Highest Area Under The ROC Curve Research Articles

Comparative analysis of SARC-F-EBM, Ishii test, and six other screening tools for sarcopenia in Chinese community-dwelling older adults: a cross-sectional diagnostic study

To assess the relative performance of simple screening methods for sarcopenia in Chinese community-dwelling older adults. Data of older adults aged ≥ 60 were collected through a cross-sectional investigation. Sarcopenia was defined according to the Asian Working Group for Sarcopenia 2019 criteria. The accuracy of screening methods was evaluated using sensitivity, specificity, receiver operating characteristic (ROC) curves and area under the ROC curves (AUC). The AUC value greater than 0.8 represented the good screening ability. A total of 918 older adults (44.3% men, mean age 70.4 ± 6.5 years) were included in this study. The overall prevalence rates of possible sarcopenia, confirmed sarcopenia, and severe sarcopenia were 59.5%, 12.8%, and 5.9%, respectively. In men, the SARC-F-EBM and Ishii tests indicated good screening capabilities for confirmed sarcopenia, with an AUC of 0.81 (95% CI: 0.77–0.85) and 0.80 (95% CI: 0.76–0.84), respectively. In women, the highest AUC was also achieved using the SARC-F-EBM at 0.79 (95% CI: 0.75–0.82), followed by the Ishii test at 0.77 (95% CI: 0.74–0.81), showing the moderate efficacy. A ranking diagram showed that SARC-F-EBM was most likely to be considered the best method for diagnosing sarcopenia in terms of AUC and sensitivity, regardless of sex. We recommend the SARC-F-EBM for sarcopenia screening in community-dwelling Chinese older adults when respondents are able to answer the questionnaire accurately; otherwise, the Ishii test consisting entirely of objective measures could be used.

Diagnostic Performance for Detection of Glaucomatous Structural Damage Using Pixelwise Analysis of Retinal Thickness Measurements.

To compare the diagnostic accuracy of thickness measurements of individual and combined macular retinal layers to discriminate 188 glaucomatous and 148 glaucoma suspect eyes from 362 healthy control (HC) eyes on a pixel-by-pixel basis. For this retrospective study, we manually corrected the segmentations of posterior pole optical coherence tomography (OCT) scans to determine the thickness of the nerve fiber layer (NFL), ganglion cell layer (GCL), inner plexiform layer (IPL), the ganglion cell complex (GCC), and the total neural retina (TR). For each eye, the total number of pixels with thickness values less than the fifth percentile of the HC distribution was used to create a receiver operating characteristic (ROC) curve for each layer and for layer combinations. Using total abnormal pixel count criteria to discriminate glaucoma from HC eyes, the individual layers with the highest area under the ROC curve (AUC) were the NFL and GCL; IPL performance was significantly lower (P < 0.05). GCC had a significant higher AUC (94.3%) than individual the AUC of the NFL (92.3%) (P = 0.0231) but not higher than AUC of the GCL (93.4%) (P = 0.3487). The highest AUC (95.4%) and sensitivity (85.1%) at 95% specificity was found for the Boolean combination of NFL or GCL. The highest AUC is not significantly higher (P = 0.0882) than the AUC of the GCC but the highest sensitivity is significantly higher than the sensitivity of the GCC. This pattern was similar for discriminating between suspect and HC eyes (P = 0.0356). Using pixel-based methods, the diagnostic accuracy of NFL and GCL exceeded that of IPL and TR. GCC had equivalent performance as NFL and GCL. The specific spatial locations within the posterior pole that exhibit best performance vary depending on which layer is being assessed. Recognizing this dependency highlights the importance of considering multiple layers independently, as they offer complementary information for effective and comprehensive diagnosis.

Multicenter Development and Prospective Validation of eCARTv5: A Gradient Boosted Machine Learning Early Warning Score.

Early detection of clinical deterioration using machine learning early warning scores may improve outcomes. However, most implemented scores were developed using logistic regression, only underwent retrospective validation, and were not tested in important subgroups. Our objective was to develop and prospectively validate a gradient boosted machine model (eCARTv5) for identifying clinical deterioration on the wards. Multicenter retrospective and prospective observational study. Inpatient admissions to the medical-surgical wards at seven hospitals in three health systems for model development (2006-2022) and at 21 hospitals from three health systems for retrospective (2009-2023) and prospective (2023-2024) external validation. All adult patients hospitalized at each participating health system during the study years. None MEASUREMENTS AND MAIN RESULTS: Predictor variables (demographics, vital signs, documentation, and laboratory values) were used in a gradient boosted trees algorithm to predict intensive care unit transfer or death in the next 24 hours. The developed model (eCART) was compared to the Modified Early Warning Score (MEWS) and the National Early Warning Score (NEWS) using the area under the receiver operating characteristic curve (AUROC). The development cohort included 901,491 admissions, the retrospective validation cohort included 1,769,461 admissions, and the prospective validation cohort included 205,946 admissions. In retrospective validation, eCART had the highest AUROC (0.835; 95%CI 0.834, 0.835), followed by NEWS (0.766 (95%CI 0.766, 0.767)), and MEWS (0.704 (95%CI 0.703, 0.704)). eCART's performance remained high (AUROC ≥0.80) across a range of patient demographics, clinical conditions, and during prospective validation. We developed eCART, which performed better than the NEWS and MEWS retrospectively, prospectively, and across a range of subgroups. These results served as the foundation for Food and Drug Administration clearance for its use in identifying deterioration in hospitalized ward patients.

Development and validation of machine learning models for predicting cancer-related fatigue in lymphoma survivors

BackgroundNew cases of lymphoma are rising, and the symptom burden, like cancer-related fatigue (CRF), severely impacts the quality of life of lymphoma survivors. However, clinical diagnosis and treatment of CRF are inadequate and require enhancement. ObjectiveThe main objective of this study is to construct machine learning-based CRF prediction models for lymphoma survivors to help healthcare professionals accurately identify the CRF population and better personalize treatment and care for patients. MethodsA cross-sectional study in China recruited lymphoma patients from June 2023 to March 2024, dividing them into two datasets for model construction and external validation. Six machine learning algorithms were used in this study: Logistic Regression (LR), Random Forest, Single Hidden Layer Neural Network, Support Vector Machine, eXtreme Gradient Boosting, and Light Gradient Boosting Machine (LightGBM). Performance metrics like the area under the receiver operating characteristic (AUROC) and calibration curves were compared. The clinical applicability was assessed by decision curve, and Shapley additive explanations was employed to explain variable significance. ResultsCRF incidence was 40.7 % (dataset I) and 44.8 % (dataset II). LightGBM showed strong performance in training and internal validation. LR excelled in external validation with the highest AUROC and best calibration. Pain, total protein, physical function, and sleep disturbance were important predictors of CRF. ConclusionThe study presents a machine learning-based CRF prediction model for lymphoma patients, offering dynamic, data-driven assessments that could enhance the development of automated CRF screening tools for personalized management in clinical practice.

An optimized isolation forest based intrusion detection system for heterogeneous and streaming data in the industrial Internet of Things (IIoT) networks

While conventional Intrusion Detection Systems (IDS) are essential for defending against intruders in the Industrial Internet of Things (IIoT), handling data from heterogeneous and streaming data sources should receive more attention. This work introduces a novel Optimized IForest-based Intrusion Detection System (OIFIDS) which is designed to handle both heterogeneous and streaming data efficiently. The suggested approach employs a collection of optimized binary trees, each of which is trained on a distinctive subset of data, and in which the location of empty leaves determines the anomaly score assigned to a certain data point. Optimizing isolation Forest (iForest) utilizing a modified version of the Harris Hawks Optimization algorithm, which exploits both Exploration factor and Random walk strategies (ERHHO) decreases the dataset's dimension, decreases its learning time, and enhances the detection precision, accuracy, F1-score, FPR, and recall. To demonstrate how effective the proposed approach is, it is evaluated using three datasets: CICIDS-2018, NSL-KDD, and UNSW-NB15. The experimental results prove the ability of the suggested approach in handling both heterogeneous and streaming data efficiently and delivering results that were comparable to the cutting-edge baseline techniques. Moreover, it performs effectively when there are no anomalies in the training sample and when dealing with challenging scenarios with several irrelevant features and high dimensions. Based on the comparison with various state-of-the-art IDSs, the suggested approach is able to detect intrusion with greater accuracies of 95.6%, 94.8%, and 99% than the other approaches on the NSL-KDD, UNSW-NB15, and CICIDS-2018 datasets, respectively. Experiments on heterogenous data reveal that Area Under the ROC Curve (AUC) of OIFIDS beats the baseline approach for UNSW-NB15 dataset and is higher than the second-best method by 8% and 2.4% for NSL-KDD and CICIDS-2018 respectively. Evaluating the proposed system on streaming data illustrates that it can address the concept drift problem well with high AUC value of 0.948, 0.97, and 0.922 on the NSL-KDD, CICIDS-2018, and UNSW-NB15 datasets, respectively.

Using machine learning to predict bacteremia in urgent care patients on the basis of triage data and laboratory results

BackgroundDespite advancements in antimicrobial therapies, bacteremia remains a life-threatening condition. Appropriate antimicrobials must be promptly administered to ensure patient survival. However, diagnosing bacteremia based on blood cultures is time-consuming and not something emergency department (ED) personnel are routinely trained to do. MethodsThis retrospective cohort study developed several machine learning (ML) models to predict bacteremia in adults initially presenting with fever or hypothermia, comprising logistic regression, random forest, extreme gradient boosting, support vector machine, k-nearest neighbor, multilayer perceptron, and ensemble models. Random oversampling and synthetic minority oversampling techniques were adopted to balance the dataset. The variables included demographic characteristics, comorbidities, immunocompromised status, clinical characteristics, subjective symptoms reported during ED triage, and laboratory data. The study outcome was an episode of bacteremia. ResultsOf the 5063 patients with initial fever or hypothermia from whom blood cultures were obtained, 128 (2.5 %) were diagnosed with bacteremia. We combined 36 selected variables and 10 symptoms subjectively reported by patients into features for analysis in our models. The ensemble model outperformed other models, with an area under the receiver operating characteristic curve (AUROC) of 0.930 and an F1-score of 0.735. The AUROC of all models was higher than 0.80. ConclusionThe ML models developed effectively predicted bacteremia among febrile or hypothermic patients in the ED, with all models demonstrating high AUROC values and rapid processing times. The findings suggest that ED clinicians can effectively utilize ML techniques to develop predictive models for addressing clinical challenges.

PhosBERT: A self-supervised learning model for identifying phosphorylation sites in SARS-CoV-2-infected human cells

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is a single-stranded RNA virus, which mainly causes respiratory and enteric diseases and is responsible for the outbreak of coronavirus disease 19 (COVID-19). Numerous studies have demonstrated that SARS-CoV-2 infection will lead to a significant dysregulation of protein post-translational modification profile in human cells. The accurate recognition of phosphorylation sites in host cells will contribute to a deep understanding of the pathogenic mechanisms of SARS-CoV-2 and also help to screen drugs and compounds with antiviral potential. Therefore, there is a need to develop cost-effective and high-precision computational strategies for specifically identifying SARS-CoV-2-infected phosphorylation sites. In this work, we first implemented a custom neural network model (named PhosBERT) on the basis of a pre-trained protein language model of ProtBert, which was a self-supervised learning approach developed on the Bidirectional Encoder Representation from Transformers (BERT) architecture. PhosBERT was then trained and validated on serine (S) and threonine (T) phosphorylation dataset and tyrosine (Y) phosphorylation dataset with 5-fold cross-validation, respectively. Independent validation results showed that PhosBERT could identify S/T phosphorylation sites with high accuracy and AUC (area under the receiver operating characteristic) value of 81.9% and 0.896. The prediction accuracy and AUC value of Y phosphorylation sites reached up to 87.1% and 0.902. It indicated that the proposed model was of good prediction ability and stability and would provide a new approach for studying SARS-CoV-2 phosphorylation sites.

Management of Mild Postpartum Anemia: Is Iron Administration Effective?

Background This study aimed to investigate the efficacy of iron therapy in the treatment of mild postpartum anemia. Methods We conducted a case-control study involving women who underwent cesarean section at our hospital between 2015 and 2020. Following propensity score matching, participants were categorized into two groups based on whether or not they received iron therapy. These patients were evaluated for mean hemoglobin (Hb) levels on the seventh postoperative day (POD 7), the percentage of subjects achieving Hb greater than 10 g/dL on POD 7, and the incidence of adverse events. The efficacy of iron administration was evaluated using a superiority test, and receiver operating characteristic analyses were employed to generate area under the receiver operating characteristic curves (AUROC). Results The mean Hb level on POD 7 was 10.12 g/dL in the iron group and 9.89 g/dL in the iron-free group (P = 0.206). The superiority test revealed that the percentage of subjects achieving Hb levels greater than 10 g/dL on POD 7 was 56.1% in the iron group and 48.8% in the iron-free group (P = 0.880), indicating that the iron group did not demonstrate superiority over the iron-free group. The incidence of adverse events was significantly higher in the iron group (P = 0.027). The highest AUROC was observed with preoperative mean corpuscular Hb, measuring 0.632 (95% CI: 0.509-0.755), with a cutoff point of 28.5 pg. Conclusion Consideration should be given to the uniform administration of iron for the management of mild postpartum anemia.

Continuous Glucose Monitoring Prediction of Gestational Diabetes Mellitus and Perinatal Complications.

Objective: To assess the performance of continuous glucose monitoring (CGM)-measured glycemic metrics in predicting development of gestational diabetes mellitus (GDM) and select perinatal complications. Research Methods: In a prospective observational study, CGM data were collected from 760 pregnant females throughout gestation after study enrollment. GDM was diagnosed using the oral glucose tolerance test (OGTT) at 24-34 weeks of gestation. Predictive models were built using logistic and elastic net regression. Predictive performance was assessed by the area under the receiver-operating characteristic (AUROC) curve. Results: The AUROCs of using second trimester percent time >140 mg/dL (TA140) and week 13-14 TA140 in predicting GDM were 0.81 and 0.74, respectively. The AUROCs for predicting large-for-gestational-age (LGA) births and hypertensive disorders of pregnancy (HDP) using second trimester TA140 were both 0.58. When matching the specificity of OGTT, a model using TA140 in weeks 13-14 achieved similar sensitivity to OGTT in predicting HDP (13% vs. 10%, respectively) and LGA (6% for both methods). Elastic net also demonstrated similar AUROC and diagnostic performance with no meaningful improvement by using multiple predictors. Conclusion: CGM-measured hyperglycemic metrics such as TA140 predicted GDM with high AUROCs as early as 13-14 weeks of gestation. These metrics were also similar statistically to the OGTT at 24-34 weeks in predicting perinatal complications, although sensitivity was low for both. CGM could potentially be used as an early screening tool for elevated hyperglycemia during gestation, which could be used in addition to or instead of the OGTT.

O-220 Ovarify: a deep learning model for automatic calculation of ovarian morphological features in pelvic ultrasonography

Abstract Study question Can a deep learning model accurately characterize ovarian features on pelvic ultrasounds? Summary answer We developed a deep learning model, Ovarify, that can accurately classify and segment ovarian features on pelvic ultrasounds. What is known already Ovarian morphology, as detected using pelvic ultrasonography, is used widely in the assessment of gynecological health. Morphological metrics such as ovarian volume and antral follicle count have known diagnostic and prognostic value for a variety of gynecological outcomes, ranging from risk for ovulatory disorders to likelihood of response to assisted reproductive technologies. However, a lack of standardization in the interpretation of ovarian features, and the time-intensive nature of validated image analysis techniques (such as 2D grid analysis), limits their utility in clinical practice. Deep learning provides an opportunity to overcome these limitations. Study design, size, duration The training dataset for the model comprised 426 3D pelvic ultrasounds obtained using the GE Voluson ultrasound platform, yielding 56,538 individual frames with potential ovarian content. In addition, a similarly collected validation dataset consisted of 95 ultrasounds, yielding 11,304 frames. The held-out test set comprised 107 ultrasounds with 14,767 frames. All frames were manually labeled with the associated 2D ovarian contour, and 163 of the ultrasounds had an associated ovarian volume label. Participants/materials, setting, methods Ovarify consists of two stages: 1) a lightweight classifier that labels ultrasound frames as either positive or negative (having or not having ovarian content) and 2) a subsequent U-Net-inspired segmentation module with an Xception backbone and attention-based decoder that segments the ovary and antral follicles in all frames previously classified as positive. Classification performance was evaluated using F1-score and area under the receiver operating characteristic curve (AUROC). Segmentation performance was evaluated using the Jaccard index. Main results and the role of chance For the classification stage, Ovarify achieved an AUROC of 0.980±0.002 and 0.976±0.004 on the training and validation sets, respectively. On the held-out test set, the model achieved an AUROC of 0.971±0.005. In addition, the model achieved an F1-score of 0.825±0.006, 0.810±0.013, and 0.784±0.014 on the training, validation, and held-out testing sets, respectively. The high AUROC and F1-scores across both seen and unseen datasets support a high-performing classifier. In the segmentation stage, Ovarify reached a median Jaccard index of 0.815±0.004 on the unseen test set when tasked with segmenting the ovary, and a median Jaccard index of 0.657±0.018 on the unseen test set when segmenting antral follicles. High agreement between the ground truth ovary and follicle labels and the predicted segmented regions increases confidence in the downstream morphological analysis. For example, the Pearson correlation coefficient between the Ovarify-calculated ovarian volume and manually-validated ovarian volume was 0.813, indicating that an automatic ovarian volume calculation highly aligns with the ground truth. Ovarify was able to provide predictions in under one minute for the entirety of the held-out test set. Limitations, reasons for caution Ovarify was trained on images generated by the GE Voluson platform and has not been evaluated for generalizability across ultrasound images generated using machines whose output does not match the DICOM format standard. Wider implications of the findings The model developed is relevant for automatically calculating ovarian morphological features from pelvic ultrasounds. This approach has implications in removing user-specific bias and error when calculating biomarkers associated with female reproductive health, including but not limited to ovulatory disorders such as polycystic ovary syndrome (PCOS). Trial registration number not applicable

Development and Validation of a Risk Mortality Prediction Model for Patients with Pulmonary Tuberculosis Complicated by Severe Community-Acquired Pneumonia in the Intensive Care Unit.

The mortality rate from pulmonary tuberculosis (PTB) complicated by severe community-acquired pneumonia (SCAP) in the intensive care unit (ICU) remains high. We aimed to develop a rapid and simple model for the early assessment and stratification of prognosis in these patients. All adult patients with PTB complicated by SCAP admitted to the ICU of a tertiary hospital in Chengdu, Sichuan, China between 2019 and 2021 (development cohort) and 2022 (validation cohort) were retrospectively included. Data on demographics, comorbidities, laboratory values, and interventions were collected. The outcome was the 28-day mortality. Stepwise backward multivariate Cox analysis was used to develop a mortality risk prediction score model. Receiver operating characteristic (ROC) and calibration curves were used to evaluate the model's predictive efficiency. Decision curve analysis (DCA) was used to validate the model's clinical value and impact on decision making. Overall, 357 and 168 patients were included in the development and validation cohorts, respectively. The Pulmonary Tuberculosis Severity Index (PTSI) score included long-term use of glucocorticoid, body mass index (BMI) <18.5 kg/m2, diabetes, blood urea nitrogen (BUN) ≥7.14 mmol/L, PO2/FiO2 <150 mmHg, and vasopressor use. The area under the ROC curve (AUC) values were 0.817 (95% CI: 0.772-0.863) and 0.814 for the development and validation cohorts, respectively. The PTSI score had a higher AUC than the APACHE II, SOFA, and CURB-65 score. The calibration curves indicated good calibration in both cohorts. The DCA of the PTSI score indicated the high clinical application of the model compared with the APACHE II and SOFA scores. This prognostic tool was designed to rapidly evaluate the 28-day mortality risk in individuals with PTB complicated by SCAP. It can stratify this patient group into relevant risk categories, guide targeted interventions, and enhance clinical decision making, thereby optimizing patient care and improving outcomes.

Clinical Utility of the aMAP Score for Predicting Hepatocellular Carcinoma Development in Patients with Chronic Hepatitis B.

This study aimed to evaluate the efficacy of the aMAP score and compare it with other risk scores for predicting hepatocellular carcinoma (HCC) development in Thai patients with chronic hepatitis B (CHB). We retrospectively analyzed patients with CHB between 1 January 2008 and 31 December 2019. Data on demographics, clinical parameters, cirrhosis status, HCC imaging, and alpha fetoprotein surveillance were collected to calculate the aMAP score (0-100) based on age, sex, albumin-bilirubin level, and platelet count. Of the 1060 patients analyzed, 789 were eligible, of whom 51 developed HCC. The cumulative HCC incidences in the low-, moderate-, and high-risk groups at 3, 5, and 10 years were significantly different (log-rank, p < 0.0001). The area under the receiver operating characteristic curves (AUROCs) of the aMAP scores for predicting HCC at 3, 5, and 10 years were 0.748, 0.777, and 0.784, respectively. Among the risk scores, the CU-HCC score had the highest AUROCs (0.823) for predicting 5-year HCC development. The aMAP score is a valuable tool for predicting HCC risk in Thai patients with CHB and can enhance surveillance strategies. However, its performance is inferior to that of the CU-HCC score, suggesting the need for new predictive tools for HCC surveillance.

Development and validation of machine learning algorithms based on electrocardiograms for cardiovascular diagnoses at the population level

Artificial intelligence-enabled electrocardiogram (ECG) algorithms are gaining prominence for the early detection of cardiovascular (CV) conditions, including those not traditionally associated with conventional ECG measures or expert interpretation. This study develops and validates such models for simultaneous prediction of 15 different common CV diagnoses at the population level. We conducted a retrospective study that included 1,605,268 ECGs of 244,077 adult patients presenting to 84 emergency departments or hospitals, who underwent at least one 12-lead ECG from February 2007 to April 2020 in Alberta, Canada, and considered 15 CV diagnoses, as identified by International Classification of Diseases, 10th revision (ICD-10) codes: atrial fibrillation (AF), supraventricular tachycardia (SVT), ventricular tachycardia (VT), cardiac arrest (CA), atrioventricular block (AVB), unstable angina (UA), ST-elevation myocardial infarction (STEMI), non-STEMI (NSTEMI), pulmonary embolism (PE), hypertrophic cardiomyopathy (HCM), aortic stenosis (AS), mitral valve prolapse (MVP), mitral valve stenosis (MS), pulmonary hypertension (PHTN), and heart failure (HF). We employed ResNet-based deep learning (DL) using ECG tracings and extreme gradient boosting (XGB) using ECG measurements. When evaluated on the first ECGs per episode of 97,631 holdout patients, the DL models had an area under the receiver operating characteristic curve (AUROC) of <80% for 3 CV conditions (PTE, SVT, UA), 80–90% for 8 CV conditions (CA, NSTEMI, VT, MVP, PHTN, AS, AF, HF) and an AUROC > 90% for 4 diagnoses (AVB, HCM, MS, STEMI). DL models outperformed XGB models with about 5% higher AUROC on average. Overall, ECG-based prediction models demonstrated good-to-excellent prediction performance in diagnosing common CV conditions.

Development and Validation of a Prediction Model for Acute Hypotensive Events in Intensive Care Unit Patients.

Background: Persistent hypotension in the intensive care unit (ICU) is associated with increased mortality. Predicting acute hypotensive events can lead to timely intervention. We aimed to develop a prediction model of acute hypotensive events in patients admitted to the ICU. Methods: We included adult patients admitted to the Nagoya City University (NCU) Hospital ICU between January 2018 and December 2021 for model training and internal validation. The MIMIC-III database was used for external validation. A hypotensive event was defined as a mean arterial pressure < 60 mmHg for at least 5 min in 10 min. The input features were age, sex, and time-series data for vital signs. We compared the area under the receiver-operating characteristic curve (AUROC) of three machine-learning algorithms: logistic regression, the light gradient boosting machine (LightGBM), and long short-term memory (LSTM). Results: Acute hypotensive events were found in 1325/1777 (74.6%) and 2691/5266 (51.1%) of admissions in the NCU and MIMIC-III cohorts, respectively. In the internal validation, the LightGBM model had the highest AUROC (0.835), followed by the LSTM (AUROC 0.834) and logistic regression (AUROC 0.821) models. Applying only blood pressure-related features, the LSTM model achieved the highest AUROC (0.843) and consistently showed similar results in external and internal validation. Conclusions: The LSTM model using only blood pressure-related features had the highest AUROC with comparable performance in external validation.

Automated annotation of disease subtypes

BackgroundDistinguishing diseases into distinct subtypes is crucial for study and effective treatment strategies. The Open Targets Platform (OT) integrates biomedical, genetic, and biochemical datasets to empower disease ontologies, classifications, and potential gene targets. Nevertheless, many disease annotations are incomplete, requiring laborious expert medical input. This challenge is especially pronounced for rare and orphan diseases, where resources are scarce. MethodsWe present a machine learning approach to identifying diseases with potential subtypes, using the approximately 23,000 diseases documented in OT. We derive novel features for predicting diseases with subtypes using direct evidence. Machine learning models were applied to analyze feature importance and evaluate predictive performance for discovering both known and novel disease subtypes. ResultsOur model achieves a high (89.4%) ROC AUC (Area Under the Receiver Operating Characteristic Curve) in identifying known disease subtypes. We integrated pre-trained deep-learning language models and showed their benefits. Moreover, we identify 515 disease candidates predicted to possess previously unannotated subtypes. ConclusionsOur models can partition diseases into distinct subtypes. This methodology enables a robust, scalable approach for improving knowledge-based annotations and a comprehensive assessment of disease ontology tiers. Our candidates are attractive targets for further study and personalized medicine, potentially aiding in the unveiling of new therapeutic indications for sought-after targets.

Validation of Modified Objective Prognostic Score in Patients with Advanced Cancer in Taiwan.

Modified versions of the Objective Prognostic Score (mOPS) needs to be validated to reflect practical palliative care circumstances in Taiwan. We compared the abilities of an mOPS score of 1.5 or higher versus a Karnofsky Performance Status (KPS) score of 30 or lower to predict 2-week mortality in patients with advanced cancer in Taiwan. Observational study. We performed a secondary analysis of an international multicenter cohort study of patients in East Asia. Participants were inpatients with advanced cancer in palliative care units (PCUs) in Taiwan. We compared the mOPS-B model, which does not require laboratory tests, with the KPS in a 2-week survival timeframe. We compared the accuracy of the prognostic models using sensitivity, specificity, and area under the receiver operating characteristic curve (AUROC). Calibration plots and net reclassification indices (NRI) for 2-week survival were compared between the two models. Differences in survival between the higher- and lower-scoring groups of each model were identified using the log-rank test. We included 317 patients, with a median survival of 14.0 days. The mOPS-B had a high sensitivity (0.82) and high AUROC value (0.69). By contrast, the KPS demonstrated good sensitivity (0.77) and an acceptable AUROC value (0.65) for predicting 2-week survival. The calibration plot did not demonstrate satisfactory agreement between the actual and predicted survival times in either the mOPS-B or the KPS groups. Our NRI was positive (absolute value: 22%), indicating that mOPS-B predicted 2-week survival better than KPS. The mOPS-B may serve better than the KPS as a screening tool for admission to PCUs in Taiwan because it was more accurate at predicting 2-week survival.

BiMPADR: A Deep Learning Framework for Predicting Adverse Drug Reactions in New Drugs.

Detecting the unintended adverse reactions of drugs (ADRs) is a crucial concern in pharmacological research. The experimental validation of drug-ADR associations often entails expensive and time-consuming investigations. Thus, a computational model to predict ADRs from known associations is essential for enhanced efficiency and cost-effectiveness. Here, we propose BiMPADR, a novel model that integrates drug gene expression into adverse reaction features using a message passing neural network on a bipartite graph of drugs and adverse reactions, leveraging publicly available data. By combining the computed adverse reaction features with the structural fingerprints of drugs, we predict the association between drugs and adverse reactions. Our models obtained high AUC (area under the receiver operating characteristic curve) values ranging from 0.861 to 0.907 in an external drug validation dataset under differential experiment conditions. The case study on multiple BET inhibitors also demonstrated the high accuracy of our predictions, and our model's exploration of potential adverse reactions for HWD-870 has contributed to its research and development for market approval. In summary, our method would provide a promising tool for ADR prediction and drug safety assessment in drug discovery and development.

Validation of three health literacy screening questions compared with S-TOFHLA in a low-income diverse English- and Spanish-Speaking population.

Clinicians need a tool to gauge patients' ability to understand health conditions and treatment options. The Short-form Test of Functional Health Literacy in Adults (S-TOFHLA) is the gold standard for this, but its length is prohibitive for use in clinical settings. This study seeks to validate a novel three-item question set for predicting health literacy. This cross-sectional study utilized an in-person questionnaire alongside the S-TOFHLA. The sample included 2027 English- and Spanish-speaking adults (≥18years) recruited from primary care practices serving a low-income eastern Pennsylvania community. Most patients (57.7%) identified as Hispanic. Diagnostic accuracy of each question and aggregated scores were assessed against the validated survey by calculating the area under the receiver operating characteristic (AUROC) curve. Questions in the 'Problems Learning' and 'Help Reading' domains (AUROC 0.66 for each) performed better than the 'Confident Forms' question (AUROC 0.64). Summing all three scores resulted in an even higher AUROC curve (0.71). Cronbach's alpha of the combined items was 0.696. Study results suggest that any of the three questions are viable options for screening health literacy levels of diverse patients in primary care clinical settings. However, they perform better as a summed score than when used individually.

Artificial intelligence for ultrasound microflow imaging in breast cancer diagnosis.

To develop and evaluate artificial intelligence (AI) algorithms for ultrasound (US) microflow imaging (MFI) in breast cancer diagnosis. We retrospectively collected a dataset consisting of 516 breast lesions (364 benign and 152 malignant) in 471 women who underwent B-mode US and MFI. The internal dataset was split into training (n = 410) and test datasets (n = 106) for developing AI algorithms from deep convolutional neural networks from MFI. AI algorithms were trained to provide malignancy risk (0-100%). The developed AI algorithms were further validated with an independent external dataset of 264 lesions (229 benign and 35 malignant). The diagnostic performance of B-mode US, AI algorithms, or their combinations was evaluated by calculating the area under the receiver operating characteristic curve (AUROC). The AUROC of the developed three AI algorithms (0.955-0.966) was higher than that of B-mode US (0.842, P < 0.0001). The AUROC of the AI algorithms on the external validation dataset (0.892-0.920) was similar to that of the test dataset. Among the AI algorithms, no significant difference was found in all performance metrics combined with or without B-mode US. Combined B-mode US and AI algorithms had a higher AUROC (0.963-0.972) than that of B-mode US (P < 0.0001). Combining B-mode US and AI algorithms significantly decreased the false-positive rate of BI-RADS category 4A lesions from 87% to 13% (P < 0.0001). AI-based MFI diagnosed breast cancers with better performance than B-mode US, eliminating 74% of false-positive diagnoses in BI-RADS category 4A lesions.

Abstract LB243: Deep learning-based molecular characterization of lung cancers from never smokers using hematoxylin and eosin-stained whole slide images

Abstract Purpose: This study employs cutting-edge deep learning techniques to comprehensively analyze hematoxylin and eosin-stained (H&amp;E) whole slide images (WSIs) to predict which tumors carry driver gene alterations, mutational signatures, and additional molecular features among lung cancers from never-smokers (LCINS)— the seventh leading cause of cancer death worldwide. Method: A total of 464 H&amp;E-stained WSIs of lung adenocarcinomas from the Sherlock-Lung study were utilized, with 325 WSIs for model training and 139 for testing. Alongside the WSIs, genomic data, including mutations (overall nonsynonymous mutations and driver mutations only), fusions, or copy numbers in driver genes (TP53, KRAS, EGFR, CDKN2A, MDM2, ALK, RBM10), mutational signatures (APOBEC), molecular features (Kataegies, Whole Genome Doubling (WGD) status, Tumor Mutational Burden (TMB)), and specific hotspot driver mutation in EGFR (p.L858R and p.E746_A750del) and KRAS (p.G12C, p.G12V, and p.G12D), were included for all tumors. Each WSI was assigned different labels based on the presence or absence of these genomic alterations or features. We used a customized multilabel binary deep learning model based on ResNet50, a residual convolutional neural network, for analyzing the data. The network was trained from scratch, with an initial epoch set to 100. Results: Our methodology demonstrated high predictive performance, measured by the area under the receiver operating characteristic (AUROC). In the presence or absence category, we achieved high AUROC for detecting any alterations in these driver genes: TP53 (0.81), KRAS (0.92), EGFR (0.93), CDKN2A deletion (0.88), MDM2 amplification (0.94), and ALK fusion (0.86). We also evaluated molecular features, obtaining AUROC scores of 0.79 for WGD status and 0.86 for Kataegies. Moderate AUROC scores were observed for tumors with RBM10 mutations (0.67), high TMB (0.64), and APOBEC signatures (0.56). When focusing on driver mutations only, we achieved high AUROC scores for EGFR (0.88), KRAS (0.84), TP53 (0.93), and RBM10 (0.78). In addition, our model successfully predicted tumors with specific mutations in EGFR (p.L858R = 0.79; p.E746_A750del = 0.77) and KRAS (p.G12C = 0.77). Performance was suboptimal for KRAS p.G12V (0.41) and KRAS p.G12D (0.47). Conclusions: Our deep learning network achieves high prediction scores in identifying tumors with critical driver gene alterations and actionable mutations, holding promise for potential clinical use. In the future, the model could be optimized as a screening assay to guide molecular testing and therapeutic management of patients with LCINS. Citation Format: Monjoy Saha, Tongwu Zhang, Praphulla Bhawsar, Wei Zhao, Jianxin Shi, Soo Ryum Yang, Jonas Almeida, Maria Teresa Landi. Deep learning-based molecular characterization of lung cancers from never smokers using hematoxylin and eosin-stained whole slide images [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(7_Suppl):Abstract nr LB243.