Discriminative Ability Research Articles

Risk factor analysis and optimal cutoff value selection of PSAD for diagnosing clinically significant prostate cancer in patients with negative mpMRI: results from a high-volume center in Southeast China

BackgroundMulti-parametric magnetic resonance imaging (mpMRI) is a diagnostic tool used for screening, localizing, and staging prostate cancer. Patients with Prostate Imaging Reporting and Data System (PI-RADS) score of 1 and 2 are considered negative mpMRI, with a lower likelihood of detecting clinically significant prostate cancer (csPCa). However, relying solely on mpMRI is insufficient to completely exclude csPCa, necessitating further stratification of csPCa patients using biomarkers.MethodsA retrospective study was conducted on mpMRI-negative patients who underwent prostate biopsy at the First Affiliated Hospital of Zhejiang University from January 2022 to June 2023. A total of 607 patients were included based on inclusion and exclusion criteria. Univariate and multivariate logistic regression analysis were performed to identify risk factors for diagnosing csPCa in patients with negative mpMRI. Receiver Operating Characteristic (ROC) curves were plotted to compare the discriminatory ability of different Prostate-Specific Antigen Density (PSAD) cutoff values for csPCa.ResultsAmong the 607 patients with negative mpMRI, 73 patients were diagnosed with csPCa. In univariate logistic regression analysis, age, PSA, f/tPSA, prostate volume, and PSAD were all associated with diagnosing csPCa in patients with negative mpMRI (P < 0.05), with PSAD being the most accurate predictor. In multivariate logistic regression analysis, f/tPSA, age, and PSAD were independent predictors of csPCa (P < 0.05). PSAD cutoff value of 0.20 ng/ml/ml has better discriminatory ability for predicting csPCa and is a significant risk factor for csPCa in multivariate analysis.ConclusionAge, f/tPSA, and PSAD are independent predictors of diagnosing csPCa in patients with negative mpMRI. It is suggested that patients with negative mpMRI and PSAD less than 0.20 ng/ml/ml could avoid prostate biopsy, as a PSAD cutoff value of 0.20 ng/ml/ml has better diagnostic performance than the traditional cutoff value of 0.15 ng/ml/ml.

Skin Cancer Segmentation using CNN

This groundbreaking research introduces a comprehensive strategy for advancing medical image segmentation, merging two pivotal concepts to significantly enhance both the accuracy and efficiency of the segmentation process. The first component of our approach involves the integration of polar transformations as a preprocessing step applied to the original dataset. This transformative technique is designed to address the challenges associated with segmenting single structures of elliptical shape in medical images, such as organs (e.g., heart and kidneys), skin lesions, polyps, and various abnormalities. By centering the polar transformation on the object's focal point, a reduction in dimensionality is achieved, coupled with a distinct separation of segmentation and localization tasks. Two distinct methodologies for selecting an optimal polar origin are proposed: one involving estimation through a segmentation neural network trained on non-polar images, and the other employing a dedicated neural network trained to pre dict the optimal origin. The second key element of our approach is around the integration of the DoubleU-Net architecture, a powerful encoder-decoder model specifically designed for the task of semantic image segmentation. DoubleU- Net is a group of two U-Net architectures, each with a specific purpose. The initial U-Net is pre-trained on VGG-19 as the encoder and uses features learned from ImageNet to provide efficient information transfer. In order to store more semantic information and content, a second U-Net was added to the base to enhance the capabilities of the network. Join Atrous Spatial Pyramid Pooling (ASPP) to develop network data extraction content. The combination of DoubleU-Net architecture and joint transformation as a step forward shows good segmentation performance in different clinical tasks, including liver segmentation, polyp detection vision, skin segmentation, and epicardial fat tissue segmentation. It shows that various medical projects, including various diagnostic methods such as colonoscopy, dermoscopy, microscopy, have a positive impact on the plan. More importantly, the method performs well in difficult cases, such as the segmentation of small and flat polyps in CVC-ClinicDB and the 2015 subset of the MICCAI Automated Polyp Detection dataset. The results demonstrate the accuracy and generality of the combination, making it the best way to evaluate medical images in context; Our study has revealed a new method of skin cancer diagnosis that combines the power of deep learning with innovation. Advanced technology. The combination of dual U-Net architecture and polar coordinate transformation not only improves the accuracy of classification of lesions but also improves the robustness of the model to changes in image features. This study contributes to the development of computer-aided diagnostic systems for early diagnosis. Experimental results show that our method provides good accuracy, sensitivity, and specificity in detecting malignant and benign tumors. Additionally, we are conducting ablation studies to determine the contribution of each presentation and treatment of skin cancer to ultimately benefit patients and determine these benefits. We also apply the transformation of the joint as the first step to improve the discrimination ability of the model. This mechanical change effectively reduces the impact caused by changes in wound size, shape, and direction by displaying the original image of the polar system. By standardizing the representation of skin diseases, polar transformation improves the model's ability to generalize across different data sets and improves overall performance.

Neural adaptations to congenital deafness: enhanced tactile discrimination through cross-modal neural plasticity - an fMRI study.

This study explores the compensatory neural mechanisms associated with congenital deafness through anexamination of tactile discrimination abilities using high-resolution functional magnetic resonance imaging (fMRI). To analyze the neural substrates underlying tactile processing in congenitally deaf individuals and compare themwith hearing controls. Our participant pool included thirty-five congenitally deaf individuals and thirty-five hearing controls. All participantsengaged in tactile discrimination tasks involving the identification of common objects by touch. We utilized ananalytical suite comprising voxel-based statistics, functional connectivity multivariate/voxel pattern analysis (fc-MVPA), and seed-based connectivity analysis to examine neural activity. Our findings revealed pronounced neural activity in congenitally deaf participants within regions typically associatedwith auditory processing, including the bilateral superior temporal gyrus, right middle temporal gyrus, and rightrolandic operculum. Additionally, unique activation and connectivity patterns were observed in the right insula andbilateral supramarginal gyrus, indicating a strategic reorganization of neural pathways for tactile informationprocessing. Behaviorally, both groups demonstrated high accuracy in the tactile tasks, exceeding 90%. However, thedeaf participants outperformed their hearing counterparts in reaction times, showcasing significantly enhancedefficiency in tactile information processing. These insights into the brain's adaptability to sensory loss through compensatory neural reorganization highlight theintricate mechanisms by which tactile discrimination is enhanced in the absence of auditory input. Understandingthese adaptations can help develop strategies to harness the brain's plasticity to improve sensory processing inindividuals with sensory impairments, ultimately enhancing their quality of life through improved tactile perceptionand sensory integration.

Deep learning nomogram for predicting neoadjuvant chemotherapy response in locally advanced gastric cancer patients.

Developed and validated a deep learning radiomics nomogram using multi-phase contrast-enhanced computed tomography (CECT) images to predict neoadjuvant chemotherapy (NAC) response in locally advanced gastric cancer (LAGC) patients. This multi-center study retrospectively included 322 patients diagnosed with gastric cancer from January 2013 to June 2023 at two hospitals. Handcrafted radiomics technique and the EfficientNet V2 neural network were applied to arterial, portal venous, and delayed phase CT images to extract two-dimensional handcrafted and deep learning features. A nomogram model was built by integrating the handcrafted signature, the deep learning signature, with clinical features. Discriminative ability was assessed using the receiver operating characteristics (ROC) curve and the precision-recall (P-R) curve. Model fitting was evaluated using calibration curves, and clinical utility was assessed through decision curve analysis (DCA). The nomogram exhibited excellent performance. The area under the ROC curve (AUC) was 0.848 [95% confidence interval (CI), 0.793-0.893)], 0.802 (95% CI 0.688-0.889), and 0.751 (95% CI 0.652-0.833) for the training, internal validation, and external validation sets, respectively. The AUCs of the P-R curves were 0.838 (95% CI 0.756-0.895), 0.541 (95% CI 0.329-0.740), and 0.556 (95% CI 0.376-0.722) for the corresponding sets. The nomogram outperformed the clinical model and handcrafted signature across all sets (all P < 0.05). The nomogram model demonstrated good calibration and provided greater net benefit within the relevant threshold range compared to other models. This study created a deep learning nomogram using CECT images and clinical data to predict NAC response in LAGC patients undergoing surgical resection, offering personalized treatment insights.

Discriminatory ability of perioperative heart rate variability in predicting postoperative complications in major urologic surgery: a prospective cohort study

We aimed to determine if continuous perioperative heart rate variability (HRV) monitoring could improve risk stratification compared to a short preoperative measurement in radical cystectomy patients. Electrocardiography (ECG) recordings were collected continuously preoperatively to discharge in 83 patients. Two, 5-min ECG signal segments (preoperative and at 24-h post ECG placement) were analyzed offline to extract HRV metrics. HRV metric discriminatory ability to identify patients with 30-day postoperative complications were analyzed using receiver operating characteristics curves. Sixty participants were included for analysis of which 27 (45%) developed a complication within 30 days postoperative. HRV was reduced in patients with complications. Postoperative standard deviation NN intervals and root mean square of successive differences had area under the curves (AUC) of 0.67 (95% CI 0.54 to 0.81) and 0.68 (95% CI 0.54 to 0.82), respectively. Significant discriminatory abilities were also reported for postoperative frequency metrics of absolute low frequency (LF) [AUC = 0.65 (95% CI 0.51 to 0.79)] and high frequency (HF) powers [AUC = 0.69 (95% CI 0.55 to 0.83)] and total power [AUC = 0.66 (95% CI 0.53 to 0.80)]. Postoperative acquired HRV metrics demonstrated improved discriminatory ability. Our findings suggest that longer-term perioperative HRV monitoring presents with superior ability to stratify complication risk.

Establishment of a Clinical Prediction Model Based on the Study of Risk Factors for Pain Recurrence after Percutaneous Radiofrequency Thermocoagulation in Patients with Primary Trigeminal Neuralgia

Objectives: This study aimed to investigate the follow-up outcomes and risk factors associated with pain recurrence after percutaneous radiofrequency thermocoagulation (PRT) among patients with primary trigeminal neuralgia (PTN) and to establish a clinical prediction model based on these risk factors. Methods: The data of PTN patients who underwent PRT were collected in our study. All subjects were randomly divided at a 7:3 ratio into a training group (T Group) and a test group (C Group) to select risk factors. According to the follow-up results, the patients were divided into a recurrence group (F Group) and a nonrecurrence group (NF Group). Predictive factors were selected through LASSO regression analysis based on T Group. The identified variables were subjected to multivariate logistic regression analysis to construct a nomogram. Receiver operating characteristic (ROC) curves and calibration curves were uesd to evaluate discrimination and calibration ability separately. Results: A total of 884 patients were initially included, 857 patients achieved satisfactory results, and the pain relief rate at discharge was 96.95%. A total of 529 subjects were included after screening, and the recurrence rate was 27.22% after 1 year. Six non-zero variables were selected through LASSO regression analysis: the disease course, atypical pain, previous surgery, facial numbness before PRT, neurovascular contact (NVC), and operation duration. Six variables were included in the multivariate logistic regression analysis, and the results showed that they were independent risk factors (P&lt;0.05). The predictive model is represented by a nomogram. The area under the curves (AUC) of the ROC curves were 0.868 (0.826~0.909) and 0.874 (0.802~0.950) for T Group and C Group, respectively. The prediction curves for T Group (P=0.784) and C Group (P=0.293) fit the ideal prediction curve, and the Brier scores were 0.120 and 0.099 for T Group and C Group, which indicates that the probability predicted by the model is consistent with the actual occurrence. Conclusion: The Pain relief rate at discharge was 96.95% for PTN patients after PRT, and the recurrence rate was 27.22% after 1 year. The disease course, atypical pain, previous surgery, facial numbness before PRT, NVC, and operation duration are independent risk factors for the recurrence of pain. A nomogram model for pain recurrence was established, which has good predictive ability.

Comparison of atherosclerotic cardiovascular disease (ASCVD) and Framingham risk scores (FRS) in an Iranian population

BackgroundFramingham risk score (FRS) and Atherosclerotic Cardiovascular Disease risk score (ASCVDrs) are widely used tools developed based on the American population. This study aimed to compare the ASCVDrs and FRS in an Iranian population. MethodThe participants of the Fasa Adult Cohort Study and the patients of the cardiovascular database of Vali-Asr Hospital of Fasa, aged 40–80 years, were involved in the present cross-sectional study. After excluding non-eligible participants, the individuals with a history of myocardial infarction or admission to the cardiology ward due to heart failure were considered high-risk, and the others were considered low-risk. The discriminative ability of FRS and ASCVDrs was evaluated and compared using receiver operating characteristic curve analysis. The correlation and agreement of ASCVDrs and FRS were tested using Cohen Kappa and Spearman. ResultsFinally, 8983 individuals (mean age:53.9 ± 9.5 y, 49.2 % male), including 1827 high-risk participants, entered the study. ASCVDrs detected a greater portion of participants as high-risk in comparison with FRS (28.7 % vs. 15.7 %). ASVD (AUC:0.794) had a higher discriminative ability than FRS (AUC:0.746), and both showed better discrimination in women. Optimal cut-off points for both ASCVDrs (4.36 %) and FRS (9.05 %) were lower than the original ones and in men. Compared to FRS, ASCVDrs had a higher sensitivity (79.3 % vs. 71.6 %) and lower specificity (64.5 % vs. 65.1 %). FRS and ASCVDrs had a moderate agreement (kappa:0.593,p-value<0.001) and were significantly correlated (Spearman:0.772,p-value<0.001). ConclusionsASCVDrs had a more accurate prediction of cardiovascular events and identified a larger number of people as high-risk in the Iranian population.

Retrospective Analysis of Risk Factors in Geriatric Hip Fracture Patients Predictive of Surgical Intensive Care Unit Admission.

Although numerous risk factors and prediction models affecting morbidity and mortality in geriatric hip fracture patients have been previously identified, there are scant published data on predictors for perioperative Surgical Intensive Care Unit (SICU) admission in this patient population. Determining if a patient will need an SICU admission would not only allow for the appropriate allocation of resources and personnel but also permit targeted clinical management of these patients with the goal of improving morbidity and mortality outcomes. The purpose of this study was to identify specific risk factors predictive of SICU admission in a population of geriatric hip fracture patients. Unlike previous studies which have investigated predominantly demographic, comorbidity, and laboratory data, the present study also considered a frailty index and length of time from injury to presentation in the Emergency Department (ED). A total of 501 geriatric hip fracture patients admitted to a Level 1 trauma center were included in this retrospective, single-center, quantitative study from January 1, 2019, to December 31, 2022. Using a logistical regression analysis, more than 25 different variables were included in the regression model to identify values predictive of SICU admission. Predictive models of planned versus unplanned SICU admissions were also estimated. The discriminative ability of variables in the final models to predict SICU admission was assessed with receiver operating characteristic curves' area under the curve estimates. Frailty,serum lactate > 2, andpresentation to the ED > 12 hours after injurywere significant predictors of SICU admission overall (P = 0.03, 0.038, and 0.05 respectively). Additionally, the predictive model for planned SICU admission had no common significant predictors with unplanned SICU admission. Planned SICU admission significant predictors included an Injury Severity Score (ISS) of 15 and greater, a higher total serum protein, serum sodium <135, systolic blood pressure (BP) under 100, increased heart rate on admission to ED, thrombocytopenia (<120), and higher Anesthesia Society Association physical status classification (ASA) score (P = 0.007, 0.04, 0.05, 0.002, 0.041, 0.05, and 0.005 respectively). Each SICU prediction model (overall, planned, and unplanned) demonstrated sufficient discriminative ability with the area under the curve (AUC) values of 0.869, 0.601, and 0.866 respectively.Finally, mean hospital Length of Stay (LOS) and mortality were increased in SICU admissions when compared to non-SICU admissions. Of the three risk factors predictive of SICU admission identified in this study, two have not been extensively studied previously in this patient population. Frailty has been associated with increased mortality and postoperative complications in hip fracture patients, but this is the first study to date to use a novel frailty index specifically designed and validated for use in hip fracture patients. The other risk factor, time from injury to presentation to the ED serves as an indicator for time a hip fracture patient spent without receiving medical attention. This risk factor has not been investigated heavily in the past as a predictor of SICU admissions in this patient population.

A nomogram model for the occurrence of bladder spasm after TURP in patients with prostate enlargement based on serum prostacyclin and 5-hydroxytryptamine and clinical characteristics.

With the development of analytical methods, mathematical models based on humoral biomarkers have become more widely used in the medical field. This study aims to investigate the risk factors associated with the occurrence of bladder spasm after transurethral resection of the prostate (TURP) in patients with prostate enlargement, and then construct a nomogram model. Two hundred and forty-two patients with prostate enlargement who underwent TURP were included. Patients were divided into Spasm group (n=65) and non-spasm group (n=177) according to whether they had bladder spasm after surgery. Serum prostacyclin (PGI2) and 5-hydroxytryptamine (5-HT) levels were measured by enzyme-linked immunoassay. Univariate and multivariate logistic regression were used to analyze the risk factors. Postoperative serum PGI2 and 5-HT levels were higher in patients in the Spasm group compared with the Non-spasm group (P<0.05). Preoperative anxiety, drainage tube obstruction, and elevated postoperative levels of PGI2 and 5-HT were independent risk factors for bladder spasm after TURP (P<0.05). The C-index of the model was 0.978 (0.959-0.997), with a χ2 = 4.438 (p = 0.816) for Hosmer-Lemeshow goodness-of-fit test. The ROC curve to assess the discrimination of the nomogram model showed an AUC of 0.978 (0.959-0.997). Preoperative anxiety, drainage tube obstruction, and elevated postoperative serum PGI2 and 5-HT levels are independent risk factors for bladder spasm after TURP. The nomogram model based on the aforementioned independent risk factors had good discrimination and predictive abilities, which may provide a high guidance value for predicting the occurrence of bladder spasm in clinical practice.

Prediction of atrial fibrillation in patients with stroke: a systematic review with meta-analysis

Abstract Background Atrial fibrillation-related strokes have a high rate of recurrence,1 and are associated with substantial morbidity, long-term disability, and mortality.2,3 Detection of AF at the time of stroke may require extended cardiac rhythm monitoring,4-6 which is costly.7 Identifying patients with stroke at high risk for AF can be challenging but important for preventing recurrent strokes. Purpose To provide an overview of prediction tools for AF in patients with an acute stroke, and to determine suitability for clinical use by synthesising discriminatory abilities. Methods We performed a systematic review of prediction tools (including multivariable scores and individual parameters) derived, validated and/or augmented for AF prediction in patients with strokes using Medline and Embase through 25 October 2023. Discrimination measures for tools with c-statistic data from ≥3 cohorts were pooled by Bayesian meta-analysis, with heterogeneity assessed through a 95% prediction interval (PI). Summary discrimination values of &amp;lt;0.60, 0.60-0.70, 0.70-0.80, and &amp;gt;0.80 were defined a priori as inadequate, adequate, good and excellent.1 Results From 8125 unique records we included 71 studies from 135 cohorts with a mean age and percentage of women ranging from 43.0 years to 77.5 years, and 25.2% to 55.3%, respectively. We identified 99 prediction tools for incident AF after a stroke. 33 predictive tools used clinical variables, 13 only used Echocardiographic data, 14 only used electrocardiogram (ECG) data, and 39 used combinations of the above modalities. 87 tools were developed with Cox or logistic regression and 12 tools through machine learning approaches. 36 had undergone external validation, with 27 demonstrating a c-statistic &amp;gt;0.75 on external validation. Nine multivariable prediction models and two predictive factors (B-type natriuretic peptide and age) were eligible for meta-analysis. Of multivariable prediction models, 4 showed good summary discrimination with wide heterogeneity (Figure 1): LADS (n = 1286, summary c-statistic 0.713, 95% CI 0.534-0.882 95% PI 0.379 -0.979), CHASE-LESS (n = 22747 summary c-statistic 0.726, 95% CI 0.667-0.785 95% PI 0.608-0.836), AS5F (n = 31517, summary c-statistic 0.726, 95% CI 0.693-0.754 95% PI 0.645-0.808), STAF (n = 3759, summary c-statistic 0.771, 95% CI 0.679-0.859 95% PI 0.456-0.973). BNP showed excellent summary discrimination with acceptable heterogeneity (Figure 2), (n = 1343 summary c-statistic 0.832, 95% CI 0.776-0.886 95% PI 0.718-0.927). Conclusions Five prediction tools externally validated for prediction of risk of AF in those who have had an acute stroke demonstrate good predictive ability. In meta-analysis, BNP performed the best out of all the models and variables tested.

Preoperative abnormal bone mineral density as a prognostic indicator in patients undergoing gastrectomy for gastric cancer: A cohort study.

Predicting postgastrectomy relapse and mortality in patients with gastric cancer (GC) remains challenging, with limitations to traditional staging systems such as the tumor-node-metastasis (TNM) system. This study aimed to investigate the impact of preoperative Hounsfield unit (HU) values, which serve as a surrogate marker for bone mineral density (BMD), in predicting survival outcomes in patients with GC. A retrospective analysis was conducted on data from patients with GC who underwent curative-intent gastrectomy. Opportunistic abdominopelvic computed tomography images were used to assess HU values at the 3rd lumbar vertebra (L3). These values were then categorized using a cutoff value of 110 HU, which has been established in previous studies as a determinant for abnormal versus normal BMD. Cox regression analysis established predictor models for overall survival (OS). Among 501 initial patients, 478 met the inclusion criteria. Multivariate analyses revealed HU values (hazard ratio, 1.51), along with other factors (the 5-factor modified frailty index, type of gastrectomy, TNM stage, anemia, and serum albumin level), as significant predictors of OS. The full model (FM) incorporating these variables demonstrated superior discrimination ability compared to the baseline model (BM), which is based solely on the TNM stage (concordance index: 0.807 vs 0.709; P < .001). Furthermore, the FM outperformed the BM in predicting OS risks at 36- and 60-months post-surgery. In conclusion, among patients undergoing gastrectomy for GC, those with HU values ≤ 110 (indicating abnormal BMD) at the L3 level, as determined through opportunistic CT scans, exhibited a poorer prognosis than those with HU values > 110 (indicating normal BMD). Integrating HU with other clinicopathological parameters enhances predictive accuracy, facilitating individualized risk stratification and treatment decision-making, which could potentially lead to improved survival outcomes.

Results of external validation for anatomical site recognition of ablation sites in atrial fibrillation using the CARTONET R12 model

Abstract Background CARTONET® is a cloud-based management system designed for retrospective analysis of ablation procedures using the CARTO system. Leveraging a proprietary machine learning model, it automates procedural information analysis. However, the efficacy and accuracy of the CARTONET R12 model's automated location identification remain unverified. Methods This study involved a total of 396 cases. Pulmonary vein isolation (PVI), including cavo-tricuspid isthmus (CTI) ablation, was performed in 313 patients (PVI group). PVI and additional ablation, such as a Box isolation and superior vena cava isolation (SVCI), was performed in 83 patients (PVI+ group). We compared the automated location identification of R12 model and the location identification of arrhythmia experts. Results In this study, 29,422 points were analyzed. The sensitivity of the pulmonary vein carina was low in both PVI group and PVI+ group. Excluding the inferior, anterior, and ridge segmentations, PVI group demonstrated significantly higher sensitivity compared to PVI + group. Regarding the ablation lines for box isolation, both the roof line and post line showed results below 30%. The pulmonary vein carina showed a low PPV in both PVI group and PVI+ group. In segmentations excluding the anterior, PVI group demonstrated significantly higher PPV compared to the group PVI+. Regarding the ablation lines for box isolation, both the roof line and post line showed results below 50%.The R12 model couldn't auto-identify CTI and SVC. Conclusion Excluding the carina, it can be said that in cases where PVI was performed, both PPV and sensitivity demonstrate high accuracy. Conversely, in cases where Box isolation was performed, accuracy decreases due to the low discrimination ability between ablation information near the roof, posterior wall, and inferior wall of the pulmonary veins and the ablation lines of Box isolation. Regarding the right atrium, it was found that the R12 model is entirely incapable of discriminating SVC and CTI, indicating a need for revisions in the current model.The segmentation of anatomyResults of the positive predict value

Identification of efficient inbred testers to strengthen established heterotic groups in pearl millet (Pennisetum glaucum [L.] R. Br)

AbstractPearl millet has two major heterotic groups, HGB (Heterotic Group B) and HGR (Heterotic Group R), each with unique genetic backgrounds and traits. Leveraging appropriate genetic resources is crucial for enhancing the potential of these heterotic groups in hybrid breeding programmes. Testers are essential in this context, determining the combining ability of new breeding lines and accurately assessing the performance and potential of parental lines in hybrid combinations. Two sets of test crosses were produced using line × tester mating design: Set 1 with six HGB testers crossed with new restorer lines and Set 2 with six HGR testers crossed with new seed parental lines, evaluated at multiple locations in India. Testers were evaluated for their effectiveness using multiple criteria, such as rank correlation analysis for line performance in hybrid combinations, discrimination ability, general combining ability, per se performance and Biplot analysis. These assessments determined their relative ranking, with HGB1‐T3, HGB2‐T3, HGR1‐T1 and HGR2‐T1 identified as efficient testers. The genetic analysis helps identify efficient testers for selecting superior genotypes, thereby enhancing heterotic group performance by incorporating appropriate new breeding material.

#1469 Predicting mortality risk in incident dialysis patients using clinical predictors and patient-reported predictors

Abstract Background and Aims Prognostic prediction models can be used to identify dialysis patients that are at high risk of dying. By doing so, vulnerable patients can be monitored more closely and treatment can be intensified where needed. In addition, individualized predictions can help inform patients about their prognosis. Traditionally, prognostic models primarily rely on known clinical predictors such as age, comorbidities and laboratory markers. Although it is still unchartered territory in the field of nephrology, studies in other fields have shown that patient-reported predictors, like pain and fatigue, can be of use as predictors of survival. Predictive performance of current prognostic models for mortality risk amongst dialysis patients is subpar and we hypothesized that the addition of patient-reported predictors may improve predictions. Therefore, we aimed to develop a novel prognostic model for the prediction of two-year mortality in patients starting dialysis using both clinical predictors and patient-reported predictors. Method Data from the Netherlands Cooperative Study on the Adequacy of Dialysis (NECOSAD) were used to develop two prognostic models for the prediction of 2-year all-cause mortality in incident dialysis patients: a base model consisting of clinical predictors and an updated model containing additional patient-reported predictors. Missing data were handled using a tenfold multiple imputation with fully conditional specification. The imputation model included all candidate predictors as well as the outcome variable of 2-year mortality. Predictors for the base model were selected based on existing mortality models and clinical expertise. Patient-reported predictors to add to the base model were chosen based on availability in NECOSAD, clinical expertise and literature describing patient-reported predictors. The models were developed using multivariate logistic regression and their discriminative abilities were assessed by calculating the area under the curve (AUC). Results A total of 1956 patients from NECOSAD were used for the analyses. Of these, 447 (22.9%) died within two years after baseline (3 months after dialysis initiation). The base model contained the following predictors: age, sex, serum albumin, C-reactive protein, diabetes mellitus, hemoglobin, residual kidney function (GFR at 3 months), body mass index, cardiovascular disease, dialysis modality, malignancy and primary kidney disease. The discriminative abilities of the base model were moderate with a AUC of 0.79 (95% CI: 0.77-0.82). The base model was updated using the following patient-reported predictors: mental component score, physical component score, general health perception, number of symptoms, symptom burden (measured with the KDQOL instrument) depressive symptoms, fatigue and pain. The addition of the patient-reported predictors marginally improved the predictive performance of the model achieving an AUC of 0.81 (95% CI: 0.79-0.84). Conclusion The base model including clinical predictors was enhanced slightly by the addition of patient-reported predictors.

New Health Indicator Construction and Fault Detection Network for Rolling Bearings via Convolutional Auto-Encoder and Contrast Learning

As one of the most important components in rotating machinery, if bearings fail, serious disasters may occur. Therefore, the remaining useful life (RUL) prediction of bearings is of great significance. Health indicator (HI) construction and early fault detection play a crucial role in data-driven RUL prediction. Unfortunately, most existing HI construction methods require prior knowledge and preset trends, making it difficult to reflect the actual degradation trend of bearings. And the existing early fault detection methods rely on massive historical data, yet manual annotation is time-consuming and laborious. To address the above issues, a novel deep convolutional auto-encoder (CAE) based on envelope spectral feature extraction is developed in this work. A sliding value window is defined in the envelope spectrum to obtain initial health indicators, which are used as preliminary labels for model training. Subsequently, CAE is trained by minimizing the composite loss function. The proposed construction method can reflect the actual degradation trend of bearings. Afterwards, the autoencoder is pre-trained through contrast learning (CL) to improve its discriminative ability. The model that has undergone offline pre-training is more sensitive to early faults. Finally, the HI construction method is combined with the early fault detection method to obtain a comprehensive network for online health assessment and fault detection, thus laying a solid foundation for subsequent RUL prediction. The superiority of the proposed method has been verified through experiments.

#1751 Validation of the ReDO Death Score in a Portuguese cohort

Abstract Background and Aims Mortality within the first year after dialysis start is estimated to be 20-30%, and it is mostly due to cardiovascular disease. Prognostic assessment after starting hemodialysis is challenging. Clark et al. developed the Recovery and Death Outcome risk score, which accurately predicted the likelihood of renal recovery to dialysis independence and of death within 1 year after discharge from in-hospital dialysis initiation. These models can be used at discharge or soon after patients start outpatient dialysis. We aimed to validate the Death Outcome risk score to predict one-year mortality after dialysis start in our population. Method .Retrospective analysis of hospitalized patients who initiated hemodialysis in a tertiary care hospital (Unidade Local de Saúde Santa Maria), from January 1st of 2016 to December 31st of 2019, and were discharged to outpatient dialysis. The risk of death within one year of discharge was calculated according to the ReDO score. We evaluated all-cause mortality within one year of hospital discharge. We classified patients into death outcome risk groups (labeled D1 to D4) according to the ReDO predictive score. Cox regression method was used to determine if the risk score was predictive of mortality within the first year after discharge. The discriminatory ability for the ReDO score to predict mortality was determined using the receiver operating characteristic (ROC) curve. Results 369 patients were included, with a mean age of 71.1 ± 14.3 years. The majority were male (59.9%, n = 221) and 87% were Caucasian (n = 321). The median Charlson score was 7 ± 3. The mortality rate within one year after discharge was 22.2% (n = 82). The one-year survival was significantly lower in patients with the highest probability of death (D4 = 61.3% vs. D3 = 67.6% vs. D2 = 81.3% vs. D4 = 97.3%, p &amp;lt; 0.001). The ReDO Death score accurately predicted the one-year risk of mortality, with an area under the ROC curve of 0.741, [95% CI (0.687–0.794), p &amp;lt; 0.001] (Fig. 1). The optimal REDO Death risk cut-off was &amp;gt;30%, with a hazard ratio of 6.57 [95% CI (3.48–12.2), p &amp;lt; 0.001] for one year risk of death, with a sensitivity of 78.0% and specificity of 60.6%. Conclusion We validated the ReDO score for 1-year mortality prediction after starting hemodialysis during hospitalization in a Portuguese population. This score can be used as a tool to inform goals of care discussion at the time of transition to out-of-hospital care, involving the in-hospital nephrology care team, the patient, and, if applicable, the future care team, as it can enlighten clinical decisions and, in some cases, lead to better end-of-life planning.

Comparison of atherogenic indices for predicting the risk of metabolic syndrome in Southwest Iran: results from the Hoveyzeh Cohort Study (HCS)

BackgroundMetabolic syndrome (MetS) is a cluster of risk factors related to diabetes and cardiovascular disease (CVD). Given that early identification of MetS might decrease CVD risk, it is imperative to establish a simple and cost-effective method to identify individuals at risk of MetS. The purpose of this study was to explore the relationships between several atherogenic indices (including AIP, TyG index, non-HDL-C, LDL-c/HDL-c, and TC/HDL-c) and MetS, and to assess the ability of these indices to predict MetS.MethodsThe present cross-sectional study was conducted using baseline data from 9809 participants of the Hoveyzeh Cohort Study (HCS). MetS was defined based on the International Diabetes Federation (IDF). To examine the discriminatory abilities of each atherogenic indices in the identification of MetS, a receiver-operating characteristic curve was conducted. Logistic regression analysis was also performed to evaluate the relationship between atherogenic indices and MetS.ResultsAll of the atherogenic indices including the TyG index, AIP, non-HDL-C, TC/HDL-c, and LDL-c/HDL-c were significantly higher in participants with MetS than in those without MetS. According to the ROC curve analysis, the TyG index revealed the highest area under the curve (0.79 and 0.85 in men and women, respectively), followed by the AIP (0.76 and 0.83 in men and women, respectively). The best cutoff values for the TyG index and AIP were 8.96 and 0.16 for men and 8.84 and 0.05 for women, respectively. The TyG index and AIP were also strongly associated with MetS.ConclusionAmong the 5 atherogenic indices evaluated, the TyG index and AIP were strongly related to MetS. The TyG index also demonstrated superior discriminative ability compared to other atherogenic indices in predicting MetS.

Incorporating low haemoglobin into a risk prediction model for conversion in minimally invasive gynaecologic oncology surgeries

Background A well-known complication of laparoscopic management of gynaecologic masses and cancers is the need to perform an intraoperative conversion to laparotomy. The purpose of this study was to identify novel patient risk factors for conversion from minimally invasive to open surgeries for gynaecologic oncology operations. Methods This was a retrospective cohort study of 1356 patients ≥18 years of age who underwent surgeries for gynaecologic masses or malignancies between February 2015 and May 2020 at a single academic medical centre. Multivariable logistic regression was used to study the effects of older age, higher body mass index (BMI), higher American Society of Anaesthesiologist (ASA) physical status, and lower preoperative haemoglobin (Hb) on odds of converting from minimally invasive to open surgery. Receiver operating characteristic (ROC) curve analysis assessed the discriminatory ability of a risk prediction model for conversion. Results A total of 704 planned minimally invasive surgeries were included with an overall conversion rate of 6.1% (43/704). Preoperative Hb was lowest for conversion cases, compared to minimally invasive and open cases (11.6 ± 1.9 vs 12.8 ± 1.5 vs 11.8 ± 1.9 g/dL, p<.001). Patients with preoperative Hb <10 g/dL had an adjusted odds ratio (OR) of 3.94 (CI: 1.65–9.41, p=.002) for conversion while patients with BMI ≥30 kg/m2 had an adjusted OR of 2.86 (CI: 1.50–5.46, p=.001) for conversion. ROC curve analysis using predictive variables of age >50 years, BMI ≥30 kg/m2, ASA physical status >2, and preoperative haemoglobin <10 g/dL resulted in an area under the ROC curve of 0.71. Patients with 2 or more risk factors were at highest risk of requiring an intraoperative conversion (12.0%). Conclusions Lower preoperative haemoglobin is a novel risk factor for conversion from minimally invasive to open gynaecologic oncology surgeries and stratifying patients based on conversion risk may be helpful for preoperative planning.

A nomogram for predicting survival in patients with gastrointestinal stromal tumor: a study based on the surveillance, epidemiology, and end results database

PurposeThe objective of this investigation was to construct and validate a nomogram for prognosticating cancer-specific survival (CSS) in patients afflicted with gastrointestinal stromal tumor (GIST) at 3-, 5-, and 8-years post-diagnosis.MethodsData pertaining to patients diagnosed with GIST were acquired from the Surveillance, Epidemiology, and End Results (SEER) database. Through random selection, a training cohort (70%) and a validation cohort (30%) were established from the patient population. Employing a backward stepwise Cox regression model, independent prognostic factors were identified. Subsequently, these factors were incorporated into the nomogram to forecast CSS rates at 3-, 5-, and 8-years following diagnosis. The nomogram’s performance was assessed using indicators such as the consistency index (C-index), the area under the time-dependent receiver operating characteristic curve (AUC), the net reclassification improvement (NRI), the integrated discrimination improvement (IDI), calibration curves, and decision-curve analysis (DCA).ResultsThis investigation encompassed a cohort of 3,062 GIST patients. By analyzing the Cox regression model within the training cohort, nine prognostic factors were identified: age, sex, race, marital status, AJCC (American Joint Committee on Cancer) stage, surgical status, chemotherapy status, radiation status, and income status. The nomogram was subsequently developed and subjected to both internal and external validation. The nomogram exhibited favorable discrimination abilities, as evidenced by notably high C-indices and AUC values. Calibration curves confirmed the nomogram’s reliability. Moreover, the nomogram outperformed the AJCC model, as demonstrated by enhanced NRI and IDI values. The DCA curves validated the clinical utility of the nomogram.ConclusionThe present study has successfully constructed and validated the initial nomogram for predicting prognosis in GIST patients. The nomogram’s performance and practicality suggest its potential utility in clinical settings. Nevertheless, further external validation is warranted.

#2586 Time to event prediction of new-onset diabetes after transplantation in kidney transplant recipients using machine learning models

Abstract Background and Aims New-onset diabetes after transplantation (NODAT) is a common and serious complication of kidney transplantation that contributes to long-term morbidity and mortality after the transplant procedure. Conventional methods may predict the risk of, but not time to developing NODAT. We aimed to develop a machine learning model that predicts the time to NODAT in kidney transplant recipients (KTR). Method Two large prospective cohort studies of stable KTR (inclusion &amp;gt;1 year after transplantation) from the University Medical Center Groningen were combined. Data collection took place between 2001 and 2003 for cohort A, and between 2008 and 2011 for cohort B, resulting in the inclusion of 750 KTR free of NODAT at baseline. To predict absolute time to NODAT, a bootstrap aggregating (bagging) approach was utilized which consisted of several extreme gradient boosting (XGBoost) models with the accelerated failure time (AFT) objective. The performance of the resulting bagging model was internally evaluated by assessing the concordance index (c-index) in 20% of the dataset not used for training. Results From the 750 KTR included, 76 (10.1%) developed NODAT during a median (interquartile range [IQR]) follow-up of 5.8 (4.7 to 9.2) years. The median (IQR) age was 51.2 (40.5 to 60.7) years and 342 (45.6%) of the KTR were female. During feature engineering, 18 features from 53 were selected. After hyperparameter tuning, model performance during internal validation was 0.84. Furthermore, SHapley Additive exPlanations (SHAP) values were used to estimate feature importance, which revealed that glucose, HbA1c, mean arterial pressure, cyclosporine concentrations and daily vegetable intake were the most important features. Conclusion The developed machine learning model shows good discriminative ability of time to NODAT prediction. Due to the methodology used, the model allows for flexible incremental learning, meaning that the bagging model can be retrained with additional data without compromising on performance.