A clinical text classification paradigm using weak supervision and deep representation

Extraction of sleep information from clinical notes of Alzheimer's disease patients using natural language processing

OBJECTIVES

Alzheimer's disease (AD) is the most common form of dementia in the United States. Sleep is one of the lifestyle-related factors that has been shown critical for optimal cognitive function in old age. However, there is a lack of research studying the association between sleep and AD incidence. A major bottleneck for conducting such research is that the traditional way to acquire sleep information is time-consuming, inefficient, non-scalable, and limited to patients' subjective experience. We aim to automate the extraction of specific sleep-related patterns, such as snoring, napping, poor sleep quality, daytime sleepiness, night wakings, other sleep problems, and sleep duration, from clinical notes of AD patients. These sleep patterns are hypothesized to play a role in the incidence of AD, providing insight into the relationship between sleep and AD onset and progression.

MATERIALS AND METHODS

A gold standard dataset is created from manual annotation of 570 randomly sampled clinical note documents from the adSLEEP, a corpus of 192 000 de-identified clinical notes of 7266 AD patients retrieved from the University of Pittsburgh Medical Center (UPMC). We developed a rule-based natural language processing (NLP) algorithm, machine learning models, and large language model (LLM)-based NLP algorithms to automate the extraction of sleep-related concepts, including snoring, napping, sleep problem, bad sleep quality, daytime sleepiness, night wakings, and sleep duration, from the gold standard dataset.

RESULTS

The annotated dataset of 482 patients comprised a predominantly White (89.2%), older adult population with an average age of 84.7 years, where females represented 64.1%, and a vast majority were non-Hispanic or Latino (94.6%). Rule-based NLP algorithm achieved the best performance of F1 across all sleep-related concepts. In terms of positive predictive value (PPV), the rule-based NLP algorithm achieved the highest PPV scores for daytime sleepiness (1.00) and sleep duration (1.00), while the machine learning models had the highest PPV for napping (0.95) and bad sleep quality (0.86), and LLAMA2 with finetuning had the highest PPV for night wakings (0.93) and sleep problem (0.89).

DISCUSSION

Although sleep information is infrequently documented in the clinical notes, the proposed rule-based NLP algorithm and LLM-based NLP algorithms still achieved promising results. In comparison, the machine learning-based approaches did not achieve good results, which is due to the small size of sleep information in the training data.

CONCLUSION

The results show that the rule-based NLP algorithm consistently achieved the best performance for all sleep concepts. This study focused on the clinical notes of patients with AD but could be extended to general sleep information extraction for other diseases.

Predicting Suicidal Ideation, Planning, and Attempts among the Adolescent Population of the United States

Suicide is the second leading cause of death among individuals aged 5 to 24 in the United States (US). However, the precursors to suicide often do not surface, making suicide prevention challenging. This study aims to develop a machine learning model for predicting suicide ideation (SI), suicide planning (SP), and suicide attempts (SA) among adolescents in the US during the coronavirus pandemic. We used the 2021 Adolescent Behaviors and Experiences Survey Data. Class imbalance was addressed using the proposed data augmentation method tailored for binary variables, Modified Synthetic Minority Over-Sampling Technique. Five different ML models were trained and compared. SHapley Additive exPlanations analysis was conducted for explainability. The Logistic Regression model, identified as the most effective, showed superior performance across all targets, achieving high scores in recall: 0.82, accuracy: 0.80, and area under the Receiver Operating Characteristic curve: 0.88. Variables such as sad feelings, hopelessness, sexual behavior, and being overweight were noted as the most important predictors. Our model holds promise in helping health policymakers design effective public health interventions. By identifying vulnerable sub-groups within regions, our model can guide the implementation of tailored interventions that facilitate early identification and referral to medical treatment.

Health professionals' acceptance of mobile-based clinical guideline application in a resource-limited setting: using a modified UTAUT model

INTRODUCTION

Clinical guidelines are crucial for assisting health professionals to make correct clinical decisions. However, manual clinical guidelines are not accessible, and this increases the workload. So, a mobile-based clinical guideline application is needed to provide real-time information access. Hence, this study aimed to assess health professionals' intention to accept mobile-based clinical guideline applications and verify the unified theory of acceptance and technology utilization model.

METHODS

Institutional-based cross-sectional study design was used among 803 study participants. The sample size was determined based on structural equation model parameter estimation criteria with stratified random sampling. Amos version 23 software was used for analysis. Internal consistency of latent variable items, and convergent and divergent validity, were evaluated using composite reliability, AVE, and a cross-loading matrix. Model fitness of the data was assessed based on a set of criteria, and it was achieved. P-value < 0.05 was considered for assessing the formulated hypothesis.

RESULTS

Effort expectancy and social influence had a significant effect on health professionals' attitudes, with path coefficients of (β = 0.61, P-value < 0.01), and (β = 0.510, P-value < 0.01) respectively. Performance expectancy, facilitating condition, and attitude had significant effects on health professionals' acceptance of mobile-based clinical guideline applications with path coefficients of (β = 0.37, P-value < 0.001), (β = 0.44, P-value < 0.001) and (β = 0.57, P-value < 0.05) respectively. Effort expectancy and social influence were mediated by attitude and had a significant partial relationship with health professionals' acceptance of mobile-based clinical guideline application with standardized estimation coefficients of (β = 0.22, P-value = 0.027), and (β = 0.19, P-value = 0.031) respectively. All the latent variables accounted for 57% of health professionals' attitudes, and latent variables with attitudes accounted for 63% of individuals' acceptance of mobile-based clinical guideline applications.

CONCLUSIONS

The unified theory of acceptance and use of the technology model was a good model for assessing individuals' acceptance of mobile-based clinical guidelines applications. So, enhancing health professionals' attitudes, and computer literacy through training are needed. Mobile application development based on user requirements is critical for technology adoption, and people's support is also important for health professionals to accept and use the application.

The Role of Artificial Intelligence in Healthcare: Enhancing Coronary Computed Tomography Angiography for Coronary Artery Disease Management

This review aims to explore the potential of artificial intelligence (AI) in coronary CT angiography (CCTA), a key tool for diagnosing coronary artery disease (CAD). Because CAD is still a major cause of death worldwide, effective and accurate diagnostic methods are required to identify and manage the condition. CCTA certainly is a noninvasive alternative for diagnosing CAD, but it requires a large amount of data as input. We intend to discuss the idea of incorporating AI into CCTA, which enhances its diagnostic accuracy and operational efficiency. Using such AI technologies as machine learning (ML) and deep learning (DL) tools, CCTA images are automated to perfection and the analysis is significantly refined. It enables the characterization of a plaque, assesses the severity of the stenosis, and makes more accurate risk stratifications than traditional methods, with pinpoint accuracy. Automating routine tasks through AI-driven CCTA will reduce the radiologists' workload considerably, which is a standard benefit of such technologies. More importantly, it would enable radiologists to allocate more time and expertise to complex cases, thereby improving overall patient care. However, the field of AI in CCTA is not without its challenges, which include data protection, algorithm transparency, as well as criteria for standardization encoding. Despite such obstacles, it appears that the integration of AI technology into CCTA in the future holds great promise for keeping CAD itself in check, thereby aiding the fight against this disease and begetting better clinical outcomes and more optimized modes of healthcare. Future research on AI algorithms for CCTA, making ethical use of AI, and thereby overcoming the technical and clinical barriers to widespread adoption of this new tool, will hopefully pave the way for profound AI-driven transformations in healthcare.

Perceived social support and treatment adherence in Chileans with Type 2 diabetes

This study aimed to determine if greater perceived social support was directly associated with better Type 2 diabetes (T2D) treatment adherence and if better T2D treatment adherence was related to lower HbA1c levels in Chilean adults with T2D. For this purpose, 200 adults were recruited from the Chilean Diabetic Association. Participants were asked to complete self-report instruments and provide a capillary blood sample to measure HbA1c. Structural equation model analyses were performed to determine direct associations. The study's results indicate that greater perceived social support was associated with healthier dietary habits, regular foot care, more frequent physical activity, and lower medication intake. Likewise, blood sugar testing and physical activity were related to HbA1c. These findings provide evidence of how perceived social support relates to T2D treatment adherence behaviors in Latino patients from South America and could be used for interventions to enhance social support from patients' families, partners, and friends.

Developmental changes of bodily self-consciousness in adolescent girls

The body and the self change markedly during adolescence, but how does bodily self-consciousness, the pre-reflexive experience of being a bodily subject, change? We addressed this issue by studying embodiment towards virtual avatars in 70 girls aged 10-17 years. We manipulated the synchrony between participants' and avatars' touch or movement, as well as the avatar visual shape or size relative to each participant's body. A weaker avatar's embodiment in case of mismatch between the body seen in virtual reality and the real body is indicative of a more robust bodily self-consciousness. In both the visuo-tactile and the visuo-motor experiments, asynchrony decreased ownership feeling to the same extent for all participants, while the effect of asynchrony on agency feeling increased with age. In the visuo-tactile experiment, incongruence in visual appearance did not affect agency feeling but impacted ownership, especially in older teenage girls. These findings highlight the higher malleability of bodily self-consciousness at the beginning of adolescence and suggest some independence between body ownership and agency.

Comparative effect of physical health training and psychological training of the theory of reasoned action (TRA) model on the life quality of patients with diabetes in Tehran, Iran: utilization of message texting

BACKGROUND AND PURPOSE

Providing physical health and mental health training promotion is necessary for a sustainable change in attitude and lifestyle of diabetic patients. The present study was conducted with the aim of comparing the effect of physical health training and psychological training of the theory of reasoned action (TRA) model on the life quality of patients with type 2 diabetes.

METHODS

This experimental study was conducted in 2022 with two intervention groups and one control group consisting of 129 patients with type 2 diabetes who were referred to Imam Khomeini Hospital in Tehran. Over the course of one month, each individual in intervention group 1 received 15 text messages focusing on physical health, while intervention group 2 received 15 psychological text messages related to the TRA. The control group did not receive any text messages during this period. The data collection tool used was the "Audit of Diabetes-Dependent Quality of Life (ADDQoL)" questionnaire, which was completed by the participants before and after the intervention. The data were analyzed using SPSS version 16 software at a statistical significance level of 0.05.

RESULTS

In the intervention-1 group, the average life quality score was 8.51 units (P < 0.001), while in the intervention-2 group, it was 19.25 units (P < 0.001) higher than the control group. The psychological training group had a 17.62 units (P < 0.05) lower average fasting blood sugar (FBS) and a 10.74 units (P < 0.001) higher average quality of life compared to the physical training group.

CONCLUSION

The results of this study showed that the effectiveness of psychological training of the TRA model in improving life quality and reducing FBS in patients with diabetes is greater than physical health training. It is suggested that policy makers and health managers base future plans on physical health promotion training along with TRA model mental health training for the development of education in patients with diabetes. Specialists and healthcare workers can also act to improve personal health characteristics, especially those related to reducing FBS and increasing the quality of life of patients with diabetes, by using training through mobile phone text messages, particularly with psychological content TRA based.

Exploring clinical specialists' perspectives on the future role of AI: evaluating replacement perceptions, benefits, and drawbacks

BACKGROUND OF STUDY

Over the past few decades, the utilization of Artificial Intelligence (AI) has surged in popularity, and its application in the medical field is witnessing a global increase. Nevertheless, the implementation of AI-based healthcare solutions has been slow in developing nations like Pakistan. This unique study aims to assess the opinion of clinical specialists on the future replacement of AI, its associated benefits, and its drawbacks in form southern region of Pakistan.

MATERIAL AND METHODS

A cross-sectional selective study was conducted from 140 clinical specialists (Surgery = 24, Pathology = 31, Radiology = 35, Gynecology = 35, Pediatric = 17) from the neglected southern Punjab region of Pakistan. The study was analyzed using χ2 - the test of association and the nexus between different factors was examined by multinomial logistic regression.

RESULTS

Out of 140 respondents, 34 (24.3%) believed hospitals were ready for AI, while 81 (57.9%) disagreed. Additionally, 42(30.0%) were concerned about privacy violations, and 70(50%) feared AI could lead to unemployment. Specialists with less than 6 years of experience are more likely to embrace AI (p = 0.0327, OR = 3.184, 95% C.I; 0.262, 3.556) and those who firmly believe that AI knowledge will not replace their future tasks exhibit a lower likelihood of accepting AI (p = 0.015, OR = 0.235, 95% C.I: (0.073, 0.758). Clinical specialists who perceive AI as a technology that encompasses both drawbacks and benefits demonstrated a higher likelihood of accepting its adoption (p = 0.084, OR = 2.969, 95% C.I; 0.865, 5.187).

CONCLUSION

Clinical specialists have embraced AI as the future of the medical field while acknowledging concerns about privacy and unemployment.

Optimizing type 2 diabetes management: AI-enhanced time series analysis of continuous glucose monitoring data for personalized dietary intervention

Despite advanced health facilities in many developed countries, diabetic patients face multifold health challenges. Type 2 diabetes mellitus (T2DM) go along with conspicuous symptoms due to frequent peaks, hypoglycemia <=70 mg/dL (while fasting), or hyperglycemia >=180 mg/dL two hours postprandial, according to the American Diabetes Association (ADA)). The worse effects of Type 2 diabetes mellitus are precisely associated with the poor lifestyle adopted by patients. In particular, a healthy diet and nutritious food are the key to success for such patients. This study was done to help T2DM patients improve their health by developing a favorable lifestyle under an AI-assisted Continuous glucose monitoring (CGM) digital system. This study aims to reduce the blood glucose level fluctuations of such patients by rectifying their daily diet and maintaining their exertion vs. food consumption records. In this study, a well-precise prediction is obtained by training the ML model on a dataset recorded from CGM sensor devices attached to T2DM patients under observation. As the data obtained from the CGM sensor is time series, to predict blood glucose levels, the time series analysis and forecasting are done with XGBoost, SARIMA, and Prophet. The results of different Models are then compared based on performance metrics. This helped in monitoring various trends, specifically irregular patterns of the patient's glucose data, collected by the CGM sensor. Later, keeping track of these trends and seasonality, the diet is adjusted accordingly by adding or removing particular food and keeping track of its nutrients with the intervention of a commercially available all-in-one AI solution for food recognition. This created an interactive assistive system, where the predicted results are compared to food contents to bring the blood glucose levels within the normal range for maintaining a healthy lifestyle and to alert about blood glucose fluctuations before the time that are going to occur sooner. This study will help T2DM patients get in managing diabetes and ultimately bring HbA1c within the normal range (<= 5.7%) for diabetic and pre-diabetic patients, three months after the intervention.

Employing supervised machine learning algorithms for classification and prediction of anemia among youth girls in Ethiopia

In developing countries, one-quarter of young women have suffered from anemia. However, the available studies in Ethiopia have been usually used the traditional stastical methods. Therefore, this study aimed to employ multiple machine learning algorithms to identify the most effective model for the prediction of anemia among youth girls in Ethiopia. A total of 5642 weighted samples of young girls from the 2016 Ethiopian Demographic and Health Survey dataset were utilized. The data underwent preprocessing, with 80% of the observations used for training the model and 20% for testing. Eight machine learning algorithms were employed to build and compare models. The model performance was assessed using evaluation metrics in Python software. Various data balancing techniques were applied, and the Boruta algorithm was used to select the most relevant features. Besides, association rule mining was conducted using the Apriori algorithm in R software. The random forest classifier with an AUC value of 82% outperformed in predicting anemia among all the tested classifiers. Region, poor wealth index, no formal education, unimproved toilet facility, rural residence, not used contraceptive method, religion, age, no media exposure, occupation, and having more than 5 family size were the top attributes to predict anemia. Association rule mining was identified the top seven best rules that most frequently associated with anemia. The random forest classifier is the best for predicting anemia. Therefore, making it potentially valuable as decision-support tools for the relevant stakeholders and giving emphasis for the identified predictors could be an important intervention to halt anemia among youth girls.

The use of medical health applications by primary care physicians in Israel: a cross-sectional study

BACKGROUND

The use of medical health applications (mHealth apps) by patients, caregivers, and physicians is widespread. mHealth apps are often employed by physicians to quickly access professional knowledge, guide treatment, easily retrieve medical records, and monitor and manage patients. This study sought to characterize the use of mHealth apps among primary care physicians (PCPs) in Israel. The reasons for using apps and barriers to their use were also investigated.

METHODS

From all MHS' PCPs, we randomly selected 700 PCPs and invited them to complete a questionnaire regarding the use of mHealth apps and attitudes toward them.

RESULTS

From August 2020 to December 2020, 191 physicians completed the questionnaire (response rate 27.3%). 68.0% of PCPs reported using mHealth apps. Telemedicine service apps were the most frequently used. Medical calculators (used for clinical scoring) and differential diagnosis apps were the least frequently used. The most common reason for mHealth app use was accessibility, followed by time saved and a sense of information reliability. Among infrequent users of apps, the most common barriers reported were unfamiliarity with relevant apps and preference for using a computer. Concerns regarding information reliability were rarely reported by PCPs. Physician gender and seniority were not related to mHealth app use. Physician age was related to the use of mHealth apps.

CONCLUSIONS

mHealth apps are widely used by PCPs in this study, regardless of physician gender or seniority. Information from mHealth apps is considered reliable by PCPs. The main barrier to app use is unfamiliarity with relevant apps and preference for computer use.

Risk assessment of imported malaria in China: a machine learning perspective

BACKGROUND

Following China's official designation as malaria-free country by WHO, the imported malaria has emerged as a significant determinant impacting the malaria reestablishment within China. The objective of this study is to explore the application prospects of machine learning algorithms in imported malaria risk assessment of China.

METHODS

The data of imported malaria cases in China from 2011 to 2019 was provided by China CDC; historical epidemic data of malaria endemic country was obtained from World Malaria Report, and the other data used in this study are open access data. All the data processing and model construction based on R, and map visualization used ArcGIS software.

RESULTS

A total of 27,088 malaria cases imported into China from 85 countries between 2011 and 2019. After data preprocessing and classification, clean dataset has 765 rows (85 * 9) and 11 cols. Six machine learning models was constructed based on the training set, and Random Forest model demonstrated the best performance in model evaluation. According to RF, the highest feature importance were the number of malaria deaths and Indigenous malaria cases. The RF model demonstrated high accuracy in forecasting risk for the year 2019, achieving commendable accuracy rate of 95.3%. This result aligns well with the observed outcomes, indicating the model's reliability in predicting risk levels.

CONCLUSIONS

Machine learning algorithms have reliable application prospects in risk assessment of imported malaria in China. This study provides a new methodological reference for the risk assessment and control strategies adjusting of imported malaria in China.

Natural Language Processing for Radiation Oncology: Personalizing Treatment Pathways

Natural language processing (NLP), a technology that translates human language into machine-readable data, is revolutionizing numerous sectors, including cancer care. This review outlines the evolution of NLP and its potential for crafting personalized treatment pathways for cancer patients. Leveraging NLP's ability to transform unstructured medical data into structured learnable formats, researchers can tap into the potential of big data for clinical and research applications. Significant advancements in NLP have spurred interest in developing tools that automate information extraction from clinical text, potentially transforming medical research and clinical practices in radiation oncology. Applications discussed include symptom and toxicity monitoring, identification of social determinants of health, improving patient-physician communication, patient education, and predictive modeling. However, several challenges impede the full realization of NLP's benefits, such as privacy and security concerns, biases in NLP models, and the interpretability and generalizability of these models. Overcoming these challenges necessitates a collaborative effort between computer scientists and the radiation oncology community. This paper serves as a comprehensive guide to understanding the intricacies of NLP algorithms, their performance assessment, past research contributions, and the future of NLP in radiation oncology research and clinics.

The validity of electronic health data for measuring smoking status: a systematic review and meta-analysis

BACKGROUND

Smoking is a risk factor for many chronic diseases. Multiple smoking status ascertainment algorithms have been developed for population-based electronic health databases such as administrative databases and electronic medical records (EMRs). Evidence syntheses of algorithm validation studies have often focused on chronic diseases rather than risk factors. We conducted a systematic review and meta-analysis of smoking status ascertainment algorithms to describe the characteristics and validity of these algorithms.

METHODS

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines were followed. We searched articles published from 1990 to 2022 in EMBASE, MEDLINE, Scopus, and Web of Science with key terms such as validity, administrative data, electronic health records, smoking, and tobacco use. The extracted information, including article characteristics, algorithm characteristics, and validity measures, was descriptively analyzed. Sources of heterogeneity in validity measures were estimated using a meta-regression model. Risk of bias (ROB) in the reviewed articles was assessed using the Quality Assessment of Diagnostic Accuracy Studies-2 tool.

RESULTS

The initial search yielded 2086 articles; 57 were selected for review and 116 algorithms were identified. Almost three-quarters (71.6%) of algorithms were based on EMR data. The algorithms were primarily constructed using diagnosis codes for smoking-related conditions, although prescription medication codes for smoking treatments were also adopted. About half of the algorithms were developed using machine-learning models. The pooled estimates of positive predictive value, sensitivity, and specificity were 0.843, 0.672, and 0.918 respectively. Algorithm sensitivity and specificity were highly variable and ranged from 3 to 100% and 36 to 100%, respectively. Model-based algorithms had significantly greater sensitivity (p = 0.006) than rule-based algorithms. Algorithms for EMR data had higher sensitivity than algorithms for administrative data (p = 0.001). The ROB was low in most of the articles (76.3%) that underwent the assessment.

CONCLUSIONS

Multiple algorithms using different data sources and methods have been proposed to ascertain smoking status in electronic health data. Many algorithms had low sensitivity and positive predictive value, but the data source influenced their validity. Algorithms based on machine-learning models for multiple linked data sources have improved validity.

Enhancing breast cancer classification via histopathological image analysis: Leveraging self-supervised contrastive learning and transfer learning

Breast cancer, a significant threat to women's health, demands early detection. Automating histopathological image analysis offers a promising solution to enhance efficiency and accuracy in diagnosis. This study addresses the challenge of breast cancer histopathological image classification by leveraging the ResNet architecture, known for its depth and skip connections. In this work, two distinct approaches were pursued, each driven by unique motivations. The first approach aimed to improve the learning process through self-supervised contrastive learning. It utilizes a small subset of the training data for initial model training and progressively expands the training set by incorporating confidently labeled data from the unlabeled pool, ultimately achieving a reliable model with limited training data. The second approach focused on optimizing the architecture by combining ResNet50 and Inception module to get a lightweight and efficient classifier. The dataset utilized in this work comprises histopathological images categorized into benign and malignant classes at varying magnification levels (40X, 100X, 200X, 400X), all originating from the same source image. The results demonstrate state-of-the-art performance, achieving 98% accuracy for images magnified at 40X and 200X, and 94% for 100X and 400X. Notably, the proposed architecture boasts a substantially reduced parameter count of approximately 3.6 million, contrasting with existing leading architectures, which possess parameter sizes at least twice as large.

The effect of close and intensive therapeutic monitoring of patients with poorly controlled type 2 diabetes with different glycemic background

Patients with type 2 diabetes who have HbA1c values ≥ 10% have different previous glycemic trends, including new diagnosis of diabetes. We aimed to assess the efficacy of 3 months of intensive and facilitated antihyperglycemic treatment in patients with different glycemic backgrounds. In this observational study, patients with type 2 diabetes and poor glycemic control (indicated by an HbA1c level of > = 10%) were divided into groups based on their previous HbA1c levels (group 1; newly diagnosed type 2 diabetics, group 2; patients with previously controlled but now deteriorated HbA1c levels, group 3; patients whose HbA1c was not previously in the target range but was now above 10%, and group 4; patients whose HbA1c was above 10% from the start). Patients received intensive diabetes management with close monitoring and facilitated hospital visits. For further analysis, patients who were known to have previously had good metabolic control (either did not have diabetes or had previously had an HbA1c value < =7) and patients who had prior poor metabolic control were analyzed separately. Of the 195 participants [female, n = 84 (43.1%)], the median age was 54 years (inter-quantile range [IQR] = 15, min = 29, max = 80) and the median baseline HbA1c was 11.8% (IQR = 2.6%, min = 10%, max = 18.3%). The median duration of diabetes was 10 years (IQR = 9, min = 1, max = 35) when newly diagnosed patients were excluded. The ≥ 20% reduction in HbA1c at month 3 was observed in groups 1 to 4 in 97%, 88.1%, 69.1%, and 55.4%, respectively. The percentage of patients who achieved an HbA1c level of 7% or less was 60.6%, 38.1%, 16.4%, and 6.2% in the groups, respectively. The rate of those who achieved an HbA1c of 7% or less was nearly 50% of patients with type 2 diabetes mellitus who had previously had good metabolic control, whereas successful control was achieved in only 1 in 10 patients with persistently high HbA1c levels. Patients' glycemic history played an important role in determining their HbA1c levels at 3 months, suggesting that previous glycemic management patterns may indicate future success in diabetes control.

Machine learning analyses identify multi-modal frailty factors that selectively discriminate four cohorts in the Alzheimer's disease spectrum: a COMPASS-ND study

BACKGROUND

Frailty indicators can operate in dynamic amalgamations of disease conditions, clinical symptoms, biomarkers, medical signals, cognitive characteristics, and even health beliefs and practices. This study is the first to evaluate which, among these multiple frailty-related indicators, are important and differential predictors of clinical cohorts that represent progression along an Alzheimer's disease (AD) spectrum. We applied machine-learning technology to such indicators in order to identify the leading predictors of three AD spectrum cohorts; viz., subjective cognitive impairment (SCI), mild cognitive impairment (MCI), and AD. The common benchmark was a cohort of cognitively unimpaired (CU) older adults.

METHODS

The four cohorts were from the cross-sectional Comprehensive Assessment of Neurodegeneration and Dementia dataset. We used random forest analysis (Python 3.7) to simultaneously test the relative importance of 83 multi-modal frailty indicators in discriminating the cohorts. We performed an explainable artificial intelligence method (Tree Shapley Additive exPlanation values) for deep interpretation of prediction effects.

RESULTS

We observed strong concurrent prediction results, with clusters varying across cohorts. The SCI model demonstrated excellent prediction accuracy (AUC = 0.89). Three leading predictors were poorer quality of life ([QoL]; memory), abnormal lymphocyte count, and abnormal neutrophil count. The MCI model demonstrated a similarly high AUC (0.88). Five leading predictors were poorer QoL (memory, leisure), male sex, abnormal lymphocyte count, and poorer self-rated eyesight. The AD model demonstrated outstanding prediction accuracy (AUC = 0.98). Ten leading predictors were poorer QoL (memory), reduced olfaction, male sex, increased dependence in activities of daily living (n = 6), and poorer visual contrast.

CONCLUSIONS

Both convergent and cohort-specific frailty factors discriminated the AD spectrum cohorts. Convergence was observed as all cohorts were marked by lower quality of life (memory), supporting recent research and clinical attention to subjective experiences of memory aging and their potentially broad ramifications. Diversity was displayed in that, of the 14 leading predictors extracted across models, 11 were selectively sensitive to one cohort. A morbidity intensity trend was indicated by an increasing number and diversity of predictors corresponding to clinical severity, especially in AD. Knowledge of differential deficit predictors across AD clinical cohorts may promote precision interventions.

Uncovering the factors that affect earthquake insurance uptake using supervised machine learning

The escalating threat of natural disasters to public safety worldwide underlines the crucial role of effective environmental risk management tools, such as insurance. This is particularly evident in the case of earthquakes that occurred in Oklahoma between 2011 and 2020, which were linked to wastewater injection, underscoring the need for earthquake insurance. In this regard, from a survey of 812 respondents in Oklahoma, USA, we used supervised machine learning techniques (i.e., logit, ridge, least absolute shrinkage and selection operator (LASSO), decision tree, and random forest classifiers) to identify the factors that influence earthquake insurance uptake and to predict individuals who would acquire earthquake insurance. Our findings reveal that influential factors that affect earthquake insurance uptake include demographic factors such as older age, male gender, race, and ethnicity. These were found to significantly influence the decision to purchase earthquake insurance. Additionally, individuals residing in rental properties were less likely to purchase earthquake insurance, while longer residency in Oklahoma had a positive influence. Past experience of earthquakes was also found to positively influence the decision to purchase earthquake insurance. Both decision trees and random forests demonstrated good predictive capabilities for identifying earthquake insurance uptake. Notably, random forests exhibited higher precision and robustness, emerging as an encouraging choice for earthquake insurance modeling and other classification problems. Empirically, we highlight the importance of insurance as an environmental risk management tool and emphasize the need for awareness and education on earthquake insurance as well as the use of supervised machine learning algorithms for classification problems.

Improving preoperative prediction of surgery duration

BACKGROUND

Operating rooms (ORs) are one of the costliest units in a hospital, therefore the cumulative consequences of any kind of inefficiency in OR management lead to a significant loss of revenue for the hospital, staff dissatisfaction, and patient care disruption. One of the possible solutions to improving OR efficiency is knowing a reliable estimate of the duration of operations. The literature suggests that the current methods used in hospitals, e.g., a surgeon's estimate for the given surgery or taking the average of only five previous records of the same procedure, have room for improvement.

METHODS

We used over 4 years of elective surgery records (n = 52,171) from one of the major metropolitan hospitals in Australia. We developed robust Machine Learning (ML) approaches to provide a more accurate prediction of operation duration, especially in the absence of surgeon's estimation. Individual patient characteristics and historic surgery information attributed to medical records were used to train predictive models. A wide range of algorithms such as Extreme Gradient Boosting (XGBoost) and Random Forest (RF) were tested for predicting operation duration.

RESULTS

The results show that the XGBoost model provided statistically significantly less error than other compared ML models. The XGBoost model also reduced the total absolute error by 6854 min (i.e., about 114 h) compared to the current hospital methods.

CONCLUSION

The results indicate the potential of using ML methods for reaching a more accurate estimation of operation duration compared to current methods used in the hospital. In addition, using a set of realistic features in the ML models that are available at the point of OR scheduling enabled the potential deployment of the proposed approach.

LDS-CNN: a deep learning framework for drug-target interactions prediction based on large-scale drug screening

Background

Drug-target interaction (DTI) is a vital drug design strategy that plays a significant role in many processes of complex diseases and cellular events. In the face of challenges such as extensive protein data and experimental costs, it is suggested to apply bioinformatics approaches to exploit potential interactions to design new targeted medications. Different data and interaction types bring difficulties to study involving incompatible and heterology formats. The analysis of drug-target interactions in a comprehensive and unified model is a significant challenge.

Method

Here, we propose a general method for predicting interactions between small-molecule drugs and protein targets, Large-scale Drug target Screening Convolutional Neural Network (LDS-CNN), which used unified encoding to achieve the calculation of the different data formats in an integrated model to realize feature abstraction and potential object prediction.

Result

On 898,412 interaction data involving 1683 small-molecule compounds and 14,350 human proteins from 8.8 billion records, the proposed method achieved an area under the curve (AUC) of 0.96, an area under the precision-recall curve (AUPRC) of 0.95, and an accuracy of 90.13%. The experimental results illustrated that the proposed method attained high accuracy on the test set, indicating its high predictive ability in drug-target interaction prediction. LDS-CNN is effective for the prediction of large-scale datasets and datasets composed of data with different formats.

Conclusion

In this study, we propose a DTI prediction method to solve the problems of unified encoding of large-scale data in multiple formats. It provides a feasible way to efficiently abstract the features among different types of drug-related data, thus reducing experimental costs and time consumption. The proposed method can be used to identify potential drug targets and candidates for the treatment of complex diseases. This work provides a reference for DTI to process large-scale data and different formats with deep learning methods and provides certain suggestions for future research.

The 3Ranker: An AI-based Algorithm for Finding Non-animal Alternative Methods

The search for existing non-animal alternative methods for use in experiments is currently challenging because of the lack of both comprehensive structured databases and balanced keyword-based search strategies to mine unstructured textual databases. In this paper we describe 3Ranker, which is a fast, keyword-independent algorithm for finding non-animal alternative methods for use in biomedical research. The 3Ranker algorithm was created by using a machine learning approach, consisting of a Random Forest model built on a dataset of 35 million abstracts and constructed with weak supervision, followed by iterative model improvement with expert curated data. We found a satisfactory trade-off between sensitivity and specificity, with Area Under the Curve (AUC) values ranging from 0.85-0.95. Trials showed that the AI-based classifier was able to identify articles that describe potential alternatives to animal use, among the thousands of articles returned by generic PubMed queries on dermatitis and Parkinson's disease. Application of the classification models on time series data showed the earlier implementation and acceptance of Three Rs principles in the area of cosmetics and skin research, as compared to the area of neurodegenerative disease research. The 3Ranker algorithm is freely available at www.open3r.org; the future goal is to expand this framework to cover multiple research domains and to enable its broad use by researchers, policymakers, funders and ethical review boards, in order to promote the replacement of animal use in research wherever possible.

Extracting cancer concepts from clinical notes using natural language processing: a systematic review

BACKGROUND

Extracting information from free texts using natural language processing (NLP) can save time and reduce the hassle of manually extracting large quantities of data from incredibly complex clinical notes of cancer patients. This study aimed to systematically review studies that used NLP methods to identify cancer concepts from clinical notes automatically.

METHODS

PubMed, Scopus, Web of Science, and Embase were searched for English language papers using a combination of the terms concerning "Cancer", "NLP", "Coding", and "Registries" until June 29, 2021. Two reviewers independently assessed the eligibility of papers for inclusion in the review.

RESULTS

Most of the software programs used for concept extraction reported were developed by the researchers (n = 7). Rule-based algorithms were the most frequently used algorithms for developing these programs. In most articles, the criteria of accuracy (n = 14) and sensitivity (n = 12) were used to evaluate the algorithms. In addition, Systematized Nomenclature of Medicine-Clinical Terms (SNOMED-CT) and Unified Medical Language System (UMLS) were the most commonly used terminologies to identify concepts. Most studies focused on breast cancer (n = 4, 19%) and lung cancer (n = 4, 19%).

CONCLUSION

The use of NLP for extracting the concepts and symptoms of cancer has increased in recent years. The rule-based algorithms are well-liked algorithms by developers. Due to these algorithms' high accuracy and sensitivity in identifying and extracting cancer concepts, we suggested that future studies use these algorithms to extract the concepts of other diseases as well.

Facilitators and barriers of mHealth interventions during the Covid-19 pandemic: systematic review

BACKGROUND

With the spread of Covid-19 disease, health interventions related to the control, prevention, and treatment of this disease and other diseases were given real attention. The purpose of this systematic review is to express facilitators and barriers of using mobile health (mHealth) interventions during the Covid-19 pandemic.

METHODS

In this systematic review, original studies were searched using keywords in the electronic database of PubMed until August 2022. The objectives and outcomes of these studies were extracted. Finally, to identify the facilitators and barriers of mHealth interventions, a qualitative content analysis was conducted based on the strengths, weaknesses, opportunities, and threats (SWOT) analysis method with Atlas.ti 8 software. We evaluated the studies using the Mixed Methods Appraisal Tool (MMAT).

RESULTS

In total, 1598 articles were identified and 55 articles were included in this study. Most of the studies used mobile applications to provide and receive health services during the Covid-19 pandemic (96.4%). The purpose of the applications was to help prevention (17), follow-up (15), treatment (12), and diagnosis (8). Using SWOT analysis, 13 facilitators and 18 barriers to patients' use of mHealth services were identified.

CONCLUSION

Mobile applications are very flexible technologies that can be customized for each person, patient, and population. During the Covid-19 pandemic, the applications designed due to lack of interaction, lack of time, lack of attention to privacy, and non-academic nature have not met their expectations of them.

Machine Learning Models to Predict Future Frailty in Community-Dwelling Middle-Aged and Older Adults: The ELSA Cohort Study

BACKGROUND

Machine learning (ML) models can be used to predict future frailty in the community setting. However, outcome variables for epidemiologic data sets such as frailty usually have an imbalance between categories, that is, there are far fewer individuals classified as frail than as nonfrail, adversely affecting the performance of ML models when predicting the syndrome.

METHODS

A retrospective cohort study with participants (50 years or older) from the English Longitudinal Study of Ageing who were nonfrail at baseline (2008-2009) and reassessed for the frailty phenotype at 4-year follow-up (2012-2013). Social, clinical, and psychosocial baseline predictors were selected to predict frailty at follow-up in ML models (Logistic Regression, Random Forest [RF], Support Vector Machine, Neural Network, K-nearest neighbor, and Naive Bayes classifier).

RESULTS

Of all the 4 378 nonfrail participants at baseline, 347 became frail at follow-up. The proposed combined oversampling and undersampling method to adjust imbalanced data improved the performance of the models, and RF had the best performance, with areas under the receiver-operating characteristic curve and the precision-recall curve of 0.92 and 0.97, respectively, specificity of 0.83, sensitivity of 0.88, and balanced accuracy of 85.5% for balanced data. Age, chair-rise test, household wealth, balance problems, and self-rated health were the most important frailty predictors in most of the models trained with balanced data.

CONCLUSIONS

ML proved useful in identifying individuals who became frail over time, and this result was made possible by balancing the data set. This study highlighted factors that may be useful in the early detection of frailty.

A resilient workforce: patient safety and the workforce response to a cyber-attack on the ICT systems of the national health service in Ireland

BACKGROUND

In May 2021, the Irish public health service was the target of a cyber-attack. The response by the health service resulted in the widespread removal of access to ICT systems. While services including radiology, diagnostics, maternity, and oncology were prioritised for reinstatement, recovery efforts continued for over four months. This study describes the response of health service staff to the loss of ICT systems, and the risk mitigation measures introduced to safely continue health services. The resilience displayed by frontline staff whose rapid and innovative response ensured continuity of safe patient care is explored.

METHODS

To gain an in-depth understanding of staff experiences of the cyber-attack, eight focus groups (n = 36) were conducted. Participants from a diverse range of health services were recruited, including staff from radiology, pathology/laboratories, radiotherapy, maternity, primary care dental services, health and wellbeing, COVID testing, older person's care, and disability services. Thematic Analysis was applied to the data to identify key themes.

RESULTS

The impact of the cyber-attack varied across services depending on the type of care being offered, the reliance on IT systems, and the extent of local IT support. Staff stepped-up to the challenges and quickly developed and implemented innovative solutions, exhibiting great resilience, teamwork and adaptability, with a sharp focus on ensuring patient safety. The cyber-attack resulted in a flattening of the healthcare hierarchy, with shared decision-making at local levels leading to an empowered frontline workforce. However, participants in this study felt the stress placed on staff by the attack was more severe than the cumulative effect of the COVID-19 pandemic.

CONCLUSIONS

Limited contingencies within the health system IT infrastructure - what we call a lack of system resilience - was compensated for by a resilient workforce. Within the context of the prevailing COVID-19 pandemic, this was an enormous burden on a dedicated workforce. The adverse impact of this attack may have long-term and far-reaching consequences for staff wellbeing. Design and investment in a resilient health system must be prioritised.

Machine learning clinical prediction models for acute kidney injury: the impact of baseline creatinine on prediction efficacy

BACKGROUND

There are many Machine Learning (ML) models which predict acute kidney injury (AKI) for hospitalised patients. While a primary goal of these models is to support clinical decision-making, the adoption of inconsistent methods of estimating baseline serum creatinine (sCr) may result in a poor understanding of these models' effectiveness in clinical practice. Until now, the performance of such models with different baselines has not been compared on a single dataset. Additionally, AKI prediction models are known to have a high rate of false positive (FP) events regardless of baseline methods. This warrants further exploration of FP events to provide insight into potential underlying reasons.

OBJECTIVE

The first aim of this study was to assess the variance in performance of ML models using three methods of baseline sCr on a retrospective dataset. The second aim was to conduct an error analysis to gain insight into the underlying factors contributing to FP events.

MATERIALS AND METHODS

The Intensive Care Unit (ICU) patients of the Medical Information Mart for Intensive Care (MIMIC)-IV dataset was used with the KDIGO (Kidney Disease Improving Global Outcome) definition to identify AKI episodes. Three different methods of estimating baseline sCr were defined as (1) the minimum sCr, (2) the Modification of Diet in Renal Disease (MDRD) equation and the minimum sCr and (3) the MDRD equation and the mean of preadmission sCr. For the first aim of this study, a suite of ML models was developed for each baseline and the performance of the models was assessed. An analysis of variance was performed to assess the significant difference between eXtreme Gradient Boosting (XGB) models across all baselines. To address the second aim, Explainable AI (XAI) methods were used to analyse the XGB errors with Baseline 3.

RESULTS

Regarding the first aim, we observed variances in discriminative metrics and calibration errors of ML models when different baseline methods were adopted. Using Baseline 1 resulted in a 14% reduction in the f1 score for both Baseline 2 and Baseline 3. There was no significant difference observed in the results between Baseline 2 and Baseline 3. For the second aim, the FP cohort was analysed using the XAI methods which led to relabelling data with the mean of sCr in 180 to 0 days pre-ICU as the preferred sCr baseline method. The XGB model using this relabelled data achieved an AUC of 0.85, recall of 0.63, precision of 0.54 and f1 score of 0.58. The cohort size was 31,586 admissions, of which 5,473 (17.32%) had AKI.

CONCLUSION

In the absence of a widely accepted method of baseline sCr, AKI prediction studies need to consider the impact of different baseline methods on the effectiveness of ML models and their potential implications in real-world implementations. The utilisation of XAI methods can be effective in providing insight into the occurrence of prediction errors. This can potentially augment the success rate of ML implementation in routine care.

Facilitators and barriers for tuberculosis preventive treatment among patients with latent tuberculosis infection: a qualitative study

BACKGROUND

Various factors influence tuberculosis preventive treatment (TPT) decisions thus it is important to understand the health beliefs and concerns of patients before starting TPT to ensure treatment compliance. This study aims to explore facilitators and barriers for TPT among patients diagnosed with Latent Tuberculosis infection (LTBI) attending six primary healthcare clinics in Selangor, Malaysia.

METHOD

In-depth interviews were conducted face-to-face or via telephone among patients with a clinical diagnosis of LTBI using a semi-structured topic guide developed based on the common-sense model of self-regulation and literature review. Audio recordings of interviews were transcribed verbatim and analysed thematically.

RESULTS

We conducted 26 In-depth interviews; Good knowledge of active tuberculosis (TB) and its associated complications, including the perceived seriousness and transmissibility of active TB, facilitates treatment. LTBI is viewed as a concern when immune status is compromised, thus fostering TPT. However, optimal health is a barrier for TPT. Owing to the lack of knowledge, patients rely on healthcare practitioners (HCPs) to determine their treatment paths. HCPs possessing comprehensive knowledge play a role in facilitating TPT whereas barriers to TPT encompass misinterpretation of tuberculin skin test (TST), inadequate explanation of TST, and apprehensions about potential medication side effects.

CONCLUSIONS

Knowledge of LTBI can influence TPT uptake and patients often entrust their HCPs for treatment decisions. Improving knowledge of LTBI both among patients and HCPs can lead to more effective doctor-patient consultation and consequently boost the acceptance of TPT. Quality assurance should be enhanced to ensure the effective usage of TST as a screening tool.

Transformers for extracting breast cancer information from Spanish clinical narratives

The wide adoption of electronic health records (EHRs) offers immense potential as a source of support for clinical research. However, previous studies focused on extracting only a limited set of medical concepts to support information extraction in the cancer domain for the Spanish language. Building on the success of deep learning for processing natural language texts, this paper proposes a transformer-based approach to extract named entities from breast cancer clinical notes written in Spanish and compares several language models. To facilitate this approach, a schema for annotating clinical notes with breast cancer concepts is presented, and a corpus for breast cancer is developed. Results indicate that both BERT-based and RoBERTa-based language models demonstrate competitive performance in clinical Named Entity Recognition (NER). Specifically, BETO and multilingual BERT achieve F-scores of 93.71% and 94.63%, respectively. Additionally, RoBERTa Biomedical attains an F-score of 95.01%, while RoBERTa BNE achieves an F-score of 94.54%. The findings suggest that transformers can feasibly extract information in the clinical domain in the Spanish language, with the use of models trained on biomedical texts contributing to enhanced results. The proposed approach takes advantage of transfer learning techniques by fine-tuning language models to automatically represent text features and avoiding the time-consuming feature engineering process.

The effect of capacity building evidence-based medicine training on its implementation among healthcare professionals in Southwest Ethiopia: a controlled quasi-experimental outcome evaluation

BACKGROUND

Evidence-based medicine (EBM) bridges research and clinical practice to enhance medical knowledge and improve patient care. However, clinical decisions in many African countries don't base on the best available scientific evidence. Hence, this study aimed to determine the effect of training interventions on background knowledge and awareness of EBM sources, attitude, competence, and practice of EBM among healthcare professionals.

METHOD

We designed a controlled group quasi-experimental pre-post test study to evaluate the effect of capacity-building EBM training. A total of 192 healthcare professionals were recruited in the study (96 from the intervention and 96 from the control group). We used a difference-in-differences (DID) analysis to determine the effect of the training. Along the way, we used a fixed effect panel-data regression model to assess variables that could affect healthcare professionals' practice of EBM. The cut point to determine the significant effect of EBM training on healthcare professionals' background knowledge and awareness of EBM sources, attitude, and competence was at a P-value < 0.05.

RESULT

The DID estimator showed a significant net change of 8.0%, 17.1%, and 11.4% at P < 0.01 on attitude, competence, and practice of EBM, respectively, whereas no significant increment in the background knowledge and awareness of EBM sources. The fixed effect regression model showed that the attitude [OR = 2.288, 95% CI: (1.049, 4.989)], competence [OR = 4.174, 95% CI: 1.984, 8.780)], technical support [OR = 2.222, 95% CI: (1.043, 3.401)], and internet access [OR = 1.984, 95% CI: (1.073, 4.048)] were significantly affected EBM practice.

CONCLUSION

The capacity-building training improved attitude, competence, and EBM practice. Policymakers, government, and other concerned bodies recommended focusing on a well-designed training strategy to enhance the attitude, competence, and practice towards EBM among healthcare professionals. It was also recommended to enhance internet access and set mechanisms to provide technical support at health facilities.

Performance of ChatGPT, human radiologists, and context-aware ChatGPT in identifying AO codes from radiology reports

While radiologists can describe a fracture's morphology and complexity with ease, the translation into classification systems such as the Arbeitsgemeinschaft Osteosynthesefragen (AO) Fracture and Dislocation Classification Compendium is more challenging. We tested the performance of generic chatbots and chatbots aware of specific knowledge of the AO classification provided by a vector-index and compared it to human readers. In the 100 radiological reports we created based on random AO codes, chatbots provided AO codes significantly faster than humans (mean 3.2 s per case vs. 50 s per case, p < .001) though not reaching human performance (max. chatbot performance of 86% correct full AO codes vs. 95% in human readers). In general, chatbots based on GPT 4 outperformed the ones based on GPT 3.5-Turbo. Further, we found that providing specific knowledge substantially enhances the chatbot's performance and consistency as the context-aware chatbot based on GPT 4 provided 71% consistent correct full AO codes for the compared to the 2% consistent correct full AO codes for the generic ChatGPT 4. This provides evidence, that refining and providing specific context to ChatGPT will be the next essential step in harnessing its power.

Exploring the relationships between health literacy, social support, self-efficacy and self-management in adults with multiple chronic diseases

BACKGROUND

Self-management in chronic diseases is essential to slowing disease progression and preventing complications. However, empirical research on the associations of critical factors, such as health literacy, social support, and self-efficacy with self-management in the context of multiple chronic diseases is scarce. This study aimed to investigate these associations and provides insights for healthcare providers to develop effective educational strategies for people with multiple chronic diseases.

METHODS

Using a cross-sectional survey design, adults (n = 600) diagnosed with at least two chronic diseases were conveniently recruited. To measure health literacy, social support, self-efficacy, and chronic disease self-management behaviours, the Health Literacy Questionnaire (HLQ), Medical Outcome Study - Social Support Survey, Self-efficacy in Managing Chronic Disease, and Self-management in Chronic Diseases instruments were utilized respectively. Comorbidity status was assessed using Age-adjusted Charlson Comorbidity Index (ACCI). A generalised linear regression model was used with a backward technique to identify variables associated with self-management.

RESULTS

Participants' mean age was 61 years (SD = 15.3), 46% were female, and most had up to 12 years of education (82.3%). Mean scores for HLQ domains 1-5 varied from 2.61 to 3.24 (possible score 1-4); domains 6-9 from 3.29 to 3.65 (possible score 1-5). The mean scores were 52.7 (SD = 10.4, possible score 0-95), 5.46 (SD = 1.9, possible score 0-10) and 82.1 (SD = 12.4, possible score 30-120) for social support, self-efficacy, and self-management, respectively. Mean ACCI was 6.7 (SD = 2.1). Eight factors (age > 65 years, being female, 4 health literacy domains, greater social support, and higher self-efficacy levels) were significantly associated with greater self-management behaviours while comorbidity status was not. The factors that showed the strongest associations with self-management were critical health literacy domains: appraisal of health information, social support for health, and healthcare provider support.

CONCLUSIONS

Developing critical health literacy abilities is a more effective way to enhance self-management behaviours than relying solely on self-confidence or social support, especially for people with multiple chronic diseases. By facilitating communication and patient education, healthcare providers can help patients improve their critical health literacy, which in turn can enhance their self-management behaviours.

Assessing the transportability of clinical prediction models for cognitive impairment using causal models

BACKGROUND

Machine learning models promise to support diagnostic predictions, but may not perform well in new settings. Selecting the best model for a new setting without available data is challenging. We aimed to investigate the transportability by calibration and discrimination of prediction models for cognitive impairment in simulated external settings with different distributions of demographic and clinical characteristics.

METHODS

We mapped and quantified relationships between variables associated with cognitive impairment using causal graphs, structural equation models, and data from the ADNI study. These estimates were then used to generate datasets and evaluate prediction models with different sets of predictors. We measured transportability to external settings under guided interventions on age, APOE ε4, and tau-protein, using performance differences between internal and external settings measured by calibration metrics and area under the receiver operating curve (AUC).

RESULTS

Calibration differences indicated that models predicting with causes of the outcome were more transportable than those predicting with consequences. AUC differences indicated inconsistent trends of transportability between the different external settings. Models predicting with consequences tended to show higher AUC in the external settings compared to internal settings, while models predicting with parents or all variables showed similar AUC.

CONCLUSIONS

We demonstrated with a practical prediction task example that predicting with causes of the outcome results in better transportability compared to anti-causal predictions when considering calibration differences. We conclude that calibration performance is crucial when assessing model transportability to external settings.

Natural language processing in narrative breast radiology reporting in University Malaya Medical Centre

Radiology reporting is narrative, and its content depends on the clinician's ability to interpret the images accurately. A tertiary hospital, such as anonymous institute, focuses on writing reports narratively as part of training for medical personnel. Nevertheless, free-text reports make it inconvenient to extract information for clinical audits and data mining. Therefore, we aim to convert unstructured breast radiology reports into structured formats using natural language processing (NLP) algorithm. This study used 327 de-identified breast radiology reports from the anonymous institute. The radiologist identified the significant data elements to be extracted. Our NLP algorithm achieved 97% and 94.9% accuracy in training and testing data, respectively. Henceforth, the structured information was used to build the predictive model for predicting the value of the BIRADS category. The model based on random forest generated the highest accuracy of 92%. Our study not only fulfilled the demands of clinicians by enhancing communication between medical personnel, but it also demonstrated the usefulness of mineable structured data in yielding significant insights.

Weakly supervised spatial relation extraction from radiology reports

Objective

Weak supervision holds significant promise to improve clinical natural language processing by leveraging domain resources and expertise instead of large manually annotated datasets alone. Here, our objective is to evaluate a weak supervision approach to extract spatial information from radiology reports.

Materials and Methods

Our weak supervision approach is based on data programming that uses rules (or labeling functions) relying on domain-specific dictionaries and radiology language characteristics to generate weak labels. The labels correspond to different spatial relations that are critical to understanding radiology reports. These weak labels are then used to fine-tune a pretrained Bidirectional Encoder Representations from Transformers (BERT) model.

Results

Our weakly supervised BERT model provided satisfactory results in extracting spatial relations without manual annotations for training (spatial trigger F1: 72.89, relation F1: 52.47). When this model is further fine-tuned on manual annotations (relation F1: 68.76), performance surpasses the fully supervised state-of-the-art.

Discussion

To our knowledge, this is the first work to automatically create detailed weak labels corresponding to radiological information of clinical significance. Our data programming approach is (1) adaptable as the labeling functions can be updated with relatively little manual effort to incorporate more variations in radiology language reporting formats and (2) generalizable as these functions can be applied across multiple radiology subdomains in most cases.

Conclusions

We demonstrate a weakly supervision model performs sufficiently well in identifying a variety of relations from radiology text without manual annotations, while exceeding state-of-the-art results when annotated data are available.

Using multi-criteria decision analysis to describe stakeholder preferences for new quality improvement initiatives that could optimise prescribing in England

Background

Hospital decision-makers have limited resources to implement quality improvement projects. To decide which interventions to take forward, trade-offs must be considered that inevitably turn on stakeholder preferences. The multi-criteria decision analysis (MCDA) approach could make this decision process more transparent.

Method

An MCDA was conducted to rank-order four types of interventions that could optimise medication use in England's National Healthcare System (NHS) hospitals, including Computerised Interface, Built Environment, Written Communication, and Face-to-Face Interactions. Initially, a core group of quality improvers (N = 10) was convened to determine criteria that could influence which interventions are taken forward according to the Consolidated Framework for Implementation Research. Next, to determine preference weightings, a preference survey was conducted with a diverse group of quality improvers (N = 356) according to the Potentially All Pairwise Ranking of All Possible Alternatives method. Then, rank orders of four intervention types were calculated according to models with criteria unweighted and weighted according to participant preferences using an additive function. Uncertainty was estimated by probabilistic sensitivity analysis using 1,000 Monte Carlo Simulation iterations.

Results

The most important criteria influencing what interventions were preferred was whether they addressed "patient needs" (17.6%)' and their financial "cost (11.5%)". The interventions' total scores (unweighted score out of 30 | weighted out of 100%) were: Computerised Interface (25 | 83.8%), Built Environment (24 | 79.6%), Written Communication (22 | 71.6%), and Face-to-Face (22 | 67.8%). The probabilistic sensitivity analysis revealed that the Computerised Interface would be the most preferred intervention over various degrees of uncertainty.

Conclusions

An MCDA was conducted to rank order intervention types that stand to increase medication optimisation across hospitals in England. The top-ranked intervention type was the Computerised Interface. This finding does not imply Computerised Interface interventions are the most effective interventions but suggests that successfully implementing lower-ranked interventions may require more conversations that acknowledge stakeholder concerns.

Optimizing healthcare system by amalgamation of text processing and deep learning: a systematic review

The explosion of clinical textual data has drawn the attention of researchers. Owing to the abundance of clinical data, it is becoming difficult for healthcare professionals to take real-time measures. The tools and methods are lacking when compared to the amount of clinical data generated every day. This review aims to survey the text processing pipeline with deep learning methods such as CNN, RNN, LSTM, and GRU in the healthcare domain and discuss various applications such as clinical concept detection and extraction, medically aware dialogue systems, sentiment analysis of drug reviews shared online, clinical trial matching, and pharmacovigilance. In addition, we highlighted the major challenges in deploying text processing with deep learning to clinical textual data and identified the scope of research in this domain. Furthermore, we have discussed various resources that can be used in the future to optimize the healthcare domain by amalgamating text processing and deep learning.

Ontology-driven and weakly supervised rare disease identification from clinical notes

BACKGROUND

Computational text phenotyping is the practice of identifying patients with certain disorders and traits from clinical notes. Rare diseases are challenging to be identified due to few cases available for machine learning and the need for data annotation from domain experts.

METHODS

We propose a method using ontologies and weak supervision, with recent pre-trained contextual representations from Bi-directional Transformers (e.g. BERT). The ontology-driven framework includes two steps: (i) Text-to-UMLS, extracting phenotypes by contextually linking mentions to concepts in Unified Medical Language System (UMLS), with a Named Entity Recognition and Linking (NER+L) tool, SemEHR, and weak supervision with customised rules and contextual mention representation; (ii) UMLS-to-ORDO, matching UMLS concepts to rare diseases in Orphanet Rare Disease Ontology (ORDO). The weakly supervised approach is proposed to learn a phenotype confirmation model to improve Text-to-UMLS linking, without annotated data from domain experts. We evaluated the approach on three clinical datasets, MIMIC-III discharge summaries, MIMIC-III radiology reports, and NHS Tayside brain imaging reports from two institutions in the US and the UK, with annotations.

RESULTS

The improvements in the precision were pronounced (by over 30% to 50% absolute score for Text-to-UMLS linking), with almost no loss of recall compared to the existing NER+L tool, SemEHR. Results on radiology reports from MIMIC-III and NHS Tayside were consistent with the discharge summaries. The overall pipeline processing clinical notes can extract rare disease cases, mostly uncaptured in structured data (manually assigned ICD codes).

CONCLUSION

The study provides empirical evidence for the task by applying a weakly supervised NLP pipeline on clinical notes. The proposed weak supervised deep learning approach requires no human annotation except for validation and testing, by leveraging ontologies, NER+L tools, and contextual representations. The study also demonstrates that Natural Language Processing (NLP) can complement traditional ICD-based approaches to better estimate rare diseases in clinical notes. We discuss the usefulness and limitations of the weak supervision approach and propose directions for future studies.

Machine learning to predict virological failure among HIV patients on antiretroviral therapy in the University of Gondar Comprehensive and Specialized Hospital, in Amhara Region, Ethiopia, 2022

BACKGROUND

Treatment with effective antiretroviral therapy (ART) reduces viral load as well as HIV-related morbidity and mortality in HIV-positive patients. Despite the expanded availability of antiretroviral therapy around the world, virological failure remains a serious problem for HIV-positive patients. Thus, Machine learning predictive algorithms have the potential to improve the quality of care and predict the needs of HIV patients by analyzing huge amounts of data, and enhancing prediction capabilities. This study used different machine learning classification algorithms to predict the features that cause virological failure in HIV-positive patients.

METHOD

An institution-based secondary data was used to conduct patients who were on antiretroviral therapy at the University of Gondar Comprehensive and Specialized Hospital from January 2020 to May 2022. Patients' data were extracted from the electronic database using a structured checklist and imported into Python version three software for data pre-processing and analysis. Then, seven supervised classification machine-learning algorithms for model development were trained. The performances of the predictive models were evaluated using accuracy, sensitivity, specificity, precision, f1-score, and AUC. Association rule mining was used to generate the best rule for the association between independent features and the target feature.

RESULT

Out of 5264 study participants, 1893 (35.06%) males and 3371 (64.04%) females were included. The random forest classifier (sensitivity = 1.00, precision = 0.987, f1-score = 0.993, AUC = 0.9989) outperformed in predicting virological failure among all selected classifiers. Random forest feature importance and association rules identified the top eight predictors (Male, younger age, longer duration on ART, not taking CPT, not taking TPT, secondary educational status, TDF-3TC-EFV, and low CD4 counts) of virological failure based on the importance ranking, and the CD-4 count was recognized as the most important predictor feature.

CONCLUSION

The random forest classifier outperformed in predicting and identifying the relevant predictors of virological failure. The results of this study could be very helpful to health professionals in determining the optimal virological outcome.

Demystifying the Role of Natural Language Processing (NLP) in Smart City Applications: Background, Motivation, Recent Advances, and Future Research Directions

Smart cities provide an efficient infrastructure for the enhancement of the quality of life of the people by aiding in fast urbanization and resource management through sustainable and scalable innovative solutions. The penetration of Information and Communication Technology (ICT) in smart cities has been a major contributor to keeping up with the agility and pace of their development. In this paper, we have explored Natural Language Processing (NLP) which is one such technical discipline that has great potential in optimizing ICT processes and has so far been kept away from the limelight. Through this study, we have established the various roles that NLP plays in building smart cities after thoroughly analyzing its architecture, background, and scope. Subsequently, we present a detailed description of NLP's recent applications in the domain of smart healthcare, smart business, and industry, smart community, smart media, smart research, and development as well as smart education accompanied by NLP's open challenges at the very end. This work aims to throw light on the potential of NLP as one of the pillars in assisting the technical advancement and realization of smart cities.

Developing Artificial Intelligence Models for Extracting Oncologic Outcomes from Japanese Electronic Health Records

INTRODUCTION

A framework that extracts oncological outcomes from large-scale databases using artificial intelligence (AI) is not well established. Thus, we aimed to develop AI models to extract outcomes in patients with lung cancer using unstructured text data from electronic health records of multiple hospitals.

METHODS

We constructed AI models (Bidirectional Encoder Representations from Transformers [BERT], Naïve Bayes, and Longformer) for tumor evaluation using the University of Miyazaki Hospital (UMH) database. This data included both structured and unstructured data from progress notes, radiology reports, and discharge summaries. The BERT model was applied to the Life Data Initiative (LDI) data set of six hospitals. Study outcomes included the performance of AI models and time to progression of disease (TTP) for each line of treatment based on the treatment response extracted by AI models.

RESULTS

For the UMH data set, the BERT model exhibited higher precision accuracy compared to the Naïve Bayes or the Longformer models, respectively (precision [0.42 vs. 0.47 or 0.22], recall [0.63 vs. 0.46 or 0.33] and F1 scores [0.50 vs. 0.46 or 0.27]). When this BERT model was applied to LDI data, prediction accuracy remained quite similar. The Kaplan-Meier plots of TTP (months) showed similar trends for the first (median 14.9 [95% confidence interval 11.5, 21.1] and 16.8 [12.6, 21.8]), the second (7.8 [6.7, 10.7] and 7.8 [6.7, 10.7]), and the later lines of treatment for the predicted data by the BERT model and the manually curated data.

CONCLUSION

We developed AI models to extract treatment responses in patients with lung cancer using a large EHR database; however, the model requires further improvement.

Improving classification of correct and incorrect protein-protein docking models by augmenting the training set

Motivation

Protein-protein interactions drive many relevant biological events, such as infection, replication and recognition. To control or engineer such events, we need to access the molecular details of the interaction provided by experimental 3D structures. However, such experiments take time and are expensive; moreover, the current technology cannot keep up with the high discovery rate of new interactions. Computational modeling, like protein-protein docking, can help to fill this gap by generating docking poses. Protein-protein docking generally consists of two parts, sampling and scoring. The sampling is an exhaustive search of the tridimensional space. The caveat of the sampling is that it generates a large number of incorrect poses, producing a highly unbalanced dataset. This limits the utility of the data to train machine learning classifiers.

Results

Using weak supervision, we developed a data augmentation method that we named hAIkal. Using hAIkal, we increased the labeled training data to train several algorithms. We trained and obtained different classifiers; the best classifier has 81% accuracy and 0.51 Matthews' correlation coefficient on the test set, surpassing the state-of-the-art scoring functions.

Availability and implementation

Docking models from Benchmark 5 are available at https://doi.org/10.5281/zenodo.4012018. Processed tabular data are available at https://repository.kaust.edu.sa/handle/10754/666961. Google colab is available at https://colab.research.google.com/drive/1vbVrJcQSf6\_C3jOAmZzgQbTpuJ5zC1RP?usp=sharing.

Supplementary information

Supplementary data are available at Bioinformatics Advances online.

Application of Deep Learning in Generating Structured Radiology Reports: A Transformer-Based Technique

Since radiology reports needed for clinical practice and research are written and stored in free-text narrations, extraction of relative information for further analysis is difficult. In these circumstances, natural language processing (NLP) techniques can facilitate automatic information extraction and transformation of free-text formats to structured data. In recent years, deep learning (DL)-based models have been adapted for NLP experiments with promising results. Despite the significant potential of DL models based on artificial neural networks (ANN) and convolutional neural networks (CNN), the models face some limitations to implement in clinical practice. Transformers, another new DL architecture, have been increasingly applied to improve the process. Therefore, in this study, we propose a transformer-based fine-grained named entity recognition (NER) architecture for clinical information extraction. We collected 88 abdominopelvic sonography reports in free-text formats and annotated them based on our developed information schema. The text-to-text transfer transformer model (T5) and Scifive, a pre-trained domain-specific adaptation of the T5 model, were applied for fine-tuning to extract entities and relations and transform the input into a structured format. Our transformer-based model in this study outperformed previously applied approaches such as ANN and CNN models based on ROUGE-1, ROUGE-2, ROUGE-L, and BLEU scores of 0.816, 0.668, 0.528, and 0.743, respectively, while providing an interpretable structured report.

Not so weak PICO: leveraging weak supervision for participants, interventions, and outcomes recognition for systematic review automation

Objective

The aim of this study was to test the feasibility of PICO (participants, interventions, comparators, outcomes) entity extraction using weak supervision and natural language processing.

Methodology

We re-purpose more than 127 medical and nonmedical ontologies and expert-generated rules to obtain multiple noisy labels for PICO entities in the evidence-based medicine (EBM)-PICO corpus. These noisy labels are aggregated using simple majority voting and generative modeling to get consensus labels. The resulting probabilistic labels are used as weak signals to train a weakly supervised (WS) discriminative model and observe performance changes. We explore mistakes in the EBM-PICO that could have led to inaccurate evaluation of previous automation methods.

Results

In total, 4081 randomized clinical trials were weakly labeled to train the WS models and compared against full supervision. The models were separately trained for PICO entities and evaluated on the EBM-PICO test set. A WS approach combining ontologies and expert-generated rules outperformed full supervision for the participant entity by 1.71% macro-F1. Error analysis on the EBM-PICO subset revealed 18-23% erroneous token classifications.

Discussion

Automatic PICO entity extraction accelerates the writing of clinical systematic reviews that commonly use PICO information to filter health evidence. However, PICO extends to more entities-PICOS (S-study type and design), PICOC (C-context), and PICOT (T-timeframe) for which labelled datasets are unavailable. In such cases, the ability to use weak supervision overcomes the expensive annotation bottleneck.

Conclusions

We show the feasibility of WS PICO entity extraction using freely available ontologies and heuristics without manually annotated data. Weak supervision has encouraging performance compared to full supervision but requires careful design to outperform it.

Adverse drug event detection using natural language processing: A scoping review of supervised learning methods

To reduce adverse drug events (ADEs), hospitals need a system to support them in monitoring ADE occurrence routinely, rapidly, and at scale. Natural language processing (NLP), a computerized approach to analyze text data, has shown promising results for the purpose of ADE detection in the context of pharmacovigilance. However, a detailed qualitative assessment and critical appraisal of NLP methods for ADE detection in the context of ADE monitoring in hospitals is lacking. Therefore, we have conducted a scoping review to close this knowledge gap, and to provide directions for future research and practice. We included articles where NLP was applied to detect ADEs in clinical narratives within electronic health records of inpatients. Quantitative and qualitative data items relating to NLP methods were extracted and critically appraised. Out of 1,065 articles screened for eligibility, 29 articles met the inclusion criteria. Most frequent tasks included named entity recognition (n = 17; 58.6%) and relation extraction/classification (n = 15; 51.7%). Clinical involvement was reported in nine studies (31%). Multiple NLP modelling approaches seem suitable, with Long Short Term Memory and Conditional Random Field methods most commonly used. Although reported overall performance of the systems was high, it provides an inflated impression given a steep drop in performance when predicting the ADE entity or ADE relation class. When annotating corpora, treating an ADE as a relation between a drug and non-drug entity seems the best practice. Future research should focus on semi-automated methods to reduce the manual annotation effort, and examine implementation of the NLP methods in practice.

Which are the vital factors of mobile personal health records applications that promote continued usage? A perspective on technology fit and social capital

Introduction

With the widespread use of mobile devices and the rapid development of mobile networks, connecting mobile personal health record (mPHR) apps to wearable devices to collect personal health data for analysis and community activities has become a trend for health promotion. Therefore, the present study aims to explore the vital factors that impact the sustained usage of mPHR apps.

Objective

In this study, we identified social lock-in as a major research gap in the current era of social media and the Internet. Therefore, to explore the effects of mPHR apps on continued app usage intention, we combined technology fit (individual-technology, synchronicity-technology, and task-technology fit) and social capital (structural, relational, and cognitive capital) to develop a novel study model.

Methods

The purpose of this research is to investigate the willingness to participate in the mPHR apps. It collected 565 valid users' responses through the online questionnaire with a structural equation modeling approach.

Results

That technology and social lock-in significantly affected the willingness of users to continue using mPHR apps (β  =  0.38, P < 0.001) and that the effects of social lock-in (β  =  0.38, P < 0.001) were more pronounced than those of technology lock-in (β  =  0.22, P < 0.001).

Conclusions

The technology and social lock-in generated by technology fit and social capital had positive effects on continued app usage and the effects of both types of lock-in on continued app usage varied among different participant groups.

Recent applications of machine learning and deep learning models in the prediction, diagnosis, and management of diabetes: a comprehensive review

Diabetes as a metabolic illness can be characterized by increased amounts of blood glucose. This abnormal increase can lead to critical detriment to the other organs such as the kidneys, eyes, heart, nerves, and blood vessels. Therefore, its prediction, prognosis, and management are essential to prevent harmful effects and also recommend more useful treatments. For these goals, machine learning algorithms have found considerable attention and have been developed successfully. This review surveys the recently proposed machine learning (ML) and deep learning (DL) models for the objectives mentioned earlier. The reported results disclose that the ML and DL algorithms are promising approaches for controlling blood glucose and diabetes. However, they should be improved and employed in large datasets to affirm their applicability.

Evaluation of direct-to-angiography suite (DTAS) and conventional clinical pathways in stroke care: a simulation study

BACKGROUND

Rapid time to reperfusion is essential to minimize morbidity and mortality in acute ischemic stroke due to large vessel occlusion (LVO). We aimed to evaluate the workflow times when utilizing a direct-to-angiography suite (DTAS) pathway for patients with suspected stroke presenting at a comprehensive stroke center compared with a conventional CT pathway.

METHODS

We developed a discrete-event simulation (DES) model to evaluate DTAS workflow timelines compared with a conventional CT pathway, varying the admission NIHSS score treatment eligibility criteria. Model parameters were estimated based on 2 year observational data from our institution. Sensitivity analyses of simulation parameters were performed to assess the impact of patient volume and baseline utilization of angiography suites on workflow times utilizing DTAS.

RESULTS

Simulation modeling of stroke patients (SimStroke) demonstrated door-to-reperfusion time savings of 0.2-3.5 min (p=0.05) for a range of DTAS eligibility criteria (ie, last known well to arrival <6 hours and National Institutes of Health Stroke Scale ≥6-11), when compared with the conventional stroke care pathway. Sensitivity analyses revealed that DTAS time savings is highly dependent on baseline utilization of angiography suites.

CONCLUSIONS

The results of the SimStroke model showed comparable time intervals for door-to-reperfusion for DTAS compared with a conventional stroke care pathway. However, the DTAS pathway was very sensitive to baseline angiography suite utilization, with even a 10% increase eliminating the advantages of DTAS compared with the conventional pathway. Given the minimal time savings modeled here, further investigation of implementing the DTAS pathway in clinical care is warranted.

The openVA Toolkit for Verbal Autopsies

Verbal autopsy (VA) is a survey-based tool widely used to infer cause of death (COD) in regions without complete-coverage civil registration and vital statistics systems. In such settings, many deaths happen outside of medical facilities and are not officially documented by a medical professional. VA surveys, consisting of signs and symptoms reported by a person close to the decedent, are used to infer the COD for an individual, and to estimate and monitor the COD distribution in the population. Several classification algorithms have been developed and widely used to assign causes of death using VA data. However, the incompatibility between different idiosyncratic model implementations and required data structure makes it difficult to systematically apply and compare different methods. The openVA package provides the first standardized framework for analyzing VA data that is compatible with all openly available methods and data structure. It provides an open-source, R implementation of several most widely used VA methods. It supports different data input and output formats, and customizable information about the associations between causes and symptoms. The paper discusses the relevant algorithms, their implementations in R packages under the openVA suite, and demonstrates the pipeline of model fitting, summary, comparison, and visualization in the R environment.