Clinical Decision Aid Research Articles

Nomogram for predicting cervical lymph node metastasis of papillary thyroid carcinoma using deep learning-based super-resolution ultrasound image

ObjectivesTo investigate the feasibility and effectiveness of a deep learning (DL) super-resolution (SR) ultrasound image reconstruction model for predicting cervical lymph node status in patients with papillary thyroid carcinoma(PTC).MethodsIn this retrospective study, researchers recruited 544 patients with PTC and randomly assigned them to training and test sets. SR ultrasound images were acquired using SR technology to improve image resolution, and artificial features and DL features were extracted from the original (OR) and SR images, respectively, to construct a ML, DL model. The best model was selected and aggregated with clinical parameters to construct the nomogram. The performance of the model is evaluated by ROC curves, calibration curves and decision curves.ResultsIn distinguishing the presence or absence of metastatic lymph nodes, the predictive performance of the SR_ResNet 101 and SR_SVM models based on SR outperformed those based on OR. In the test set, SR_SVM AUC was 0.878 (95% CI 0.8203–0.9358), accuracy 0.854, while OR_SVM AUC was 0.822 (95% CI 0.7500–0.8937), accuracy 0.665. SR_ResNet 101 AUC was 0.799 (95% CI 0.7175–0.8806), accuracy 0.793, and OR_ResNet101 AUC was 0.751 (95% CI 0.6620–0.8401), accuracy 0.713. Subsequently, Nomogram_A and Nomogram_B were constructed by integrating the SR_SVM model and SR_ResNet 101 model, respectively, with clinical parameters, while Nomogram_C was constructed solely based on clinical indicators. In the test set, Nomogram_A demonstrated the best performance with an AUC of 0.930 (95% CI 0.8913–0.9682) and accuracy was 0.829. Nomogram_B AUC 0.868 (95% CI 0.8102–0.9261) and accuracy was 0.829, while Nomogram_C AUC 0.880 (95% CI 0.8257–0.9349) and accuracy was 0.787. The DeLong test revealed that the diagnostic performance of Nomogram_A based on SR_SVM was significantly higher than that of Nomogram_B, Nomogram_C, and the level of Radiologist (P < 0.05). The calibration curves and Hosmer–Lemeshow tests confirmed a high degree of fit, and the decision curve analysis demonstrated clinical value and potential patient benefit.ConclusionsThe predictive model constructed using SR reconstructed ultrasound images demonstrated superior performance in predicting preoperative cervical lymph node metastasis in PTC compared to OR images. The nomogram prediction model based on SR images has the potential to enhance the accuracy of predictive models and aid in clinical decision-making.

Investigation of the impact of the deep learning based CT fractional flow reserve on clinical decision-making and long-term prognosis in patients with obstructive coronary heart disease

Objective: To investigate the impact of the deep-learning-based CT fractional flow reserve (CT-FFR) on clinical decision-making and long-term prognosis in patients with obstructive coronary heart disease. Methods: In this single-center retrospective cohort study, consecutive patients with obstructive coronary heart disease (with at least one stenosis≥50%) on their first coronary computed tomography angiography (CCTA) in Beijing Anzhen Hospital from February 2017 to July 2018 were included. Baseline clinical and CT characteristics were collected. Deep-learning-based CT-FFR and Leiden CCTA risk score were calculated. All patients enrolled were followed up for at least 5 years. The study endpoint was major adverse cardiovascular events (MACE), defined as the composite of cardiac death, nonfatal myocardial infarction, unstable angina requiring hospitalization, and unplanned revascularization. Receiver operating characteristic (ROC) curves were drawn to define the optimal cut-off point of the Leiden score in predicting the 5-year MACE, and survival analysis and Cox regression were performed to explore the related factors of MACE. Results: A total of 622 patients, aged 61 (54, 66) years, with 407 (65.4%) males were included. Diagnostic coronary angiography was performed in 78 patients after their baseline CCTA, with 34 (43.6%) patients had CT-FFR>0.80. During a follow-up time of 2 181 (2 093, 2 355) days, 155 patients (24.9%) suffered from MACE. ROC derived optimal cut-off point of Leiden score for predicting MACE was 15.48. Survival analysis found that male patients, Leiden risk score>15 and CT-FFR≤0.80 had worse prognosis. Multivariate Cox regression analysis identified CT-FFR≤0.80 as an robust and independent predictor of MACE (HR=4.98, 95%CI 3.15-7.86, P<0.001). Conclusion: Deep-learning-based CT-FFR aids in clinical decision-making and the evaluation of long-term prognosis in patients with obstructive coronary heart disease.

Gene expression profiling in B-cell non-Hodgkin lymphomas

Abstract BACKGROUND: Gene expression profiling has become a fundamental tool in cancer diagnosis and management. B-cell non-Hodgkin lymphoma (B-NHL) is a group of malignant neoplasms originating from the lymphoid tissues, mainly the lymph nodes and the gene expression technique was used to unravel its complexity and aid in clinical decision-making. OBJECTIVES: The aims of this study were to find the significance of gene expression profiling focusing on colony-stimulating factor 1 receptor (CSF1R), myeloid differentiation factor 88 (MyD88), and tumor necrosis factor-α (TNF-α) as a promising approach in B-NHL diagnosis and their comparison with healthy controls. PATIENTS, MATERIALS AND METHODS: The current clinical prospective study was mediated from June 1, 2021, to December 30, 2022, of NHL patients in Kurdistan, Iraq. Seventy-three patients were recruited from Nanakali Hospital for Blood Diseases and Cancer, Erbil. The integration of gene expression biomarkers uses quantitative real-time polymerase chain reaction technique to diagnose B-NHL. Specifically, we focused on three key genes MyD88, TNF, and CSF1R whose expression profiles were analyzed in B-NHL patients and controls. We leveraged a dataset to explore gene expression patterns in B-NHL and applied classification algorithms to distinguish between B-NHL patients and controls. RESULTS: The initial results show the overall lower CSF1R expression in B-NHL as compared to the controls and a significant reduction in CSF1R expression in females (≤50 years and &gt;50 years). The result considers lower CSF1R expression in B-NHL males (≤50 years) and higher but not significant in males (&gt;50 years). CONCLUSIONS: These B-NHL-expressed genes may be considered potential diagnostic markers with their meaningful comparisons to control groups, and they could be proposed to guide the management of patients and facilitate their stratification into clinical trials.

Epigenome-Wide DNA Methylation in Unipolar Depression: Predictive Biomarker of Antidepressant Treatment Response?

Despite the well-documented efficacy of antidepressant agents for the treatment of major depressive disorder (MDD), initial treatment nonresponse rates are high. Recent years have seen an increase in research into predictive biomarkers toward improving diagnosis and individualized treatment. Among those, epigenetic mechanisms such as DNA methylation constitute promising candidate markers in predicting antidepressant treatment response in MDD. The present study sought to address epigenome-wide DNA methylation as a predictor of antidepressant treatment response in the largest sample to date of patients with MDD. Epigenome-wide DNA methylation was analyzed using the Infinium MethylationEPIC BeadChip in peripheral blood of n = 230 Caucasian patients with MDD receiving 6-week antidepressant treatment in a naturalistic in-patient setting as well as in a subsample of n = 107 patients primarily receiving continuous treatment with serotonin reuptake inhibitors or serotonin and norepinephrine reuptake inhibitors. Treatment response was assessed by means of the Hamilton Depression Scale. No genome-wide significant hits were observed. Suggestive (P < 1E-5) epigenome-wide evidence was discerned for altered DNA methylation at 6 CpG sites (LOC102724467, LOC100506023, RSPO2, SAG, IL16, PRKCI) to predict response to naturalistic antidepressant treatment. In patients treated with serotonin reuptake inhibitors or serotonin and norepinephrine reuptake inhibitors, differential DNA methylation at 11 CpGs, for example, mapping to the TIMP2, VDAC1, or SORL1 genes, was suggestively associated with treatment response. The present results provide preliminary evidence for altered DNA methylation patterns to be associated with antidepressant treatment response in MDD. Provided significant replication in independent and larger samples, the present findings might in the future aid in clinical decision-making toward more individualized and thus more efficacious treatments of MDD.

16 Deriviation of a Clinical Decision Aid to Rule Out Acute Aortic Syndrome in Patients Presenting to the Emergency Department With Chest Pain

16 Deriviation of a Clinical Decision Aid to Rule Out Acute Aortic Syndrome in Patients Presenting to the Emergency Department With Chest Pain

Optimizing the fairness of survival prediction models for racial/ethnic subgroups: A study on predicting post-operative survival in stage IA and IB non-small cell lung cancer.

380 Background: The recent surge of utilizing machine learning (ML) to develop prediction models for clinical decision-making aids is promising. However, these models can demonstrate racial bias due to inequities in real-world training data. In lung cancer, multiple models have been developed to predict prognosis, but none have been optimized to mitigate bias in performance among racial/ethnic subgroups. We developed a ML model to predict five-year survival in Stage 1A-1B non-small cell lung cancer (NSCLC), ensuring fairness on race. Methods: In the National Cancer Database, we identified patients with histopathologically confirmed stage 1A -1B NSCLC who underwent curative intent lobectomy from 2004 – 2017. We split the study cohort into a training and test sets (70%/30%). We trained and compared various ML models to predict 5-year overall survival. Patient demographic, clinical, and disease characteristics were used as input features for the models. To evaluate model fairness, we used the equalized odds ratio (eOR), which compares the true positive and false positive rates across groups; an eOR value of 1 represents equivalent rates across racial groups. We utilized 3 approaches to mitigate model bias and optimize for fairness of the best “naïve” model: grid search, threshold optimizer, and the exponentiated gradient methods. We evaluated model performance before and after bias mitigation using the area under the curve (AUC). Results: 124,298 patients fit our inclusion/exclusion criteria; 87% of patients were White, 8% were Black/African American, 3% Hispanic, and 2% Asian. Eighty percent of patients were diagnosed with stage 1A cancer; 20% had stage 1B cancer. The best naïve ML model, not optimized for fairness on race, had an eOR of 0.25 with an AUC of 0.66 (95% CI 0.65-0.66) overall. This model demonstrated an AUC of 0.65 (0.65-0.66) among white patients, 0.64 (0.62-0.66) among Black patients, 0.64 (0.60-0.68) among Asian patients, and 0.71 (0.68-0.74) among Hispanic patients. The threshold optimizer bias mitigation strategy improved fairness the most while maintaining similar overall performance of AUC 0.65 (0.64-0.66). With this strategy the eOR improved to 0.83 while AUC remained relatively stable across racial subgroups. Conclusions: We developed a ML model to predict 5-year survival in patients undergoing surgery for stage IA-IB NSCLC and employed model bias mitigation strategies that significantly improved model fairness, without diminishing overall performance. These strategies should be considered when developing prediction models for clinical decision making to avoid perpetuating disparities in care due to algorithm bias. Model performance metrics. Equalized Odds Ratio True Positive Rate False Positive Rate AUC Naive model 0.25 0.61 0.30 0.66 Threshold optimizer mitigated model 0.83 0.62 0.32 0.65

Comparative Effects of Neurodynamic Slider and Tensioner Mobilization Techniques on Sympathetic Nervous System Function: A Randomized Controlled Trial.

Objective: To investigate the effect of slider and tensioner neurodynamic techniques (NDTs) on the sympathetic nervous system (SNS) activity, aiming to identify which technique more effectively modulates autonomic responses in asymptomatic individuals. Materials and Methods: In this double-blind controlled trial, a total of 90 healthy participants were randomly allocated into three groups: slider, tensioner, and control. Skin conductance (SC) was continuously monitored throughout the entire 20 min experiment, while body temperature and blood pressure were measured pre- and post-intervention. Results: The SC levels significantly increased in both the slider and tensioner groups compared to the control group during the intervention and end rest period on the left leg (slider vs. control: p < 0.001, d = 1.20; tensioner vs. control: p < 0.001, d = 1.64) and on the right leg (slider vs. control: p < 0.001, d = 1.47; tensioner vs. control: p < 0.001, d = 0.73). There were no significant differences between the two NDTs on the left (p < 0.13, d = 0.89) and right legs (p < 1.00, d = 0.36). The body temperature of the slider group showed a significant increase compared to both the control group (p < 0.001, d = 0.95) and the tensioner group (p < 0.001, d = 1.48). There were no significant differences between the groups in systolic (p = 0.95) or diastolic blood pressure (p = 0.06). There were no side-specific effects on SNS activity between the left and right legs (p < 0.019) during all intervention phases. Conclusions: Significant sympathoexcitatory responses were elicited by both slider and tensioner NDTs in asymptomatic participants, demonstrating their efficacy in modulating the SNS. The differences between the two techniques were not statistically significant; however, the tensioner NDT showed a slightly more pronounced effect, suggesting that the tensioner NDT can be considered superior in terms of overall SNS effect. These findings indicate that both techniques may have the potential to enhance autonomic regulation in clinical practice; however, the tensioner NDT may be more effective. The consistent responses across participants highlight the systemic benefits of NDTs, providing a foundation for further research into their application in symptomatic populations. This study contributes to evidence-based practice by providing baseline data that support the development of theoretical frameworks and aid in clinical decision-making.

Testicular Germ Cell Tumours - features and prospects of the novel tumour marker microRNA-371a-3p (M371 test): a narrative review

Testicular germ cell tumours (GCTs) represent a paradigm for the usefulness of serum tumour markers in clinical management of diseases. However, the tumour markers currently in use, beta human chorionic gonadotropin (bHCG), alpha fetoprotein (AFP) and lactate dehydrogenase (LDH) are expressed in less than 50% of GCT cases. In 2011, microRNA-371a-3p (currently named M371) was suggested as a novel marker for the first time. Chemically, microRNAS represent small RNA molecules consisting of 18-24 base pairs. Physiologically, these microRNAs play a prominent role in the epigenetic control of protein biosynthesis. M371 can be measured in serum with PCR-techniques.There is high level evidence for a 90% sensitivity and >90% specificity for GCTs of the marker M371. That high diagnostic accuracy is true for both seminoma and nonseminoma but not for the histologic subgroup of teratoma. Testicular tumours of non-germ cell origin and malignant neoplasms of other organs do not express the marker. M371 involves a very short half-life of <24 hours.The test does likely involve the prospects of providing substantial aid in clinical decision-making with respect to instances where improvement of GCT management is still required. In particular, the following clinical scenarios will probably benefit from the employment of the M371 test: (1) diagnostic work-up of incidentally detected small testicular masses with decision-making in regard to testis sparing surgery or full orchiectomy (2) simplifying the follow-up of GCT patients with sparing of imaging procedures in a number of cases; (3) diagnostic evaluation of retroperitoneal lymphadenopathy upon clinical staging; (4) diagnostic evaluation of false-positive elevations of classical tumour markers (AFP, bHCG); (5) rapid appraisal of therapeutic success or failure by means of the very short half-life of M371; (6) diagnostic evaluation of postchemotherapy residual masses particularly those in seminoma patients.The discovery and development of the novel tumour marker M371 probably represents a milestone progress in the history of the clinical management of testicular GCTs.

Validation of classification system for isolated superior mesenteric artery dissections using image-based computational flow analysis

Background and ObjectiveThe isolated superior mesenteric artery dissection (ISMAD) is a rare but potentially fatal vascular disorder. Classifications for ISMAD were previously proposed based on morphometric features. However, the classification systems were not standardized and verified yet. This study conducted computational flow analysis to validate the latest classification system of ISMAD and aid clinical decision-making based on hemodynamic parameters. Methods62 patients with ISMAD were included and classified into different types according to false lumen structures (five types, Type I-V) and true lumen patency (two types, Type P and Type S) according to Qiu classification system. Computational fluid dynamics and three-dimensional structural analyses were conducted on the basis of computed tomography angiography datasets. Quantitative and qualitative functional analyses were performed via parameters of interest including volume flow of each minute, pressure drop, pressure gradient, the derivative parameters of wall shear stress such as time-averaged wall shear stress (TAWSS), oscillatory shear index (OSI), and the relative residence time (RRT). Statistical analyses were conducted among different ISMAD types. ResultsTAWSS, OSI and RRT showed significant difference among different types when classified using false lumen structures. In detail, Type IV showed significantly higher TAWSS than other types (p = 0.007). OSI was obviously higher in Type II (p = 0.015). Type IV also presented the lowest RRT (p = 0.005). The pressure drop, pressure gradient, OSI and RRT showed higher value in Type S than that in Type P, demonstrating a statistical significance with p values of 0.017, 0.041, 0.001 and 0.012, respectively. While Type P had larger volume flow than Type S (p = 0.041). ConclusionsThe notable differences in hemodynamic features among different types demonstrated the feasibility of Qiu classification system. The evaluation based on hemodynamic simulation might also provide insights into risk identification and guide therapeutic decisions for ISMAD.

A novel study on the quality of life index in canine chronic kidney disease treated with incremental intermittent hemodialysis.

In veterinary medicine, health-related quality-of-life index (QOLi) measurements are becoming increasingly important because they are a multifaceted concept that represents not only patients' physical well-being but also clients' emotional health. This study assessed QOLi in dogs receiving incremental intermittent hemodialysis (i-IHD) with high- and low-flux dialyzers. Thirty dogs diagnosed with chronic kidney disease (CKD) Stage IV were randomly divided into two groups of 15 dogs each. A high-flux dialyzer was used in Group I, whereas a low-flux dialyzer was used in Group II. i-IHD was performed on days 0, 2, 4, 19, and 34, whereas QOLi evaluation was performed on days 0, 15, 30, and 45. Both groups exhibited considerable decreases in post-dialysis creatinine, blood urea nitrogen, and phosphorus levels, while Group I experienced notable reductions in post-dialysis triglyceride and cholesterol levels. Dialysis adequacy did not show any significant difference between the clearance rates of high- and low-flux dialyzers. The QOLi assessment showed better post-dialysis scores in all categories except for water balance in Group I, while Group II demonstrated a worsening trend in scores for mental status, appetite, mobility, general health, and pain. In the first three sessions of i-IHD, dogs with CKD should be treated every other day, and the schedule can be extended by 15 days after that. A high-flux membrane, which effectively decreases triglyceride and cholesterol levels more than a low-flux membrane, warrants consideration for dogs with cardiovascular complications undergoing dialysis. The dialysis-related QOLi aids in clinical decision-making and encourages client engagement.

An Advanced Cardiac Life Support Application Improves Performance during Simulated Cardiac Arrest.

Variability in cardiopulmonary arrest training and management leads to inconsistent outcomes during in-hospital cardiac arrest. Existing clinical decision aids, such as American Heart Association (AHA) advanced cardiovascular life support (ACLS) pocket cards and third-party mobile apps, often lack comprehensive management guidance. We developed a novel, guided ACLS mobile app and evaluated user performance during simulated cardiac arrest according to the 2020 AHA ACLS guidelines via randomized controlled trial. Forty-six resident physicians were randomized to lead a simulated code team using the AHA pockets cards (N = 22) or the guided app (N = 24). The primary outcome was successful return of spontaneous circulation (ROSC). Secondary outcomes included code leader stress and confidence, AHA ACLS guideline adherence, and errors. A focus group of 22 residents provided feedback. Statistical analysis included two-sided t-tests and Fisher's exact tests. App users showed significantly higher ROSC rate (50 vs. 18%; p = 0.024), correct thrombolytic administration (54 vs. 23%; p = 0.029), backboard use (96 vs. 27%; p < 0.001), end-tidal CO2 monitoring (58 vs. 27%; p = 0.033), and confidence compared with baseline (1.0 vs 0.3; p = 0.005) compared with controls. A focus group of 22 residents indicated unanimous willingness to use the app, with 82% preferring it over AHA pocket cards. Our guided ACLS app shows potential to improve user confidence and adherence to the AHA ACLS guidelines and may help to standardize in-hospital cardiac arrest management. Further validation studies are essential to confirm its efficacy in clinical practice.

Exploring communication preferences and risk thresholds of clinicians and parents of febrile infants under 90 days presenting to the emergency department: a qualitative study

BackgroundFebrile infants under 3 months of age are at higher risk of invasive bacterial illness (IBI) when compared with older children. Increasingly sequential assessment based on age, clinical appearance and...

Development and validation of an interpretable machine learning for mortality prediction in patients with sepsis.

Sepsis is a leading cause of death. However, there is a lack of useful model to predict outcome in sepsis. Herein, the aim of this study was to develop an explainable machine learning (ML) model for predicting 28-day mortality in patients with sepsis based on Sepsis 3.0 criteria. We obtained the data from the Medical Information Mart for Intensive Care (MIMIC)-III database (version 1.4). The overall data was randomly assigned to the training and testing sets at a ratio of 3:1. Following the application of LASSO regression analysis to identify the modeling variables, we proceeded to develop models using Extreme Gradient Boost (XGBoost), Logistic Regression (LR), Support Vector Machine (SVM), and Random Forest (RF) techniques with 5-fold cross-validation. The optimal model was selected based on its area under the curve (AUC). Finally, the Shapley additive explanations (SHAP) method was used to interpret the optimal model. A total of 5,834 septic adults were enrolled, the median age was 66 years (IQR, 54-78 years) and 2,342 (40.1%) were women. After feature selection, 14 variables were included for developing model in the training set. The XGBoost model (AUC: 0.806) showed superior performance with AUC, compared with RF (AUC: 0.794), LR (AUC: 0.782) and SVM model (AUC: 0.687). SHAP summary analysis for XGBoost model showed that urine output on day 1, age, blood urea nitrogen and body mass index were the top four contributors. SHAP dependence analysis demonstrated insightful nonlinear interactive associations between factors and outcome. SHAP force analysis provided three samples for model prediction. In conclusion, our study successfully demonstrated the efficacy of ML models in predicting 28-day mortality in sepsis patients, while highlighting the potential of the SHAP method to enhance model transparency and aid in clinical decision-making.

Amplifying Chinese physicians' emphasis on patients' psychological states beyond urologic diagnoses with ChatGPT-A multi-center cross-sectional study.

Artificial intelligence (AI) technologies, particularly large language models (LLMs), have been widely employed by the medical community. In addressing the intricacies of urology, ChatGPT offers a novel possibility to aid in clinical decision-making. This study aimed to investigate the decision-making ability of LLMs in solving complex urology-related problems and assess its effectiveness in providing psychological support to patients with urological disorders. This study evaluated the clinical and psychological support capabilities of ChatGPT 3.5 and 4.0 in the field of urology. A total of 69 clinical and 30 psychological questions were posed to the AI models, and their responses were evaluated by both urologists and psychologists. As a control, clinicians from Chinese medical institutions provided responses under closed-book conditions. Statistical analyses were conducted separately for each subgroup. In multiple-choice tests covering diverse urological topics, ChatGPT 4.0, performed comparably to the physician group, with no significant overall score difference. Subgroup analyses revealed variable performance, based on disease type and physician experience, with ChatGPT 4.0 generally outperforming ChatGPT 3.5 and exhibiting competitive results against physicians. When assessing the psychological support capabilities of AI, it is evident that ChatGPT4.0 outperforms ChatGPT3.5 across all urology-related psychological problems. The performance of LLMs in dealing with standardized clinical problems and providing psychological support has certain advantages over clinicians. AI stands out as a promising tool for potential clinical aid.

Impact of quantitative CT texture analysis on the outcome of CT-guided bone biopsy

Texture analysis can provide new imaging-based biomarkers. Texture analysis derived from computed tomography (CT) might be able to better characterize patients undergoing CT-guided percutaneous bone biopsy. The present study evaluated this and correlated texture features with bioptic outcome in patients undergoing CT-guided bone biopsy. Overall, 123 patients (89 female patients, 72.4 %) were included into the present study. All patients underwent CT-guided percutaneous bone biopsy with an 11 Gauge coaxial needle. Clinical parameters and quantitative imaging features were investigated. Random forest classifier was used to predict a positive biopsy result. Overall, 69 patients had osteolytic metastasis (56.1 %) and 54 had osteoblastic metastasis (43.9 %). The overall positive biopsy rate was 72 %. The developed radiomics model demonstrated a prediction accuracy of a positive biopsy result with an AUC of 0.75 [95 %CI 0.65 – 0.85]. In a subgroup of breast cancer patients, the model achieved an AUC of 0.85 [95 %CI 0.73 – 0.96]. In the subgroup of non-breast cancer patients, the signature achieved an AUC of 0.80 [95 %CI 0.60 – 0.99]. Quantitative CT imaging findings comprised of conventional and texture features can aid to predict the bioptic result of CT-guided bone biopsies. The developed radiomics signature aids in clinical decision-making, and could identify patients at risk for a negative biopsy.

Clinical decision aids and computed tomography coronary angiography in patients with suspected acute coronary syndrome

BackgroundThe HEART score, the T-MACS model and the GRACE score support early decision-making for acute chest pain, which could be complemented by CT coronary angiography (CTCA). However, their performance has...

Revolutionizing Neurology: The Role of Artificial Intelligence in Advancing Diagnosis and Treatment.

Artificial intelligence (AI) has emerged as a powerful tool in the field of neurology, significantly impacting the diagnosis and treatment of neurological disorders. Recent technological breakthroughs have given us access to a plethora of information relevant to many aspects of neurology. NeuroscienceandAIshare a long history of collaboration.Along with great potential, we encounter obstacles relating to data quality, ethics, and inherent difficulty in applying data science in healthcare. Neurological disorders pose intricate challenges due to their complex manifestations and variability.Automating image interpretation tasks, AI algorithms accurately identify brain structures and detect abnormalities. This accelerates diagnosis and reduces the workload on medical professionals. Treatment optimization benefits from AI simulations that model different scenarios and predict outcomes. These AI systems can currently perform many of the sophisticated perceptual and cognitive capacities of biological systems, such as object identification and decision making. Furthermore, AI is rapidly being used as a tool in neuroscience research, altering our understanding of brain functioning. It has the ability to revolutionize healthcare as we know it into a system in which humans and robots collaborate to deliver better care for our patients.Image analysis activities such as recognizing particular brain regions, calculating changes in brain volume over time, and detecting abnormalities in brain scans can be automated by AI systems. This lessens the strain on radiologists and neurologists while improving diagnostic accuracy and efficiency.It is now obvious that cutting-edge artificial intelligence models combined with high-quality clinical data will lead to enhanced prognostic and diagnostic models in neurological illness, permitting expert-level clinical decision aids across healthcare settings.In conclusion, AI's integration into neurology has revolutionized diagnosis, treatment, and research. As AI technologies advance, they promise to unravel the complexities of neurological disorders further, leading to improved patient care and quality of life. The symbiosis of AI and neurology offers a glimpse into a future where innovation and compassion converge to reshape neurological healthcare. This abstract provides a concise overview of the role of AI in neurology and its transformative potential.

ENHANCING ELECTIVE SURGERY MANAGEMENT WITH OPENPREDICTOR: VALIDATION OF A CLINICAL DECISION AID FOR OPTIMIZING ELECTIVE PATHWAYS

OpenPredictor, a machine learning-enabled clinical decision aid, has been developed to manage backlogs in elective surgeries. It aims to optimise the use of high volume, low complexity surgical pathways by accurately stratifying patient risk, thereby facilitating the allocation of patients to the most suitable surgical sites. The tool augments elective surgical pathways by providing automated secondary opinions for perioperative risk assessments, enhancing decision-making. Its primary application is in elective sites utilising lighter pre-assessment methods, identifying patients with minimal complication risks and those high-risk individuals who may benefit from early pre-assessment.The Phase 1 clinical evaluation of OpenPredictor entailed a prospective analysis of 156 patient records from elective hip and knee joint replacement surgeries. Using a polynomial logistic regression model, patients were categorised into high, moderate, and low-risk groups. This categorisation incorporated data from various sources, including patient demographics, co-morbidities, blood tests, and overall health status.In identifying patients at risk of postoperative complications, OpenPredictor demonstrated parity with consultant-led preoperative assessments. It accurately flagged 70% of patients who later experienced complications as moderate or high risk. The tool's efficiency in risk prediction was evidenced by its balanced accuracy (75.6%), sensitivity (70% with a 95% confidence interval of 62.05% to 76.91%), and a high negative predictive value (96.7%).OpenPredictor presents a scalable and consistent solution for managing elective surgery pathways, comparable in performance to secondary consultant opinions. Its integration into pre-assessment workflows assists in efficient patient categorisation, reduces late surgery cancellations, and optimises resource allocation. The Phase 1 evaluation of OpenPredictor underscores its potential for broader clinical application and highlights the need for ongoing data refinement and system integration to enhance its performance.

Electronic Real-Time Monitoring Reveals Limited Adherence to Long-Term Opioid Prescriptions in Pain Patients.

Pain management physicians are increasingly focused on limiting prescription opioid abuse, yet existing tools for monitoring adherence have limited accuracy. Medication event monitoring system (MEMS) is an emerging technology for tracking medication usage in real-time but has not been tested in chronic pain patients on long-term opioid regimens. We conducted a pilot clinical trial to investigate the utility of MEMS for monitoring opioid adherence and compared to traditional methods including self-report diaries, urine drug screen (UDS), and physicians' opinions. Opioid-maintained chronic pain patients were recruited from a pain management clinic. Participants (n=28) were randomly assigned to either receive MEMS bottles containing their opioid medication for a 90-day period or to continue using standard medication bottles. MEMS bottles were configured to record and timestamp all bottle openings and the number of pills that were removed from the bottle (via measurement of weight change). Participants who received MEMS demonstrated highly heterogenous dosing patterns, with a substantial number of patients rapidly removing excessive amounts of medication and/or "stockpiling" medication. By comparison, physicians rated all participants as either "totally compliant" or "mostly compliant". UDS results did not reveal any illicit drug use, but 25% of participants (n=7) tested negative for their prescribed opioid metabolite. MEMS data did not correlate with physician-rated adherence (P=0.24) and UDS results (P=0.77). MEMS data consistently revealed greater non-adherence than self-report data (P<0.001). These results highlight the limits in our understanding of naturalistic patterns of daily opioid use in chronic pain patients as well as support the use of MEMS for detecting potential misuse as compared to routine adherence monitoring methods. Future research directions include the need to determine how MEMS could be used to improve patient outcomes, minimize harm, and aid in clinical decision-making. This study was preregistered on ClinicalTrials.gov (NCT03752411).

Looking Beyond Mortality Prediction: Primary Care Physician Views of Patients' Palliative Care Needs Predicted by a Machine Learning Tool.

To assess primary care physicians' (PCPs) perception of the need for serious illness conversations (SIC) or other palliative care interventions in patients flagged by a machine learning tool for high 1-year mortality risk. We surveyed PCPs from four Brigham and Women's Hospital primary care practice sites. Multiple mortality prediction algorithms were ensembled to assess adult patients of these PCPs who were either enrolled in the hospital's integrated care management program or had one of several chronic conditions. The patients were classified as high or low risk of 1-year mortality. A blinded survey had PCPs evaluate these patients for palliative care needs. We measured PCP and machine learning tool agreement regarding patients' need for an SIC/elevated risk of mortality. Of 66 PCPs, 20 (30.3%) participated in the survey. Out of 312 patients evaluated, 60.6% were female, with a mean (standard deviation [SD]) age of 69.3 (17.5) years, and a mean (SD) Charlson Comorbidity Index of 2.80 (2.89). The machine learning tool identified 162 (51.9%) patients as high risk. Excluding deceased or unfamiliar patients, PCPs felt that an SIC was appropriate for 179 patients; the machine learning tool flagged 123 of these patients as high risk (68.7% concordance). For 105 patients whom PCPs deemed SIC unnecessary, the tool classified 83 as low risk (79.1% concordance). There was substantial agreement between PCPs and the tool (Gwet's agreement coefficient of 0.640). A machine learning mortality prediction tool offers promise as a clinical decision aid, helping clinicians pinpoint patients needing palliative care interventions.