probability of death and a poor IADL outcome within 1 year. The data collected above were analyzed using logistic regression analysis. The internal calibration degree and model validity were assessed by GiViTI calibration.

Results

Sarcopenia, EuroQol-5D 1 month score, age, gender, and hypertension were identified as risk factors for death in older patients with FIF within a year by logistic regression analysis. Sarcopenia, psychotropics, BMI, and length of hospital stay were all found to be risk factors for poor IADL outcomes (P < 0.1). The calibration curves indicated that the anticipated and actual probabilities of these two models were very close. The study's reliability coefficient was 0.671, showing a satisfactory level of reliability.

Conclusion

In elderly patients with FIF, sarcopenia, EuroQol-5D score, age, gender, and hypertension were risk factors for death; sarcopenia, hospital stay length, BMI were risk factors for poor quality of life.

Preoperative Serum Calcitonin Level and Ultrasonographic Characteristics Predict the Risk of Metastatic Medullary Thyroid Carcinoma: Functional Analysis of Calcitonin-Related Genes

Background. Early cervical lymph node (LN) metastasis is an important cause of poor survival in patients with medullary thyroid cancer (MTC). This study evaluated whether the preoperative serum calcitonin level in combination with ultrasonographic features of MTC can be used to assess the LN status as well as predict the risk of metastasis in patients with MTC. Methods. We retrospectively analyzed the clinical data of 95 patients with MTC, and a nomogram model was constructed and validated. Using integrated database analysis of The Cancer Genome Atlas (TCGA) and Genotype-Tissue Expression (GTEx), we mined pathways wherein CALCA is involved, identified calcitonin-related genes, and analyzed their functions. Results. Correlation analysis revealed a significant association between the infiltrating range, diameter, calcification, blood flow, the preoperative serum calcitonin level, and metastasis. The metastasis risk-prediction model showed great accuracy in determining the risk of metastasis in MTC (area under the curve of the receiver operating characteristic [ROC] curve: 0.979 [95% confidence interval 0.946–1.000]). Decision curve analysis (DCA) showed that the model has excellent clinical utilization potential. Significantly, CALCA, the mRNA for calcitonin, was highly expressed in thyroid cancer tissues and associated with the cytokine–cytokine receptor and neuroactive ligand-receptor interaction pathways as well as the cell-adhesion molecules. ROC curve indicated that the CNTFR, CD27, GDF6, and TSLP genes, which are related to the cytokine–cytokine receptor interaction pathway, could indicate the risk of metastasis in MTC. Conclusions. The preoperative serum calcitonin level, in combination with ultrasonographic features, can be used to predict the risk of metastasis in patients with MTC and constitute a noninvasive accurate method for preoperative diagnosis of MTC.

Potential Mechanism Underlying Exercise Upregulated Circulating Blood Exosome miR-215-5p to Prevent Necroptosis of Neuronal Cells and a Model for Early Diagnosis of Alzheimer's Disease

Exercise is crucial for preventing Alzheimer's disease (AD), although the exact underlying mechanism remains unclear. The construction of an accurate AD risk prediction model is beneficial as it can provide a theoretical basis for preventive exercise prescription. In recent years, necroptosis has been confirmed as an important manifestation of AD, and exercise is known to inhibit necroptosis of neuronal cells. In this study, we extracted 67 necroptosis-related genes and 32 necroptosis-related lncRNAs and screened for key predictive AD risk genes through a random forest analysis. Based on the neural network Prediction model, we constructed a new logistic regression-based AD risk prediction model in order to provide a visual basis for the formulation of exercise prescription. The prediction model had an area under the curve (AUC) value of 0.979, indicative of strong predictive power and a robust clinical application prospect. In the exercise group, the expression of exosomal miR-215-5p was found to be upregulated; miR-215-5p could potentially inhibit the expressions of IDH1, BCL2L11, and SIRT1. The single-cell SCENIC assay was used to identify key transcriptional regulators in skeletal muscle. Among them, CEBPB and GATA6 were identified as putative transcriptional regulators of miR-215. After "skeletal muscle removal of load," the expressions of CEBPB and GATA6 increased substantially, which in turn led to the elevation of miR-215 expression, thereby suggesting a putative mechanism for negative feedback regulation of exosomal homeostasis.

'Food for Thought'-The Relationship between Diet and Cognition in Breast and Colorectal Cancer Survivors: A Feasibility Study

Survivors of cancer frequently experience persistent and troublesome cognitive changes. Little is known about the role diet and nutrition plays in survivors' cognition. We explored the feasibility of collecting cross-sectional online data from Australian survivors of breast and colorectal cancer to enable preliminary investigations of the relationships between cognition with fruit and vegetable intake, and the Omega-3 Index (a biomarker of long chain omega 3 fatty acid intake). A total of 76 participants completed online (and postal Omega-3 Index biomarker) data collection (62 breast and 14 colorectal cancer survivors): mean age 57.5 (±10.2) years, mean time since diagnosis 32.6 (±15.6) months. Almost all of the feasibility outcomes were met; however, technical difficulties were reported for online cognitive testing. In hierarchical linear regression models, none of the dietary variables of interest were significant predictors of self-reported or objective cognition. Age, BMI, and length of treatment predicted some of the cognitive outcomes. We demonstrated a viable online/postal data collection method, with participants reporting positive levels of engagement and satisfaction. Fruit, vegetable, and omega-3 intake were not significant predictors of cognition in this sample, however the role of BMI in survivors' cognitive functioning should be further investigated. Future research could adapt this protocol to longitudinally monitor diet and cognition to assess the impact of diet on subsequent cognitive function, and whether cognitive changes impact dietary habits in survivors of cancer.