Area Under The ROC Curve Research Articles

NIMG-18. RAPID INTRAOPERATIVE DIFFERENTIATION OF PRIMARY CNS LYMPHOMA FROM COMMON CNS TUMORS USING STIMULATED RAMAN HISTOLOGY AND DEEP LEARNING

Abstract Accurate intraoperative diagnosis is crucial for decision-making during tumor surgery and stereotactic-guided biopsies. Differentiating between primary CNS lymphoma (PCNSL) and non-PCNSL entities such as gliomas and metastatic cancers presents significant challenges due to overlapping histomorphological features, time constraints, and the differing further therapy strategies required. Our study utilized deep learning and stimulated Raman histology (SRH) to address this challenge. We imaged unprocessed, label-free tissue samples intraoperatively using a portable Raman scattering microscope, generating virtual H&E-like images within less than five minutes. Our training set comprised over 54,000 SRH tumor patch images, including various brain tumors. We developed a deep learning pipeline utilizing a ResNet-50 feature extractor and the BYOL (Bootstrap Your Own Latent) self-supervised learning algorithm. We trained a binary linear classifier to distinguish between PCNSL and non-PCNSL entities. Data were collected from New York University (NYU), the University of Cologne (UKK), and the University of Michigan (UM), including 263 SRH whole slide images from 177 patients sourced from surgical resections and stereotactic-guided biopsies. In an international bicentric test cohort (N = 100, NYU and UKK), the model achieved an overall accuracy of 91.1%, an area under the receiver operating characteristic curve (AUROC) of 0.983, a sensitivity of 89.3%, and a specificity of 100% compared to the final formalin-fixed, paraffin-embedded-based diagnosis. Our algorithm can accurately extract typical histomorphological features to help differentiate between PCNSL and non-PCNSL tumors. These results underscore the potential of deep learning models to enhance the accuracy of intraoperative diagnoses for PCNSL and non-PCNSL entities, thereby improving intraoperative decision-making and further treatment strategies by supporting neuropathologists and surgeons.

The Development and Validation of a Glaucoma Health Score for Glaucoma Screening Based on Clinical Parameters and Optical Coherence Tomography Metrics

Background/Objectives: This study aims to develop and validate a Glaucoma Health Score (GHS) that incorporates multiple individual glaucoma risk factors to enhance glaucoma detection in screening environments. Methods: The GHS was developed using a retrospective dataset from two clinical sites, including both eyes of glaucoma patients and controls. The model incorporated age, central corneal thickness, intraocular pressure, pattern standard deviation from a visual field threshold 24-2 test, and two parameters from an optical coherence tomography (OCT) test: the average circumpapillary retinal nerve fiber layer thickness and the minimum thickness of the six sectors of the macular ganglion cell plus the inner plexiform layer. The GHS was then validated in two independent datasets: one from primary care sites using Maestro OCT data (test dataset 1) and another from an academic center using DRI OCT Triton (test dataset 2). Results: Both eyes of 51 glaucoma patients and 67 controls were included in the development dataset. Setting the GHS cutoff at 75 points out of 100, test dataset 1, which comprised 41 subjects with glaucoma and 41 healthy controls, achieved an area under the receiver operating characteristic curve (AUROC) of 0.98, with a sensitivity of 71% and specificity of 98%; test dataset 2, which included 53 patients with glaucoma and 53 healthy controls, resulted in an AUROC of 0.95, with a sensitivity of 75% and specificity of 96%. A decision curve analysis across all datasets demonstrated a higher net benefit for the GHS model compared to individual OCT parameters. Conclusions: The GHS offers a feasible, standardized approach for early detection of glaucoma, providing strong specificity and acceptable sensitivity, with clear decision-making benefits in screening settings.

Development and validation of a predictive scoring model for complications following endoscopic endonasal skull base surgery.

The endoscopic endonasal approach (EEA) has evolved into an established technique in skull base surgery. The authors previously examined 1002 EEA procedures and reported factors associated with postoperative complications. Here they report the development and validation of a scoring model based on risk factors to better predict complications following EEA. The authors developed an optimized EEA scoring model for predicting postoperative complications as evidenced by the area under the receiver operating characteristic (AUROC) curve using their previously published data in addition to data collected from the subsequent 292 EEA procedures from years 2010-2020. The model was built systematically by evaluating the contributions that different variables had on the overall predictive ability of the model. The aim was to design a model containing as few variables as possible for practicality and to facilitate calculation and use at the bedside. The Clavien-Dindo grading system was used to classify complications into grades I-V based on the level of intervention that was required to manage the complication, with grades III-V considered to be higher-grade (i.e., those requiring reoperation or ICU-level care or death). The authors identified 1294 EEA operations performed between July 2010 and July 2020 that met their inclusion criteria. Higher-grade complications were identified following 135 EEA operations. The variables that were ultimately included in the model were age, BMI, operative time, meningioma, chordoma, expanded intradural approach, and nasoseptal flap use. The final model yielded an acceptable AUROC curve of 0.72 and predicted a stepwise increase in the rate of higher-grade complications as the score increased. A score of 0-2 (low) on the grading system was associated with an average complication rate of 5.1%. A score of 3-5 (medium) was associated with an average complication rate of 12.6%. A score of 6 or above (high) was associated with an average complication rate of 26%. This EEA complications scoring model accurately categorizes patients into low-, medium-, and high-risk groups with readily obtained variables. A high score in this complications model does not suggest that a patient is ineligible for surgery, but rather highlights the importance of thorough case selection, operating with caution, and appropriate preoperative counseling.

Performance of Risk Models for Antimicrobial Resistance in Adult Patients With Sepsis

The results of prediction models that stratify patients with sepsis and risk of resistant gram-negative bacilli (GNB) infections inform treatment guidelines. However, these models do not extrapolate well across hospitals. To assess whether patient case mix and local prevalence rates of resistance contributed to the variable performance of a general risk stratification GNB sepsis model for community-onset and hospital-onset sepsis across hospitals. This was a retrospective cohort study conducted from January 2016 and October 2021. Adult patients with sepsis at 10 acute-care hospitals in rural and urban areas across Missouri and Illinois were included. Inclusion criteria were blood cultures indicating sepsis, having received 4 days of antibiotic treatment, and having organ dysfunction (vasopressor use, mechanical ventilation, increased creatinine or bilirubin levels, and thrombocytopenia). Analyses were completed in April 2024. The model included demographic characteristics, comorbidities, vital signs, laboratory values, procedures, and medications administered. Culture results were stratified for ceftriaxone-susceptible GNB (SS), ceftriaxone-resistant but cefepime-susceptible GNB (RS), and ceftriaxone- and cefepime-resistant GNB (RR). Negative cultures and other pathogens were labeled SS. Deep learning models were developed separately for community-onset (patient presented with sepsis) and hospital-onset (sepsis developed ≥48 hours after admission) sepsis. The models were tested across hospitals and patient subgroups. Models were assessed using area under the receiver operating characteristic curve (AUROC) and area under precision recall curve (AUPRC). A total of 39 893 patients with 85 238 sepsis episodes (43 207 [50.7%] community onset; 42 031 [48.3%] hospital onset) were included. Median (IQR) age was 65 (54-74) years, 21 241 patients (53.2%) were male, and 18 830 (47.2%) had a previous episode of sepsis. RS contributed to 3.9% (1667 episodes) and 5.7% (2389 episodes) of community-onset and hospital-onset sepsis episodes, respectively, and RR contributed to 1.8% (796 episodes) and 3.9% (1626 episodes), respectively. Previous infections and exposure to antibiotics were associated with the risk of resistant GNB. For example, in community-onset sepsis, 375 RR episodes (47.1%), 420 RS episodes (25.2%) and 3483 of 40 744 (8.5%) SS episodes were among patients with resistance to antimicrobial drugs (P < .001). The AUROC and AUPRC results varied across hospitals and patient subgroups for both community-onset and hospital-onset sepsis. AUPRC values correlated with the prevalence rates of resistant GNB (R = 0.79; P = .001). In this cohort study of 39 893 patients with sepsis, variable model performance was associated with prevalence rates of antimicrobial resistance rather than patient case mix. This variability suggests caution is needed when using generalized models for predicting resistant GNB etiologies in sepsis.

Assessing polyomic risk to predict Alzheimer's disease using a machine learning model.

Alzheimer's disease (AD) is the most common form of dementia in the elderly. Given that AD neuropathology begins decades before symptoms, there is a dire need for effective screening tools for early detection of AD to facilitate early intervention. Here, we used tree-based and deep learning methods to train polyomic prediction models for AD affection status and age at onset, employing genomic, proteomic, metabolomic, and drug use data from UK Biobank. We used SHAP to determine the feature's importance. Our best-performing polyomic model achieved an area under the receiver operating characteristics curve (AUROC) of 0.87. We identified GFAP and CXCL17 proteins to be the strongest predictors of AD, besides apolipoprotein E (APOE) alleles. Increasing the number of cases by including "AD-by-proxy" cases did not improve AD prediction. Among the four modalities, genomics, and proteomics were the most informative modality based on AUROC (area under the receiver operating characteristic curve). Our data suggest that two blood-based biomarkers (glial fibrillary acidic protein [GFAP] and CXCL17) may be effective for early presymptomatic prediction of AD. We developed a polyomic model to predict AD and age-at-onset using omics and medication use data from EHR. We identified GFAP and CXCL17 proteins to be the strongest predictors of AD, besides APOE alleles. "AD-by-proxy" cases, if used in training, do not improve AD prediction. Proteomics was the most informative modality overall for affection status and AAO prediction.

Reliability and utility of blood glucose levels in the periodontal pockets of patients with type 2 diabetes mellitus: a cross-sectional study.

Several studies have measured gingival blood glucose (GBG) levels, but few have confirmed systematic bias using Bland-Altman analysis. This study compared the effectiveness of GBG levels with that of fingertip blood glucose (FTBG) levels using Bland-Altman and receiver operating characteristic (ROC) analyses. A total of 15 healthy volunteers and 15 patients with type 2 diabetes were selected according to inclusion and exclusion criteria. Each group comprised eight male and seven female participants. The GBG and FTBG levels were measured using a self-monitoring blood glucose device after periodontal examination. Pearson's product‒moment correlation and simple linear regression analyses were performed. In addition, Bland‒Altman analysis was also performed to assess the degree of agreement between the two methods. ROC analysis was conducted to determine the sensitivity, specificity, and cutoff values for patients with diabetes. The area under the ROC curve (AUC) was used to identify significant differences. The mean GBG and FTBG levels were 120 ± 44.8 mg/dL and 137 ± 45.1, respectively, for the whole sample. The mean GBG and FTBG levels were 145 ± 47.2 mg/dL and 163 ± 49.1, respectively, in the diabetes group. The mean GBG and FTBG levels in the nondiabetes group were 95.3 ± 25.2 and 111 ± 18.8, respectively. Patients with diabetes were more likely to have a probing pocket depth (PPD) of ≥4 mm at the sampled site. Pearson's product‒moment correlation and simple linear regression analyses revealed a significant correlation between the GBG and FTBG measurements. Bland-Altman analysis revealed that GBG and FTBG measurements differed significantly among all participants; however, no significant differences were observed among the patients with diabetes (mean difference (MD) ± standard deviation (SD) = -18.1 ± 34.2, 95% confidence interval (CI) [-37.0 to 0.88]) or among the participants with a PPD of ≥4 mm (MD ± SD = -15.2 ± 30.4, 95% CI [-30.8 to 0.43]). The sensitivity, specificity, and cutoff values of the GBG measurements for detecting diabetes were 80%, 93%, and 123.5 mg/dL, respectively. The sensitivity, specificity, and cutoff values of the FTBG measurements for detecting diabetes were 73%, 87%, and 134.0 mg/dL, respectively. No significant differences were observed between the AUCs (0.078, 95% CI [-0.006 to 0.161]). The GBG measurements aligned with the FTBG measurements in the patients with diabetes and among the participants with a PPD of ≥4 mm. Patients with diabetes were more likely to have a PPD of ≥4 mm at the sampled site, GBG levels can be used to screen for type 2 diabetes in dental clinics.

Improved Detection Accuracy of Chronic Vertebral Compression Fractures by Integrating Height Loss Ratio and Deep Learning Approaches

Objectives: This study aims to assess the limitations of the height loss ratio (HLR) method and introduce a new approach that integrates a deep learning (DL) model to enhance vertebral compression fracture (VCF) detection performance. Methods: We conducted a retrospective study on 589 patients with chronic VCFs. We compared four different methods: HLR-only, DL-only, a combination of HLR and DL for positive VCF, and a combination of HLR and DL for negative VCF. The models were evaluated using dice similarity coefficient, sensitivity, specificity, and area under the receiver operating characteristic curve (AUROC). Results: The combined method (HLR + DL, positive) demonstrated the best performance with an AUROC of 0.968, sensitivity (94.95%), and specificity (90.59%). The HLR-only and the HLR + DL (negative) also showed strong discriminatory power, with AUROCs of 0.948 and 0.947, respectively. The DL-only model achieved the highest specificity (95.92%) but exhibited lower sensitivity (82.83%). Conclusions: Our study highlights the limitations of the HLR method in detecting chronic VCFs and demonstrates the improved performance of combining HLR with DL models.

Unraveling Uncertainty: The Impact of Biological and Analytical Variation on the Prediction Uncertainty of Categorical Prediction Models.

Interest in prediction models, including machine learning (ML) models, based on laboratory data has increased tremendously. Uncertainty in laboratory measurements and predictions based on such data are inherently intertwined. This study developed a framework for assessing the impact of biological and analytical variation on the prediction uncertainty of categorical prediction models. Practical application was demonstrated for the prediction of renal function loss (Chronic Kidney Disease Epidemiology Collaboration [CKD-EPI] equation) and 31-day mortality (advanced ML model) in 6360 emergency department patients. Model outcome was calculated in 100 000 simulations of variation in laboratory parameters. Subsequently, the percentage of discordant predictions was calculated with the original prediction as reference. Simulations were repeated assuming increasing levels of analytical variation. For the ML model, area under the receiver operating characteristic curve (AUROC), sensitivity, and specificity were 0.90, 0.44, and 0.96, respectively. At base analytical variation, the median [2.5th-97.5th percentiles] percentage of discordant predictions was 0% [0%-28.8%]. In addition, 7.2% of patients had >5% discordant predictions. At 6× base analytical variation, the median [2.5th-97.5th percentiles] percentage of discordant predictions was 0% [0%-38.8%]. In addition, 11.7% of patients had >5% discordant predictions. However, the impact of analytical variation was limited compared with biological variation. AUROC, sensitivity, and specificity were not affected by variation in laboratory parameters. The impact of biological and analytical variation on the prediction uncertainty of categorical prediction models, including ML models, can be estimated by the occurrence of discordant predictions in a simulation model. Nevertheless, discordant predictions at the individual level do not necessarily affect model performance at the population level.

Predicting Intensive Care Admission in Children with Acute Asthma: A Meta-Analysis of Predictive Models

Background: Acute asthma is a common cause of pediatric emergency department visits and hospitalizations. Early identification of children at high risk of requiring intensive care unit (ICU) admission is crucial for optimal management and resource allocation. This meta-analysis aimed to evaluate the performance of predictive models for ICU admission in children presenting with acute asthma. Methods: A systematic search of PubMed, Embase, and Cochrane Library was conducted for studies published between 2013 and 2024 that developed or validated predictive models for ICU admission in children with acute asthma. Studies reporting sensitivity, specificity, and area under the receiver operating characteristic curve (AUROC) were included. Methodological quality was assessed using the QUADAS-2 tool. Pooled estimates of diagnostic accuracy were calculated using a random-effects model. Results: Six studies (n = 2,850 children) met the inclusion criteria. The predictive models included clinical features (respiratory rate, oxygen saturation, accessory muscle use), lung function measures (peak expiratory flow rate), and blood gas analysis. Pooled sensitivity ranged from 0.71 (95% CI 0.59-0.82) to 0.78 (95% CI 0.72-0.83), specificity from 0.79 (95% CI 0.75-0.83) to 0.86 (95% CI 0.78-0.91), and AUROC from 0.79 (95% CI 0.72-0.86) to 0.88 (95% CI 0.84-0.92). Conclusion: Several predictive models demonstrate moderate to high accuracy in identifying children with acute asthma at risk of ICU admission. However, heterogeneity in model performance highlights the need for further research to validate existing models in diverse populations and develop more robust tools to guide clinical decision-making.

Predicting temporomandibular disorders in adults using interpretable machine learning methods: a model development and validation study

IntroductionTemporomandibular disorders (TMD) have a high prevalence and complex etiology. The purpose of this study was to apply a machine learning (ML) approach to identify risk factors for the occurrence of TMD in adults and to develop and validate an interpretable predictive model for the risk of TMD in adults.MethodsA total of 949 adults who underwent oral examinations were enrolled in our study. 5 different ML algorithms were used for model development and comparison, and feature selection was performed by feature importance ranking and feature decreasing methods. Several evaluation indexes, including the area under the receiver-operating-characteristic curve (AUC), were used to compare the predictive performance. The precision-recall curve (PR), calibration curve, and decision curve analysis (DCA) further assessed the accuracy and clinical utility of the model.ResultsThe performance of the random forest (RF) model was the best among the 5 ML models. An interpretable RF model was developed with 7 features (gender, malocclusion, unilateral chewing, chewing hard substances, grinding teeth, clenching teeth, and anxiety). The AUCs of the final model on the training set, internal validation set, and external test set were 0.892, 0.854, and 0.857, respectively. Calibration and DCA curves showed high accuracy and clinical applicability of the model.DiscussionAn efficient and interpretable TMD risk prediction model for adults was successfully developed using the ML method. The model not only has good predictive performance, but also enhances the clinical application value of the model through the SHAP method. This model can provide clinicians with a practical and efficient TMD risk assessment tool that can help them better predict and assess TMD risk in adults, supporting more efficient disease management and targeted medical interventions.

A Deep Learning Model to Predict Breast Implant Texture Types Using Ultrasonography Images: Feasibility Development Study.

Breast implants, including textured variants, have been widely used in aesthetic and reconstructive mammoplasty. However, the textured type, which is one of the shell texture types of breast implants, has been identified as a possible etiologic factor for lymphoma, specifically breast implant-associated anaplastic large cell lymphoma (BIA-ALCL). Identifying the shell texture type of the implant is critical to diagnosing BIA-ALCL. However, distinguishing the shell texture type can be difficult due to the loss of human memory and medical history. An alternative approach is to use ultrasonography, but this method also has limitations in quantitative assessment. This study aims to determine the feasibility of using a deep learning model to classify the shell texture type of breast implants and make robust predictions from ultrasonography images from heterogeneous sources. A total of 19,502 breast implant images were retrospectively collected from heterogeneous sources, including images captured from both Canon and GE devices, images of ruptured implants, and images without implants, as well as publicly available images. The Canon images were trained using ResNet-50. The model's performance on the Canon dataset was evaluated using stratified 5-fold cross-validation. Additionally, external validation was conducted using the GE and publicly available datasets. The area under the receiver operating characteristic curve (AUROC) and the area under the precision-recall curve (PRAUC) were calculated based on the contribution of the pixels with Gradient-weighted Class Activation Mapping (Grad-CAM). To identify the significant pixels for classification, we masked the pixels that contributed less than 10%, up to a maximum of 100%. To assess the model's robustness to uncertainty, Shannon entropy was calculated for 4 image groups: Canon, GE, ruptured implants, and without implants. The deep learning model achieved an average AUROC of 0.98 and a PRAUC of 0.88 in the Canon dataset. The model achieved an AUROC of 0.985 and a PRAUC of 0.748 for images captured with GE devices. Additionally, the model predicted an AUROC of 0.909 and a PRAUC of 0.958 for the publicly available dataset. This model maintained the PRAUC values for quantitative validation when masking up to 90% of the least-contributing pixels and the remnant pixels in breast shell layers. Furthermore, the prediction uncertainty increased in the following order: Canon (0.066), GE (0072), ruptured implants (0.371), and no implants (0.777). We have demonstrated the feasibility of using deep learning to predict the shell texture type of breast implants. This approach quantifies the shell texture types of breast implants, supporting the first step in the diagnosis of BIA-ALCL.

TPGPred: A Mixed-Feature-Driven Approach for Identifying Thermophilic Proteins Based on GradientBoosting

Thermophilic proteins maintain their stability and functionality under extreme high-temperature conditions, making them of significant importance in both fundamental biological research and biotechnological applications. In this study, we developed a machine learning-based thermophilic protein GradientBoosting prediction model, TPGPred, designed to predict thermophilic proteins by leveraging a large-scale dataset of both thermophilic and non-thermophilic protein sequences. By combining various machine learning algorithms with feature-engineering methods, we systematically evaluated the classification performance of the model, identifying the optimal feature combinations and classification models. Trained on a large public dataset of 5652 samples, TPGPred achieved an Accuracy score greater than 0.95 and an Area Under the Receiver Operating Characteristic Curve (AUROC) score greater than 0.98 on an independent test set of 627 samples. Our findings offer new insights into the identification and classification of thermophilic proteins and provide a solid foundation for their industrial application development.

Natural Language Processing of Clinical Documentation to Assess Functional Status in Patients With Heart Failure

Serial functional status assessments are critical to heart failure (HF) management but are often described narratively in documentation, limiting their use in quality improvement or patient selection for clinical trials. To develop and validate a deep learning natural language processing (NLP) strategy for extracting functional status assessments from unstructured clinical documentation. This diagnostic study used electronic health record data collected from January 1, 2013, through June 30, 2022, from patients diagnosed with HF seeking outpatient care within 3 large practice networks in Connecticut (Yale New Haven Hospital [YNHH], Northeast Medical Group [NMG], and Greenwich Hospital [GH]). Expert-annotated notes were used for NLP model development and validation. Data were analyzed from February to April 2024. Development and validation of NLP models to detect explicit New York Heart Association (NYHA) classification, HF symptoms during activity or rest, and frequency of functional status assessments. Outcomes of interest were model performance metrics, including area under the receiver operating characteristic curve (AUROC), and frequency of NYHA class documentation and HF symptom descriptions in unannotated notes. This study included 34 070 patients with HF (mean [SD] age 76.1 [12.6] years; 17 728 [52.0]% female). Among 3000 expert-annotated notes (2000 from YNHH and 500 each from NMG and GH), 374 notes (12.4%) mentioned NYHA class and 1190 notes (39.7%) described HF symptoms. The NYHA class detection model achieved a class-weighted AUROC of 0.99 (95% CI, 0.98-1.00) at YNHH, the development site. At the 2 validation sites, NMG and GH, the model achieved class-weighted AUROCs of 0.98 (95% CI, 0.96-1.00) and 0.98 (95% CI, 0.92-1.00), respectively. The model for detecting activity- or rest-related symptoms achieved an AUROC of 0.94 (95% CI, 0.89-0.98) at YNHH, 0.94 (95% CI, 0.91-0.97) at NMG, and 0.95 (95% CI, 0.92-0.99) at GH. Deploying the NYHA model among 182 308 unannotated notes from the 3 sites identified 23 830 (13.1%) notes with NYHA mentions, specifically 10 913 notes (6.0%) with class I, 12 034 notes (6.6%) with classes II or III, and 883 notes (0.5%) with class IV. An additional 19 730 encounters (10.8%) could be classified into functional status groups based on activity- or rest-related symptoms, resulting in a total of 43 560 medical notes (23.9%) categorized by NYHA, an 83% increase compared with explicit mentions alone. In this diagnostic study of 34 070 patients with HF, the NLP approach accurately extracted a patient's NYHA symptom class and activity- or rest-related HF symptoms from clinical notes, enhancing the ability to track optimal care delivery and identify patients eligible for clinical trial participation from unstructured documentation.

Machine learning for forecasting initial seizure onset in neonatal hypoxic-ischemic encephalopathy.

This study was undertaken to develop a machine learning (ML) model to forecast initial seizure onset in neonatal hypoxic-ischemic encephalopathy (HIE) utilizing clinical and quantitative electroencephalogram (QEEG) features. We developed a gradient boosting ML model (Neo-GB) that utilizes clinical features and QEEG to forecast time-dependent seizure risk. Clinical variables included cord blood gas values, Apgar scores, gestational age at birth, postmenstrual age (PMA), postnatal age, and birth weight. QEEG features included statistical moments, spectral power, and recurrence quantification analysis (RQA) features. We trained and evaluated Neo-GB on a University of California, San Francisco (UCSF) neonatal HIE dataset, augmenting training with publicly available neonatal electroencephalogram (EEG) datasets from Cork University and Helsinki University Hospitals. We assessed the performance of Neo-GB at providing dynamic and static forecasts with diagnostic performance metrics and incident/dynamic area under the receiver operating characteristic curve (iAUC) analyses. Model explanations were performed to assess contributions of QEEG features and channels to model predictions. The UCSF dataset included 60 neonates with HIE (30 with seizures). In subject-level static forecasting at 30 min after EEG initiation, baseline Neo-GB without time-dependent features had an area under the receiver operating characteristic curve (AUROC) of .76 and Neo-GB with time-dependent features had an AUROC of .89. In time-dependent evaluation of the initial seizure onset within a 24-h seizure occurrence period, dynamic forecast with Neo-GB demonstrated median iAUC = .79 (interquartile range [IQR] .75-.82) and concordance index (C-index) = .82, whereas baseline static forecast at 30 min demonstrated median iAUC = .75 (IQR .72-.76) and C-index = .69. Model explanation analysis revealed that spectral power, PMA, RQA, and cord blood gas values made the strongest contributions in driving Neo-GB predictions. Within the most influential EEG channels, as the preictal period advanced toward eventual seizure, there was an upward trend in broadband spectral power. This study demonstrates an ML model that combines QEEG with clinical features to forecast time-dependent risk of initial seizure onset in neonatal HIE. Spectral power evolution is an early EEG marker of seizure risk in neonatal HIE.

Predicting the Reparability of Rotator Cuff Tears: Machine Learning and Comparison With Previous Scoring Systems.

Repair of rotator cuff tear is not always feasible, depending on the severity. Although several studies have investigated factors related to reparability and various methods to predict it, inconsistent scoring methods and a lack of validation have hindered the utility of these methods. To develop machine learning models to predict the reparability of rotator cuff tears, compare them with previous scoring systems, and provide an accessible online model. Cohort study; Level of evidence, 3. Arthroscopic rotator cuff repairs for tears with both anteroposterior and mediolateral diameters >1 cm on preoperative magnetic resonance imaging were included and divided into a training set (70%) and an internal validation set (30%). For external validation, rotator cuff repairs performed by 2 different surgeons were included in a test set. Machine learning models and a newly adjusted scoring system were developed using the training set. The performance of the models including the adjusted scoring system and 2 previous scoring systems were compared using the test set. The performance was assessed using metrics such as the area under the receiver operating characteristic curve (AUROC) and compared using the net reclassification improvement based on the adjusted scoring system. A total of 429 patients were included for the training and internal validation set, and 112 patients were included for the test set. An elastic-net logistic regression demonstrated the best performance, with an AUROC of 0.847 and net reclassification improvement of 0.071, compared with the adjusted scoring system in the test set. The AUROC of the adjusted scoring system was 0.786, and the AUROCs of the previous scoring systems were 0.757 and 0.687. The elastic-net logistic regression was transformed into an accessible online model. The performance of the machine learning model, which provides a probability estimation for rotator cuff reparability, is comparable with that of the adjusted scoring system. Nevertheless, when deploying prediction models beyond the original cohort, regardless of whether they rely on machine learning or scoring systems, clinicians should exercise caution and not rely solely on the output of the model.

Stacking Ensemble Technique Using Optimized Machine Learning Models with Boruta–XGBoost Feature Selection for Landslide Susceptibility Mapping: A Case of Kermanshah Province, Iran

Landslides cause significant human and financial losses in different regions of the world. A high-accuracy landslide susceptibility map (LSM) is required to reduce the adverse effects of landslides. Machine learning (ML) is a robust tool for LSM creation. ML models require large amounts of data to predict landslides accurately. This study has developed a stacking ensemble technique based on ML and optimization to enhance the accuracy of an LSM while considering small datasets. The Boruta–XGBoost feature selection was used to determine the optimal combination of features. Then, an intelligent and accurate analysis was performed to prepare the LSM using a dynamic and hybrid approach based on the Adaptive Fuzzy Inference System (ANFIS), Extreme Learning Machine (ELM), Support Vector Regression (SVR), and new optimization algorithms (Ladybug Beetle Optimization [LBO] and Electric Eel Foraging Optimization [EEFO]). After model optimization, a stacking ensemble learning technique was used to weight the models and combine the model outputs to increase the accuracy and reliability of the LSM. The weight combinations of the models were optimized using LBO and EEFO. The Root Mean Square Error (RMSE) and Area Under the Receiver Operating Characteristic Curve (AUC-ROC) parameters were used to assess the performance of these models. A landslide dataset from Kermanshah province, Iran, and 17 influencing factors were used to evaluate the proposed approach. Landslide inventory was 116 points, and the combined Voronoi and entropy method was applied for non-landslide point sampling. The results showed higher accuracy from the stacking ensemble technique with EEFO and LBO algorithms with AUC-ROC values of 94.81% and 94.84% and RMSE values of 0.3146 and 0.3142, respectively. The proposed approach can help managers and planners prepare accurate and reliable LSMs and, as a result, reduce the human and financial losses associated with landslide events.

Associations of Lipid Profiles With the Onset of HEV-Related Acute Liver Failure: A Multicenter Cohort Study.

Hepatitis E virus (HEV) is one of the major etiologies for acute liver failure. This multicenter retrospective cohort study aimed to investigate the associations of lipid profiles with the risk of HEV-related acute liver failure (HEV-ALF) among hospitalized patients with acute hepatitis E. A total of 1061 participants were obtained from three tertiary medical centers in Jiangsu, China, between February 2018 and May 2024. Univariate and multivariate Cox regression models were constructed to assess the associations between lipid profiles and the risk of HEV-ALF onset. The time-dependent area under the receiver-operating-characteristic curve (AUROC) and decision curve analysis were used to further evaluate the predictive value of blood lipids. After adjusting for potential confounders, total cholesterol (HR = 0.535, 95% CI: 0.437-0.656, p < 0.001), high-density lipoprotein-cholesterol (HR = 0.065, 95% CI: 0.027-0.154, p < 0.001), low-density lipoprotein-cholesterol (HR = 0.653, 95% CI: 0.512-0.833, p = 0.001), and apolipoprotein A (ApoA) (HR = 0.006, 95% CI: 0.002-0.020, p < 0.001) were significantly associated with a reduced risk of HEV-ALF. Moreover, blood ApoA exhibited excellent discrimination ability and net benefit for predicting 7-day (AUROC = 82.47%, 95% CI: 77.92-87.02) and 14-day (AUROC = 78.81%, 95% CI: 74.13-83.49) HEV-ALF onset. The findings may provide further evidence on the progression of HEV infection and future risk prediction.

Value of 18F-FDG PET/CT in the Diagnosis and Grading of Incidental Colorectal Adenomas

PurposeColorectal adenomas (CRAs) are at a higher risk of progressing to colorectal cancer (CRC) as their histological grade increases. Herein, this study investigated the relationship between the maximum standardized uptake value (SUVmax) on 18F-fluorodeoxyglucose positron emission tomography-computed tomography (18F-FDG PET/CT) and the histological grades of CRAs and constructed the optimal regression model for distinguishing between different histological grades. MethodsThis study retrospectively analyzed the data of 153 patients with CRAs who had colorectal 18F-FDG uptake incidentally found on PET/CT. The patients were categorized into low-grade intraepithelial neoplasia (LGIN) and high-grade intraepithelial neoplasia (HGIN) groups based on their histological grade. After the analysis of the relationship between SUVmax measured on preoperative 18F-FDG PET/CT scans and histological grades, receiver-operating characteristic (ROC) curves were analyzed to determine the optimal cut-off values for distinguishing between the two groups. Common clinical and pathological factors were included and subjected to univariate and multivariate logistic regression analyses to identify independent risk factors. A diagnostic model integrating SUVmax and several risk factors was developed with the multivariate logistic regression analysis. ResultsSUVmax was significantly different between the two groups (P < 0.001) and increased with an elevation in the malignancy degree. The area under the ROC curve (AUC) for identifying LGIN and HGIN was 0.796, and the AUC of the combination model was 0.822. Furthermore, SUVmax was an independent risk factor for distinguishing between different histological grades in pairwise comparisons. ConclusionThe regression model involving SUVmax on 18F-FDG PET/CT can distinguish between histological grades of CRAs, which therefore can be used as a noninvasive tool for the accurate diagnosis of CRAs and assist in developing patient-specific treatment strategies before surgery.

Integrative hybrid deep learning for enhanced breast cancer diagnosis: leveraging the Wisconsin Breast Cancer Database and the CBIS-DDSM dataset

The objective of this investigation was to improve the diagnosis of breast cancer by combining two significant datasets: the Wisconsin Breast Cancer Database and the DDSM Curated Breast Imaging Subset (CBIS-DDSM). The Wisconsin Breast Cancer Database provides a detailed examination of the characteristics of cell nuclei, including radius, texture, and concavity, for 569 patients, of which 212 had malignant tumors. In addition, the CBIS-DDSM dataset—a revised variant of the Digital Database for Screening Mammography (DDSM)—offers a standardized collection of 2,620 scanned film mammography studies, including cases that are normal, benign, or malignant and that include verified pathology data. To identify complex patterns and trait diagnoses of breast cancer, this investigation used a hybrid deep learning methodology that combines Convolutional Neural Networks (CNNs) with the stochastic gradients method. The Wisconsin Breast Cancer Database is used for CNN training, while the CBIS-DDSM dataset is used for fine-tuning to maximize adaptability across a variety of mammography investigations. Data integration, feature extraction, model development, and thorough performance evaluation are the main objectives. The diagnostic effectiveness of the algorithm was evaluated by the area under the Receiver Operating Characteristic Curve (AUC-ROC), sensitivity, specificity, and accuracy. The generalizability of the model will be validated by independent validation on additional datasets. This research provides an accurate, comprehensible, and therapeutically applicable breast cancer detection method that will advance the field. These predicted results might greatly increase early diagnosis, which could promote improvements in breast cancer research and eventually lead to improved patient outcomes.

Prediction of strain level phage-host interactions across the Escherichia genus using only genomic information.

Predicting bacteriophage infection of specific bacterial strains promises advancements in phage therapy and microbial ecology. Whether the dynamics of well-established phage-host model systems generalize to the wide diversity of microbes is currently unknown. Here we show that we could accurately predict the outcomes of phage-bacteria interactions at the strain level in natural isolates from the genus Escherichia using only genomic data (area under the receiver operating characteristic curve (AUROC) of 86%). We experimentally established a dataset of interactions between 403 diverse Escherichia strains and 96 phages. Most interactions are explained by adsorption factors as opposed to antiphage systems which play a marginal role. We trained predictive algorithms and pinpoint poorly predicted interactions to direct future research efforts. Finally, we established a pipeline to recommend tailored phage cocktails, demonstrating efficiency on 100 pathogenic E. coli isolates. This work provides quantitative insights into phage-host specificity and supports the use of predictive algorithms in phage therapy.