Clinical Decision Support Model Research Articles

Enhancing Periodontal Treatment Through the Integration of Deep Learning-Based Detection with Bayesian Network Models.

This study incorporated deep learning for periodontal disease detection into a Bayesian network (BN) clinical decision support model for comprehensive periodontal care. BN structure and probabilities were based on clinical data and Faster R-CNN-detected radiographic images. Receiver operating characteristic curve analysis confirmed the model's high accuracy in treatment plan recommendations.

Data sources (LLM) for a Clinical Decision Support Model (SSDC) using a Healthcare Interoperability Resources (HL7-FHIR) platform for an ICU Ecosystem

Data sources (LLM) for a Clinical Decision Support Model (SSDC) using a Healthcare Interoperability Resources (HL7-FHIR) platform for an ICU Ecosystem

Primary care clinician engagement in implementing a machine-learning algorithm for targeted screening of familial hypercholesterolemia

ObjectiveTo assess the impact of a multi-pronged educational approach on the knowledge, attitudes, and behaviors regarding Familial Hypercholesterolemia (FH) management at a large academic medical center with the aim of empowering primary care clinicians (PCC) to diagnose and treat FH. MethodsA comprehensive educational program for PCCs on FH management was developed and piloted from July 2022 to March 2024. Components of our intervention included: 1. Implementation of a novel clinical decision support tool in the electronic medical record for FH management, 2. Development and dissemination of an interactive educational website focused on FH and its management, 3. Delivery of virtual instructional sessions to increase awareness of the tool, provide education on its use, and obtain support from institutional leadership, and 4. Direct outreach to a pilot subset of PCCs whose patients had been detected using the validated FIND FH® machine learning algorithm. Participating clinicians were surveyed at baseline before the intervention and after the educational session. Results70 PCC consented to participate in the study with a survey completion rate of 79 % (n = 55) and 42 % (n = 23) for the baseline and follow-up surveys, respectively. Objective PCC knowledge scores improved from 40 to 65 % of responders correctly responding to at least 2/3rds of survey questions. Despite the fact that 87 % identified PCC's as most effective for early detection of FH, 100 % of PCCs who received direct outreach chose to defer care to an outpatient cardiologist over pursuing workup in the primary care setting. ConclusionEmpowering PCCs in management of FH serves as a key strategy in addressing this underdiagnosed and undertreated potentially life-threatening condition. A systems-based approach to addressing these aims may include leveraging EMR-based clinical decision support models and cross-disciplinary educational partnerships with medical specialists.

Prediction of The Patients at Risk for Development Hematoma After Percutaneous Coronary Angiography: A Nursing Decision Support Model Pilot Study

The study aimed to develop a nursing clinical decision support model using the machine learning method, which is one of the important fields today, to identify patients with risk of hematoma development after Percutaneous Coronary Intervention and to help plan appropriate nursing interventions. In this study, the data of 100 patients with myocardial infarction was used in the development of the decision support model. R open-source programming language was used for statistical analysis of the data and the random forest method, one of the machine learning methods was used for the development of the model. The result of this pilot study, a nursing decision support model with a sensitivity of 69% and a specificity of 64% was developed with the Random forest method using 24 features regarding the demographic, laboratory, and percutaneous coronary intervention procedures of the patients.

DermCDSM: Clinical Decision Support Model for Dermatosis Using Systematic Approaches of Machine Learning and Deep Learning

DermCDSM: Clinical Decision Support Model for Dermatosis Using Systematic Approaches of Machine Learning and Deep Learning

Predicting sequenced dental treatment plans from electronic dental records using deep learning

BackgroundDesigning appropriate clinical dental treatment plans is an urgent need because a growing number of dental patients are suffering from partial edentulism with the population getting older. ObjectivesThe aim of this study is to predict sequential treatment plans from electronic dental records. MethodsWe construct a clinical decision support model, MultiTP, explores the unique topology of teeth information and the variation of complicated treatments, integrates deep learning models (convolutional neural network and recurrent neural network) adaptively, and embeds the attention mechanism to produce optimal treatment plans. ResultsMultiTP shows its promising performance with an AUC of 0.9079 and an F score of 0.8472 over five treatment plans. The interpretability analysis also indicates its capability in mining clinical knowledge from the textual data. ConclusionsMultiTP's novel problem formulation, neural network framework, and interpretability analysis techniques allow for broad applications of deep learning in dental healthcare, providing valuable support for predicting dental treatment plans in the clinic and benefiting dental patients. Clinical implicationsThe MultiTP is an efficient tool that can be implemented in clinical practice and integrated into the existing EDR system. By predicting treatment plans for partial edentulism, the model will help dentists improve their clinical decisions.

Machine learning-based models for the prediction of breast cancer recurrence risk

Breast cancer is the most common malignancy diagnosed in women worldwide. The prevalence and incidence of breast cancer is increasing every year; therefore, early diagnosis along with suitable relapse detection is an important strategy for prognosis improvement. This study aimed to compare different machine algorithms to select the best model for predicting breast cancer recurrence. The prediction model was developed by using eleven different machine learning (ML) algorithms, including logistic regression (LR), random forest (RF), support vector classification (SVC), extreme gradient boosting (XGBoost), gradient boosting decision tree (GBDT), decision tree, multilayer perceptron (MLP), linear discriminant analysis (LDA), adaptive boosting (AdaBoost), Gaussian naive Bayes (GaussianNB), and light gradient boosting machine (LightGBM), to predict breast cancer recurrence. The area under the curve (AUC), accuracy, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and F1 score were used to evaluate the performance of the prognostic model. Based on performance, the optimal ML was selected, and feature importance was ranked by Shapley Additive Explanation (SHAP) values. Compared to the other 10 algorithms, the results showed that the AdaBoost algorithm had the best prediction performance for successfully predicting breast cancer recurrence and was adopted in the establishment of the prediction model. Moreover, CA125, CEA, Fbg, and tumor diameter were found to be the most important features in our dataset to predict breast cancer recurrence. More importantly, our study is the first to use the SHAP method to improve the interpretability of clinicians to predict the recurrence model of breast cancer based on the AdaBoost algorithm. The AdaBoost algorithm offers a clinical decision support model and successfully identifies the recurrence of breast cancer.

Machine Learning Approach for Thyroid Cancer Diagnosis Using Clinical Data

Objective: With an early diagnosis of thyroid cancer, one of the world's most significant health issues, it is feasible to treat the nodules before the spread of malignant thyroid gland cells. It has become crucial to develop models for predicting thyroid cancer. In light of this, the purpose of this study is to develop a clinical decision support model using the Bagged CART model, a machine learning (ML) model for the prediction of thyroid cancer. Methods: Between 2010 and 2012, 724 patients who applied to China Median University Shengjing Hospital comprised the study's data set. The dataset comprises information on nodule malignancies, demographic characteristics, ultrasound characteristics, and blood test results for all patients who underwent thyroidectomy. Using this open-access data set, the Bagged CART modeling technique was applied. Negative predictive value (NPV), specificity (Spe), balanced accuracy (BACC), positive predictive value (PPV), accuracy (ACC), sensitivity (Sen), and F1-score performance metrics were used to evaluate the model's predictive performance. In addition, a 10-fold cross-validation method was used to determine the validity of the model. In addition, variable importance was established, which reveals how much the input variables impact the output variable. Results: ACC, BACC, Sen, Spe, PPV, NPV, and F1-score obtained from the model performance metrics were calculated to 99.1%, 98.7%, 99.7%, 97.7%, 99.1%, 99.2%, and 99.4%, respectively, as a result of modeling. According to the variable importance values that were acquired for the input variables in the dataset that was investigated in this study, the seven variable that hold the greatest significance are as follows: size, TSH, blood flow: size, TSH, blood flow: enriched, multilateral: yes, FT4, site: isthmus, and age, in that order. Conclusion: As a result, the Bagged CART model was found to be effective at predicting thyroid cancer based on the findings of this study. In addition, in this study, risk factors for thyroid cancer were evaluated and their importance values were given. With these results, the decision-making process about the disease will be able to accelerate and thus, it will be able to effective in preventive medicine practices.

Visual Assistance in Development and Validation of Bayesian Networks for Clinical Decision Support.

The development and validation of Clinical Decision Support Models (CDSM) based on Bayesian networks (BN) is commonly done in a collaborative work between medical researchers providing the domain expertise and computer scientists developing the decision support model. Although modern tools provide facilities for data-driven model generation, domain experts are required to validate the accuracy of the learned model and to provide expert knowledge for fine-tuning it while computer scientists are needed to integrate this knowledge in the learned model (hybrid modeling approach). This generally time-expensive procedure hampers CDSM generation and updating. To address this problem, we developed a novel interactive visual approach allowing medical researchers with less knowledge in CDSM to develop and validate BNs based on domain specific data mainly independently and thus, diminishing the need for an additional computer scientist. In this context, we abstracted and simplified the common workflow in BN development as well as adjusted the workflow to medical experts needs. We demonstrate our visual approach with data of endometrial cancer patients and evaluated it with six medical researchers who are domain experts in the gynecological field.

CLINICAL DECISION SUPPORT SYSTEM FOR EARLY DIAGNOSIS OF HEART ATTACK USING MACHINE LEARNING METHODS

Heart attack which is the main cause of death for both men and women is the leader among deaths due to heart diseases. Therefore, early diagnosis is very important for patients who are having a heart attack. Therefore, the study aimed to develop a clinical decision support system for the diagnosis of a heart attack to help physicians. In the study, variables were obtained accompanied by physicians by statistical analysis methods, where the optimum variables were selected from these variables considering the patient’s unconscious state in some cases. Different decision models were developed using probit regression, decision tree, SVM, and ANN methods. As a result, the developed clinical decision support models for heart attack diagnosis were compared and evaluated. Consequently, the best diagnosis model was obtained using ANN with selected variables. In addition to these, the proposed study is significantly noticed with a sensitivity of 98% and specificity of 93.7% for heart attack diagnosis with optimum variables compared to similar studies in the literature. By using the proposed decision support system, it is possible to determine whether a patient has a heart attack or not and help the physician in the process of diagnosis of a heart attack.

Clinical decision support models for oropharyngeal cancer treatment: design and evaluation of a multi-stage knowledge abstraction and formalization process

PurposeTreatment decisions in oncology are demanding and affect survival, general health, and quality of life. Expert systems can handle the complexity of the oncological field. We propose the application of a hybrid modeling approach for decision support models consisting of expert-based implementation of a decision model structure and machine-learning (ML) based parameter generation. We demonstrate our approach for the treatment of oropharyngeal cancer.MethodsWe created a clinical decision model based on Bayesian Networks and iteratively optimized its characteristics using structured knowledge engineering approaches. We combined manual adaptation of individual concepts with automatic learning of parameters and causalities. Using data from 94 patient records, we targeted the needed objectivity and clinical significance.ResultsIn three iteration steps, we assessed the model with cross-validations. The initial aggregated accuracy of 0.529 could be increased to 0.883 in the final version. The predictive rates of the target nodes range from 0.557 to 0.960.ConclusionCombining different methodological approaches requires balancing the complexity of the clinical subject matter with the amount of information available in the dataset for ML application. Our method showed promising results because flaws of one approach can be overcome by the other approach. However, technical integrability as well as clinical acceptance must always be ensured.

650-P: Artificial Intelligence and Individualized Optimal Glycemic Target in ICU Patients

Introduction: Deep learning methods, commonly called artificial intelligence (AI) , have been widely adopted in medical research. Reinforcement learning is a particularly promising AI tool that can generate optimal strategies based on non-optimized training data. The aim of the currentstudy was to develop and validate a reinforcement learning algorithm for determining optimal glycemic targets for sepsis patients in the ICU. Methods: We developed and validated a clinical decision support model that provided an individualized daily glycemic target for each critically ill patient based on the retrospective analysis of two independent ICU databases, Medical Information for Mart for Intensive Care III (MIMIC-III) and the eICU Collaborative Research Database (eICU-CRD) . We used a Markov decision process (MDP) to formulate sequential decision-making. The discrete state was generated by quantizing all patients’ longitudinal health records using a k-means clustering algorithm. To evaluate clinician policy, we used the temporal-difference (TD) learning method. To learn AI policy, we generated a random action policy, updated the value function, and used the value function to update the policy. Results: Average return per patient was significantly higher in survivors than non-survivors, and a significant inverse relationship was observed between AI performance return and estimated 90-day mortality. Contrary to the current guidelines that recommend maintaining a glycemic range of 140-180 mg during ICU admission, AI most often recommended a glycemic range of 120-139 mg/dL for ICU patients. Analysis of mortality rate according to mean glucose level during ICU admission revealed that the greater the difference between AI recommendation and real glycemic control range, the higher the mortality rate. Conclusion: These results suggest that tighter glycemic control, guided by a similar AI policy for daily individualized optimal glycemic target, may help improve ICU patient survival. Disclosure J.Yun: None. G.Lee: None. J.Kim: None. S.Ko: None. Y.Ahn: None. D.Kim: n/a. K.Song: n/a.

A Clinical Decision Support System Based on Machine Learning For The Prediction of Diabetes Mellitus.

Aim: Early diagnosis of diabetes mellitus (DM), one of the most important health problems worldwide, and taking necessary steps are very important. Therefore, it has become very important to develop models for the prediction of the disease. The aim of this study is to create a clinical decision support model with Stochastic Gradient Boosting, a machine learning model for DM prediction. Material and Methods: In the study, modeling was done with the Stochastic Gradient Boosting method using an open access data set including the factors associated with DM. Model results were evaluated with accuracy, balanced accuracy, sensitivity, selectivity, positive predictive value, negative predictive value, and F1-score performance metrics. In addition, 5-fold cross-validation method was used in the modeling phase. Finally, variable importance values were obtained by modeling. Results: Accuracy, balanced accuracy, sensitivity, specificity, positive predictive value, negative predictive value, and F1 score from by Stochastic Gradient Boosting modeling were 93.6%, 92.8%, 91.7%, 93.9%, 73.3%, 98.4%, and 81.5%, respectively. According to the variable importance values obtained for the input variables in the data set examined in this study, the most important variables are glucose, age, systolic BP, cholesterol, chol/HDL, BMI, height, waist/hip, HDL, waist, weight, diastolic BP, hip, and gender: male. Conclusion: In the current study, it was seen that the ML model applied with the results obtained can predict diabetes. Addition, according to the results of the relevant model, the most important risk factors for DM were determined and given in degrees of importance of the risk factors. With these results, necessary precautions can be taken for the disease at early levels.

A Decision Support Model for Predicting Avoidable Re-Hospitalization of Breast Cancer Patients in Kenyatta National Hospital

This study aimed to develop a clinical Decision Support Model (DSM) which is software that provides physicians and other healthcare stakeholders with patient-specific assessments and recommendation in aiding clinical decision-making while discharging Breast cancer patient since the diagnostics and discharge problem is often overwhelming for a clinician to process at the point of care or in urgent situations. The model incorporates Breast cancer patient-specific data that are well-structured having been attained from a prestudy’s administered questionnaires and current evidence-based guidelines. Obtained dataset of the prestudy’s questionnaires is processed via data mining techniques to generate an optimal clinical decision tree classifier model which serves physicians in enhancing their decision-making process while discharging a breast cancer patient on basic cognitive processes involved in medical thinking hence new, better-formed, and superior outcomes. The model also improves the quality of assessments by constructing predictive discharging models from code attributes enabling timely detection of deterioration in the quality of health of a breast cancer patient upon discharge. The outcome of implementing this study is a decision support model that bridges the gap occasioned by less informed clinical Breast cancer discharge that is based merely on experts’ opinions which is insufficiently reinforced for better treatment outcomes. The reinforced discharge decision for better treatment outcomes is through timely deployment of the decision support model to work hand in hand with the expertise in deriving an integrative discharge decision and has been an agreed strategy to eliminate the foreseeable deteriorating quality of health for a discharged breast cancer patients and surging rates of mortality blamed on mistrusted discharge decisions. In this paper, we will discuss breast cancer clinical knowledge, data mining techniques, the classifying model accuracy, and the Python web-based decision support model that predicts avoidable re-hospitalization of a breast cancer patient through an informed clinical discharging support model.

A novel method for causal structure discovery from EHR data and its application to type-2 diabetes mellitus

Modern AI-based clinical decision support models owe their success in part to the very large number of predictors they use. Safe and robust decision support, especially for intervention planning, requires causal, not associative, relationships. Traditional methods of causal discovery, clinical trials and extracting biochemical pathways, are resource intensive and may not scale up to the number and complexity of relationships sufficient for precision treatment planning. Computational causal structure discovery (CSD) from electronic health records (EHR) data can represent a solution, however, current CSD methods fall short on EHR data. This paper presents a CSD method tailored to the EHR data. The application of the proposed methodology was demonstrated on type-2 diabetes mellitus. A large EHR dataset from Mayo Clinic was used as development cohort, and another large dataset from an independent health system, M Health Fairview, as external validation cohort. The proposed method achieved very high recall (.95) and substantially higher precision than the general-purpose methods (.84 versus .29, and .55). The causal relationships extracted from the development and external validation cohorts had a high (81%) overlap. Due to the adaptations to EHR data, the proposed method is more suitable for use in clinical decision support than the general-purpose methods.

Clinical decision support algorithm based on machine learning to assess the clinical response to anti–programmed death-1 therapy in patients with non–small-cell lung cancer

ObjectiveAnti–programmed death (PD)-1 therapy confers sustainable clinical benefits for patients with non–small-cell lung cancer (NSCLC), but only some patients respond to the treatment. Various clinical characteristics, including the PD-ligand 1 (PD-L1) level, are related to the anti–PD-1 response; however, none of these can independently serve as predictive biomarkers. Herein, we established a machine learning (ML)–based clinical decision support algorithm to predict the anti–PD-1 response by comprehensively combining the clinical information. Materials and methodsWe collected clinical data, including patient characteristics, mutations and laboratory findings, from the electronic medical records of 142 patients with NSCLC treated with anti–PD-1 therapy; these were analysed for the clinical outcome as the discovery set. Nineteen clinically meaningful features were used in supervised ML algorithms, including LightGBM, XGBoost, multilayer neural network, ridge regression and linear discriminant analysis, to predict anti–PD-1 responses. Based on each ML algorithm's prediction performance, the optimal ML was selected and validated in an independent validation set of PD-1 inhibitor–treated patients. ResultsSeveral factors, including PD-L1 expression, tumour burden and neutrophil-to-lymphocyte ratio, could independently predict the anti–PD-1 response in the discovery set. ML platforms based on the LightGBM algorithm using 19 clinical features showed more significant prediction performance (area under the curve [AUC] 0.788) than on individual clinical features and traditional multivariate logistic regression (AUC 0.759). ConclusionCollectively, our LightGBM algorithm offers a clinical decision support model to predict the anti–PD-1 response in patients with NSCLC.

Improving Recurrent Neural Network Responsiveness to Acute Clinical Events

Predictive models in acute care settings must immediately recognize precipitous changes in a patient's status when presented with data reflecting such changes. Recurrent neural networks (RNN) have become popular for clinical decision support models but exhibit a delayed response to acute events. New information must propagate through the RNN's cell state before the total impact is reflected in the model's predictions. Input data perseveration is a method to train more responsive RNN-based models. Input data is replicated k times during training and deployment. Each replication propagates through the cell state and output of the RNN, but only the output at the final replication is maintained and broadcast as the prediction for evaluation. De-identified Electronic Medical Records (EMR) of 12, 826 patients admitted to a tertiary care pediatric academic center between 01/2009-02/2019 were analyzed. A baseline Long Short-Term Memory (LSTM) model ( k=1), four LSTMs with increasing amounts of input data perseveration ( k=2 to k=5), and an LSTM with an attention mechanism were trained to predict ICU-mortality. Performance of models was compared using Area Under the Receiver Operating Characteristic Curve (AUROC) after increasing periods of observation from one to 12 hours. The average variation of the change in predicted mortality immediately following defined acute events measured responsiveness. The AUROC gains due to input perseveration were larger at the earlier times of prediction ( ≤ 6 hours), increasing at the first hour from 0.77 with no input data perseveration to 0.83 when k=5. An LSTM with k=5 was $2-3$ times more responsive to acute events than a baseline LSTM.

Coupling Wearable Devices and Decision Theory in the United States Emergency Department Triage Process: A Narrative Review.

This research was motivated by the nurses’ decision-making process in the current emergency department (ED) triage process in the United States. It explores how continuous vital signs monitoring can be integrated into the ED. The article presents four shortcomings on current ED triage systems and proposes a new conceptual clinical decision support model that exploits the benefits of combining wireless wearable devices with Multi-Attribute Utility Theory to address those shortcomings. A literature review was conducted using various engineering and medical research databases, analyzing current practices and identifying potential improvement opportunities. The results from the literature review show that advancements in wireless wearable devices provide opportunities to enhance current ED processes by monitoring patients while they wait after triage and, therefore, reduce the risk of an adverse event. A dynamic mathematical decision support model to prioritize patients is presented, creating a feedback loop in the ED. The coupling of wearable devices (to collect data) with decision theory (to synthesize and organize the information) can assist in reducing sources of uncertainty inherent to ED systems. The authors also address the feasibility of the proposed conceptual model.

Clinical information extraction for preterm birth risk prediction.

This paper contributes to the pursuit of leveraging unstructured medical notes to structured clinical decision making. In particular, we present a pipeline for clinical information extraction from medical notes related to preterm birth, and discuss the main challenges as well as its potential for clinical practice. A large collection of medical notes, created by staff during hospitalizations of patients who were at risk of delivering preterm, was gathered and analyzed. Based on an annotated collection of notes, we trained and evaluated information extraction components to discover clinical entities such as symptoms, events, anatomical sites and procedures, as well as attributes linked to these clinical entities. In a retrospective study, we show that these are highly informative for clinical decision support models that are trained to predict whether delivery is likely to occur within specific time windows, in combination with structured information from electronic health records.

Implementation and Continuous Monitoring of an Electronic Health Record Embedded Readmissions Clinical Decision Support Tool.

Unplanned hospital readmissions represent a significant health care value problem with high costs and poor quality of care. A significant percentage of readmissions could be prevented if clinical inpatient teams were better able to predict which patients were at higher risk for readmission. Many of the current clinical decision support models that predict readmissions are not configured to integrate closely with the electronic health record or alert providers in real-time prior to discharge about a patient’s risk for readmission. We report on the implementation and monitoring of the Epic electronic health record—“Unplanned readmission model version 1”—over 2 years from 1/1/2018–12/31/2019. For patients discharged during this time, the predictive capability to discern high risk discharges was reflected in an AUC/C-statistic at our three hospitals of 0.716–0.760 for all patients and 0.676–0.695 for general medicine patients. The model had a positive predictive value ranging from 0.217–0.248 for all patients. We also present our methods in monitoring the model over time for trend changes, as well as common readmissions reduction strategies triggered by the score.