Score C-index Research Articles

Multi-metric comparison of machine learning imputation methods with application to breast cancer survival

Handling missing data in clinical prognostic studies is an essential yet challenging task. This study aimed to provide a comprehensive assessment of the effectiveness and reliability of different machine learning (ML) imputation methods across various analytical perspectives. Specifically, it focused on three distinct classes of performance metrics used to evaluate ML imputation methods: post-imputation bias of regression estimates, post-imputation predictive accuracy, and substantive model-free metrics. As an illustration, we applied data from a real-world breast cancer survival study. This comprehensive approach aimed to provide a thorough assessment of the effectiveness and reliability of ML imputation methods across various analytical perspectives. A simulated dataset with 30% Missing At Random (MAR) values was used. A number of single imputation (SI) methods - specifically KNN, missMDA, CART, missForest, missRanger, missCforest - and multiple imputation (MI) methods - specifically miceCART and miceRF - were evaluated. The performance metrics used were Gower’s distance, estimation bias, empirical standard error, coverage rate, length of confidence interval, predictive accuracy, proportion of falsely classified (PFC), normalized root mean squared error (NRMSE), AUC, and C-index scores. The analysis revealed that in terms of Gower’s distance, CART and missForest were the most accurate, while missMDA and CART excelled for binary covariates; missForest and miceCART were superior for continuous covariates. When assessing bias and accuracy in regression estimates, miceCART and miceRF exhibited the least bias. Overall, the various imputation methods demonstrated greater efficiency than complete-case analysis (CCA), with MICE methods providing optimal confidence interval coverage. In terms of predictive accuracy for Cox models, missMDA and missForest had superior AUC and C-index scores. Despite offering better predictive accuracy, the study found that SI methods introduced more bias into the regression coefficients compared to MI methods. This study underlines the importance of selecting appropriate imputation methods based on study goals and data types in time-to-event research. The varying effectiveness of methods across the different performance metrics studied highlights the value of using advanced machine learning algorithms within a multiple imputation framework to enhance research integrity and the robustness of findings.

Evaluating ChatGPT's moral competence in health care-related ethical problems.

Artificial intelligence tools such as Chat Generative Pre-trained Transformer (ChatGPT) have been used for many health care-related applications; however, there is a lack of research on their capabilities for evaluating morally and/or ethically complex medical decisions. The objective of this study was to assess the moral competence of ChatGPT. This cross-sectional study was performed between May 2023 and July 2023 using scenarios from the Moral Competence Test (MCT). Numerical responses were collected from ChatGPT 3.5 and 4.0 to assess individual and overall stage scores, including C-index and overall moral stage preference. Descriptive analysis and 2-sided Student's t-test were used for all continuous data. A total of 100 iterations of the MCT were performed and moral preference was found to be higher in the latter Kohlberg-derived arguments. ChatGPT 4.0 was found to have a higher overall moral stage preference (2.325 versus 1.755) when compared to ChatGPT 3.5. ChatGPT 4.0 was also found to have a statistically higher C-index score in comparison to ChatGPT 3.5 (29.03 ± 11.10 versus 19.32 ± 10.95, P =.0000275). ChatGPT 3.5 and 4.0 trended towards higher moral preference for the latter stages of Kohlberg's theory for both dilemmas with C-indices suggesting medium moral competence. However, both models showed moderate variation in C-index scores indicating inconsistency and further training is recommended. ChatGPT demonstrates medium moral competence and can evaluate arguments based on Kohlberg's theory of moral development. These findings suggest that future revisions of ChatGPT and other large language models could assist physicians in the decision-making process when encountering complex ethical scenarios.

Myelofibrosis prognostic models validated in an underserved minority population: A retrospective analysis.

e18529 Background: Contemporary prognostic models for myelofibrosis (MF) risk stratify patients using a heterogeneous composition of clinical and genetic factors. Little is known regarding how these scores perform in underserved populations. Our study assesses how well current models stratify risk at an inner-city tertiary care center which carries unique challenges including increased diversity, later disease presentation, and less access to healthcare resources. Methods: We retrospectively reviewed 88 MF patients referred to our center in the Bronx, NY between 2000 and 2023. Clinical and genetic data were used to calculate prognostic scoring for DIPSS, DIPSS plus, MIPSS70, MIPSSV2, and GIPSS. Overall survival (OS) was calculated from the time of diagnosis and non-deceased patients were censored at last follow-up. Patients with missing clinical or genetic data were excluded from individual analysis. Models were assessed via univariate and multivariate cox proportional hazards analysis followed by concordance analysis. Results: Mean age of the cohort was 66 years (range 29-88) with median survival of 7.7 years. Ethnicities included non-hispanic white (NHW) [30 (34.1%)], NH black [23 (26.1%)], NH asian [6 (6.8%)], and hispanics of all races [22 (25%)] (unknown n=7). DIPSS (n= 88, HR 1.77, CI 1.22 - 2.41, p-value = 0.0001), DIPSS plus (n=83, HR 1.54, 1.15-2.07, p = 0.0033), MIPSS70V2 (n=68, HR 2.06, CI 1.23 - 3.61, p = 0.0088), and GIPSS (n=69, HR 2.73, CI 1.24 - 3.78, p = 0.0063) were significantly associated with OS based on univariate analyses. Both DIPSS and GIPSS were significant even upon adjustment for each other (p=.009 and p=.043 respectively). MIPSS70 (n=73, HR 1.87, CI .93 - 3.88, p = 0.0842) did not meet significance. In terms of predictive accuracy based on C-index scores, DIPSS and GIPSS combined analysis scored highest (0.773), followed by GIPSS (.7226), DIPSS (0.719), MIPSSV2 (0.717), DIPSS Plus (0.6495), then MIPSS70 (0.6414). Conclusions: Our study shows that GIPSS, DIPSS, DIPSS plus, MIPSSV2 all function as valid prognostic models in the inner-city setting and that integrating both clinical and genetic models may have higher predictive value than either individual model alone. Additionally, GIPSS predicted superior survival outcomes to what was observed in our cohort for higher risk categories, prompting the need for further investigation. [Table: see text]

A machine learning predictive model for recurrence of resected distal cholangiocarcinoma: Development and validation of predictive model using artificial intelligence

IntroductionDistal Cholangiocarcinoma (dCCA) represents a challenge in hepatobiliary oncology, that requires nuanced post-resection prognostic modeling. Conventional staging criteria may oversimplify dCCA complexities, prompting the exploration of novel prognostic factors and methodologies, including machine learning algorithms. This study aims to develop a machine learning predictive model for recurrence after resected dCCA. Material and methodsThis retrospective multicentric observational study included patients with dCCA from 13 international centers who underwent curative pancreaticoduodenectomy (PD).A LASSO-regularized Cox regression model was used to feature selection, examine the path of the coefficient and create a model to predict recurrence. Internal and external validation and model performance were assessed using the C-index score. Additionally, a web application was developed to enhance the clinical use of the algorithm. ResultsAmong 654 patients, LNR (Lymph Node Ratio) 15, neural invasion, N stage, surgical radicality, and differentiation grade emerged as significant predictors of disease-free survival (DFS). The model showed the best discrimination capacity with a C-index value of 0.8 (CI 95 %, 0.77%–0.86 %) and highlighted LNR15 as the most influential factor. Internal and external validations showed the model's robustness and discriminative ability with an Area Under the Curve of 92.4 % (95 % CI, 88.2%–94.4 %) and 91.5 % (95 % CI, 88.4%–93.5 %), respectively. The predictive model is available at https://imim.shinyapps.io/LassoCholangioca/. ConclusionsThis study pioneers the integration of machine learning into prognostic modeling for dCCA, yielding a robust predictive model for DFS following PD. The tool can provide information to both patients and healthcare providers, enhancing tailored treatments and follow-up.

Application of the Second Revision of the International Staging System (R2-ISS) in the prognostic assessment of newly diagnosed multiple myeloma

Objective: To investigate the prognostic value of the Second Revision of the International Staging System (R2-ISS) in patients with newly diagnosed multiple myeloma (NDMM) . Methods: The retrospective study was performed in 326 NDMM patients with immunomodulatory drugs and/or proteasome inhibitors as the first-line treatment attending the Department of Hematology, Nanjing Drum Tower Hospital Clinical College of Nanjing Medical University, Nanjing, China, from December 2012 to March 2022. The Kaplan-Meier method was used for the survival analysis, with the Log-rank test comparing the between-group differences and Cox proportional risk regression modeling A multifactorial analysis was performed. Results: ①326 patients were included in the study, 190 of whom were males. The median age was 63 years, and the median followup time was 37 months. R2-ISS can effectively predict prognosis, particularly for R-ISS Ⅱ patients. The median progression-free survival (PFS) time of R2-ISS Ⅰ, R2-ISS Ⅱ, R2-ISS Ⅲ, and R2-ISS Ⅳ was 52, 29, 20, and 15 months (P<0.001), while the median overall survival (OS) time was 91, 60, 44, and 36 months (P<0.001). Multifactor analysis revealed that ISS Ⅱ, ISS Ⅲ, del (17p), t (4;14), 1q+, LDH increased, and age >65 years old were independent negative prognostic factors for OS. ISS Ⅱ, ISS Ⅲ, del (17p), t (4;14), 1q+, and LDH were independent negative prognostic factors for PFS. ②The C-index score of R2-ISS was 0.724, higher than that of R-ISS (0.678), indicating high prediction efficiency. ③The median PFS for 1q+-related double-hit in R2-ISS Ⅲ and Ⅳ were 20, 15 months (P=0.084) and the median OS were 35, 36 months (P=0.786), respectively. In R2-ISS Ⅲ, there were twenty-seven cases of 1q+-related double-hit, sixty-one cases of 1q+ single abnormality, and sixty-eight cases with no 1q+. The median PFS for the three groups were 20, 18, and 21 months (P=0.974), while the median OS was 35, 47, and 56 months (P=0.042), respectively. Adjusting the assignment of 1q+ to 1, the median PFS and OS of different R2-ISS stages differed significantly after regrouping (P<0.001) . Conclusions: The prognostic stratification value of R2-ISS is higher than R-ISS, particularly in the highly heterogeneous R-ISS Ⅱ population. Adjusting the assignment of the 1q+-related double-hit can improve R2-ISS, which should be validated in future studies with multi-center and expanded cases.

A New Nephrometry Score for Predicting Positive Surgical Margin After Laparoscopic Partial Nephrectomy.

To introduce the KESKIN ratio as a novel predictor of positive surgical margin (PSM) after laparoscopic partial nephrectomy (PN) and to evaluate other clinical characteristics and nephrometry scores (including RENAL, PADUA, and C-index) for predicting PSM. We retrospectively analyzed 95 patients who underwent laparoscopic PN between June 2020 and April 2023. The KESKIN ratio was defined for all patients. The KESKIN ratio, tumor and patient-related paramaters, and nephrometry scores were analyzed to predict PSM. Positive surgical margin was found in 12 of 95 patients (12.6%). There was no statistical difference between the PSM and negative surgical margin (NSM) groups in RENAL, PADUA, and C-index scores. Only the KESKIN ratio was found to be a statistically significant predictor of PSM in both univariate and multivariate regression analysis (p = 0.007 and p = 0.043, respectively). Mean endophytic diameter and endophytic percentage were found to be statistically significant predictors of PSM in only univariate analysis (p = 0.005 and p = 0.01, respectively). The value of 0.5 was determined as the cut-off value for the KESKIN ratio. Values higher than 0.5 indicate an increase in PSM. The KESKIN ratio is a novel, easily measurable, and calculable image-based parameter that can be used to predict PSM after laparascopic PN. If externally validated in a larger patient population, the KESKIN ratio may be used in future versions of the current nephrometry scoring systems for predicting the PSM.

Evaluating the predictive significance of systemic immune-inflammatory index and tumor markers in lung cancer patients with bone metastases.

This study aims to develop a predictive model for identifying lung cancer patients at elevated risk for bone metastases, utilizing the Unified Immunoinflammatory Index and various tumor markers. This model is expected to facilitate timely and effective therapeutic interventions, especially in the context of the growing significance of immunotherapy for lung cancer treatment. A retrospective analysis was conducted on 324 lung cancer patients treated between January 2019 and January 2021. After meeting the inclusion criteria, 241 patients were selected, with 56 exhibiting bone metastases. The cohort was divided into a training group (169 patients) and a validation group (72 patients) at a 7:3 ratio. Lasso regression was employed to identify critical variables, followed by logistic regression to construct a Nomogram model for predicting bone metastases. The model's validity was ascertained through internal and external evaluations using the Concordance Index (C-index) and Receiver Operating Characteristic (ROC) curve. The study identified several factors influencing bone metastasis in lung cancer, such as the Systemic Immune-Inflammatory Index (SII), Carcinoembryonic Antigen (CEA), Neuron Specific Enolase (NSE), Cyfra21-1, and Neutrophil-to-Lymphocyte Ratio (NLR). These factors were incorporated into the Nomogram model, demonstrating high validation accuracy with C-index scores of 0.936 for internal and 0.924 for external validation. The research successfully developed an intuitive and accurate Nomogram prediction model utilizing clinical indicators to predict the risk of bone metastases in lung cancer patients. This tool can be instrumental in aiding clinicians in developing personalized treatment plans, thereby optimizing patient outcomes in lung cancer care.

Advancing Precise Syphilis Diagnosis: A Nontreponemal IgM Antibody-Based Model for Latent Syphilis Staging.

Accurate differentiation between early and late latent syphilis stages is pivotal for patient management and treatment strategies. Nontreponemal IgM antibodies have shown potential in discriminating latent syphilis staging by differentiating syphilis activity. This study aimed to develop a predictive nomogram model for latent syphilis staging based on nontreponemal IgM antibodies. We explored the correlation between nontreponemal IgM antibodies and latent syphilis staging and developed a nomogram model to predict latent syphilis staging based on 352 latent syphilis patients. Model performance was assessed using AUC, calibration curve, Hosmer-Lemeshow χ2 statistics, C-index, Brier score, decision curve analysis, and clinical impact curve. Additionally, an external validation set was used to further assess the model's stability. Nontreponemal IgM antibodies correlated with latent syphilis staging. The constructed model demonstrated a strong discriminative capability with an AUC of 0.743. The calibration curve displayed a strong fit, key statistics including Hosmer-Lemeshow χ² at 2.440 (P=0.486), a C-index score of 0.743, and a Brier score of 0.054, all suggesting favorable model calibration performance. Decision curve analysis and clinical impact curve highlighted the model's robust clinical applicability. The external validation set yielded an AUC of 0.776, Hosmer-Lemeshow χ² statistics of 2.440 (P=0.486), a C-index score of 0.767, and a Brier score of 0.054, further underscored the reliability of the model. The nontreponemal IgM antibody-based predicted model could equip clinicians with a valuable tool for the precise staging of latent syphilis and enhancing clinical decision-making.

Böbrek Tümör Skorlama Sistemleri Tümör Agresivitesini Tahmin Edebilir Mi?

Objective: The aim of this study is to investigate the relationship between R.E.N.A.L. nephrometry score (RNS), Padua score (PS), Centrality (C)-index and tumour aggressivity in T1 renal tumours and to question whether these scoring systems would provide information about the pathology of renal tumours to manage clinical judgement rather than the anatomy of tumour. Material and Methods: We evaluated 83 patients with stage 1 (T1N0M0) clear cell renal cell carcinoma (cRCC) according to preoperative radiological and pathological staging. Patients were divided according to pathological results of cRCC into two groups: Patients with Fuhrman grade 1 or 2 (FG1-2) (Non-aggresive group (NAG)) and patients with FG3-4 and/or TNM Stage 3 (Aggressive group (AG)). RNS, PS and C-index scores were calculated for each patient. Finally,the relationship between nephrometry scores and pathological aggressivity were compared. Results: The mean RNS was calculated as 7.3±2.4. Total RNS was significantly higher in AG (9.2±1.2) than in NAG (6±2.2) (p&lt;0.001). RNS was an independent predictor of pathological aggressive disease (p&lt;0.001). The cut off value of RNS at the highest area under curve was 8 (p&lt;0.001). The mean PS was calculated as 8.1±1.6. PS was also an independent predictor of pathological aggressive disease (p&lt;0.001). The cut off value of PS at the highest area under curve was 8 (p&lt;0.001). The mean C-index score of AG (1.4 ± 0.4) was significantly lower (p&lt;0.001) than NAG (2.7±2.0). C-index is significant in predicting pathological aggressiveness (p&lt;0.001). Conclusions: Our results suggested that higher RNS and PS scores, lower C-index scores were associated with tumour aggressivity of renal tumours.

Nomogram for predicting adhesive small bowel obstruction following emergency gastrointestinal surgery.

Postoperative adhesions are frequent and significant complications that typically arise following abdominal surgery. Currently, the existing evidence for predicting the risk of adhesive small bowel obstruction (ASBO) after emergency gastrointestinal surgery (EGS) remains inadequate. A reliable perioperative model that quantifies the risk of ASBO after EGS serves as a practical tool for guiding individually tailored surveillance. A consecutive series of 1296 patients who underwent EGS for radiologically confirmed bowel/visceral inflammation or perforation between 2012 and 2022 at a tertiary academic medical center were included in this study to establish a best-fit nomogram. The nomogram was externally validated by assessing discrimination and calibration using an independent cohort from a separate medical center. A total of 116 patients (8.9%) developed at least one episode of ASBO after EGS during a median follow-up duration of 26months. The results of multivariable logistic analysis indicated that male sex (P = 0.043), preoperative albumin level (P = 0.002), history of pelvic radiotherapy (P = 0.038), laparotomy (P = 0.044), and intensive care unit stay ≥ 72h (P = 0.047) were identified as independent risk factors for developing ASBO. By incorporating these predictors, the developed nomogram exhibited good accuracy in risk estimation, as evidenced by a guide-corrected C-index score of 0.852 (95% CI 0.667-0.920) in the external validation cohort. Decision curve analysis and clinical impact curve demonstrated a clinically effective predictive model. By incorporating the nomogram as a supplemental tool in perioperative management, it becomes possible to accurately assess the individual's likelihood of developing ASBOs. This quantification enables surgeons to implement appropriate preventive measures, ultimately leading to improved outcomes.

Utility of the revised FIGO2023 staging with molecular classification in endometrial cancer

ObjectivesMolecular classification was introduced in endometrial cancer staging following the transition of the International Federation of Gynecology and Obstetrics (FIGO) 2008 to FIGO2023. In the early stages, p53 abnormal endometrial carcinoma with myometrial involvement was upstaged to stage IICm, in addition to the downstaging of POLE mutation endometrial cancer to stage IAm. This study compared the goodness of fit and discriminatory ability of FIGO2008, FIGO2023 without molecular classification (FIGO2023), and FIGO2023 with molecular classification (FIGO2023m); no study has been externally validated to date. MethodsThe study included 265 patients who underwent initial surgery at the National Cancer Center Hospital between 1997 and 2019 and were pathologically diagnosed with endometrial cancer. The three classification systems were compared using Harrell's concordance index (C-index), Akaike information criterion (AIC), and time-dependent receiver operating characteristic (ROC) curves. A higher C-index score and a lower AIC value indicated a more accurate model. ResultsAmong the three classification systems, FIGO2023m had the lowest AIC value (FIGO2023m: 455.925; FIGO2023: 459.162; FIGO2008: 457.901), highest C-index (FIGO2023m: 0.768; FIGO2023: 0.743; FIGO2008: 0.740), and superior time-dependent ROC curves within 1 year after surgical resection. In the stage IIIC, patients with p53 abnormalities had considerably lower 5-year overall survival than those with a p53 wild-type pattern (24.3% vs. 83.7%, p = 0.0005). ConclusionsFIGO2023m had the best discriminatory ability compared with FIGO2008 and FIGO2023. Even in advanced stages, p53 status was a poor prognostic factor. When feasible, molecular subtypes can be added to the staging criteria to allow better prognostic prediction in all stages.

Multimodal multitask learning for predicting MCI to AD conversion using stacked polynomial attention network and adaptive exponential decay

Early identification and treatment of moderate cognitive impairment (MCI) can halt or postpone Alzheimer’s disease (AD) and preserve brain function. For prompt diagnosis and AD reversal, precise prediction in the early and late phases of MCI is essential. This research investigates multimodal framework-based multitask learning in the following situations: (1) Differentiating early mild cognitive impairment (eMCI) from late MCI and (2) predicting when an MCI patient would acquire AD. Clinical data and two radiomics features on three brain areas deduced from magnetic resonance imaging were investigated (MRI). We proposed an attention-based module, Stack Polynomial Attention Network (SPAN), to firmly encode clinical and radiomics data input characteristics for successful representation from a small dataset. To improve multimodal data learning, we computed a potent factor using adaptive exponential decay (AED). We used experiments from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) cohort study, which included 249 eMCI and 427 lMCI participants at baseline visits. The proposed multimodal strategy yielded the best c-index score in time prediction of MCI to AD conversion (0.85) and the best accuracy in MCI-stage categorization (83.19%). Moreover, our performance was equivalent to that of contemporary research.

Patterns and Influence of Lymph Nodal Metastases After Neoadjuvant Chemotherapy and Surgery for Thoracic Esophageal Squamous Cell Carcinoma.

The purpose of this retrospective study was to define the pattern of lymph nodal metastases in patients with esophageal squamous cell carcinoma (ESCC) after neoadjuvant chemotherapy (NCT) followed by esophagectomy and to evaluate its influenceonprognosis. A total of 398 patients with clinical stage T3N0 or T1-3N+ ESCC who underwent NCT and radical esophagectomy with two-field lymphadenectomy were included. The distribution and frequency of metastases were counted separately for each lymph node station. The ypN stage, number of positive lymph node stations and lymph node stations with a metastasis rate greater than 5% were analyzed by using univariate Cox regression, followed by separate multivariable Cox regression analyses after adjusting for various clinical factors. Lymph node metastases were most frequently observed in the right upper paratracheal (16.8%) and left gastric artery (13.1%) stations. Multivariable models controlling for clinical factors showed that ypN stage remained a significant independent predictor of survival (N1 vs. N0: hazard ratio [HR], 2.30, 95% confidence interval [CI] 1.38-3.83, P < 0.001; N2 vs. N0: HR, 3.76, 95% CI 2.21-6.38, P < 0.001; N3 vs. N0: HR, 7.14, 95% CI 3.78-13.48, P < 0.001). The model from the multivariable analysis with the highest c-index score, indicating superior discriminatory preference, was ypN stage (c-index, 0.72). The pattern and influence of lymph node metastases after NCT will provide guidance on the extent of lymphadenectomy. Common positive lymph node stations for thoracic ESCC after NCT include the paratracheal, subcarinal, paraesophageal, paracardial, and left gastric artery stations.

Epigenetic and Tumor Microenvironment for Prognosis of Patients with Gastric Cancer.

Epigenetics studies heritable or inheritable mechanisms that regulate gene expression rather than altering the DNA sequence. However, no research has investigated the link between TME-related genes (TRGs) and epigenetic-related genes (ERGs) in GC. A complete review of genomic data was performed to investigate the relationship between the epigenesis tumor microenvironment (TME) and machine learning algorithms in GC. Firstly, TME-related differential expression of genes (DEGs) performed non-negative matrix factorization (NMF) clustering analysis and determined two clusters (C1 and C2). Then, Kaplan-Meier curves for overall survival (OS) and progression-free survival (PFS) rates suggested that cluster C1 predicted a poorer prognosis. The Cox-LASSO regression analysis identified eight hub genes (SRMS, MET, OLFML2B, KIF24, CLDN9, RNF43, NETO2, and PRSS21) to build the TRG prognostic model and nine hub genes (TMPO, SLC25A15, SCRG1, ISL1, SOD3, GAD1, LOXL4, AKR1C2, and MAGEA3) to build the ERG prognostic model. Additionally, the signature's area under curve (AUC) values, survival rates, C-index scores, and mean squared error (RMS) curves were evaluated against those of previously published signatures, which revealed that the signature identified in this study performed comparably. Meanwhile, based on the IMvigor210 cohort, a statistically significant difference in OS between immunotherapy and risk scores was observed. It was followed by LASSO regression analysis which identified 17 key DEGs and a support vector machine (SVM) model identified 40 significant DEGs, and based on the Venn diagram, eight co-expression genes (ENPP6, VMP1, LY6E, SHISA6, TMEM158, SYT4, IL11, and KLK8) were discovered. The study identified some hub genes that could be useful in predicting prognosis and management in GC.

Deep Learning Approaches for Glioblastoma Prognosis in Resource-Limited Settings: A Study Using Basic Patient Demographic, Clinical, and Surgical Inputs

Glioblastoma (GBM) is the most common brain tumor in the United States, with an annual incidence rate of 3.21 per 100,000. It is the most aggressive type of diffuse glioma and has a median survival of months after treatment. This study aims to assess the accuracy of different novel deep learning models trained on a set of simple clinical, demographic, and surgical variables to assist in clinical practice, even in areas with constrained health care infrastructure. Our study included 37,095 patients with GBM from the SEER (Surveillance Epidemiology and End Results) database. All predictors were based on demographic, clinicopathologic, and treatment information of the cases. Our outcomes of interest were months of survival and vital status. Concordance index (C-index) and integrated Brier scores (IBS) were used to evaluate the performance of the models. The patient characteristics and the statistical analyses were consistent with the epidemiologic literature. The models C-index and IBS ranged from 0.6743 to 0.6918 and from 0.0934 to 0.1034, respectively. Probabilistic matrix factorization (0.6918), multitask logistic regression (0.6916), and logistic hazard (0.6916) had the highest C-index scores. The models with the lowest IBS were the probabilistic matrix factorization (0.0934), multitask logistic regression (0.0935), and logistic hazard (0.0936). These models had an accuracy (1-IBS) of 90.66%; 90.65%, and 90.64%, respectively. The deep learning algorithms were deployed on an interactive Web-based tool for practical use available via https://glioblastoma-survanalysis.herokuapp.com/. Novel deep learning algorithms can better predict GBM prognosis than do baseline methods and can lead to more personalized patient care regardless of extensive electronic health record availability.

PD08-05 COMPUTER-GENERATED C-INDEX NEPHROMETRY SCORES DEMONSTRATE SUPERIOR PREDICTIVE ACCURACY TO THOSE OF HUMAN-GENERATED SCORES IN PREDICTING ONCOLOGIC AND PERIOPERATIVE OUTCOMES

You have accessJournal of UrologyCME1 Apr 2023PD08-05 COMPUTER-GENERATED C-INDEX NEPHROMETRY SCORES DEMONSTRATE SUPERIOR PREDICTIVE ACCURACY TO THOSE OF HUMAN-GENERATED SCORES IN PREDICTING ONCOLOGIC AND PERIOPERATIVE OUTCOMES Andrew Wood, Nicholas Heller, Tarik Benidir, Nour Abdallah, Fabian Isensee, Resha Tejpaul, Chalairat Suk-Ouichai, Caleb Curry, Alex You, Erick Remer, Samuel Haywood, Robert Abouassaly, Steven Campbell, Nikolaos Papanikolopoulos, and Christopher Weight Andrew WoodAndrew Wood More articles by this author , Nicholas HellerNicholas Heller More articles by this author , Tarik BenidirTarik Benidir More articles by this author , Nour AbdallahNour Abdallah More articles by this author , Fabian IsenseeFabian Isensee More articles by this author , Resha TejpaulResha Tejpaul More articles by this author , Chalairat Suk-OuichaiChalairat Suk-Ouichai More articles by this author , Caleb CurryCaleb Curry More articles by this author , Alex YouAlex You More articles by this author , Erick RemerErick Remer More articles by this author , Samuel HaywoodSamuel Haywood More articles by this author , Robert AbouassalyRobert Abouassaly More articles by this author , Steven CampbellSteven Campbell More articles by this author , Nikolaos PapanikolopoulosNikolaos Papanikolopoulos More articles by this author , and Christopher WeightChristopher Weight More articles by this author View All Author Informationhttps://doi.org/10.1097/JU.0000000000003239.05AboutPDF ToolsAdd to favoritesDownload CitationsTrack CitationsPermissionsReprints ShareFacebookLinked InTwitterEmail Abstract INTRODUCTION AND OBJECTIVE: The Centrality index (C-index) score is a nephrometry scoring system that requires precise measurements and mathematical calculations of cross sectional imaging. Its implementation has been slowed by requiring time investment and interobserver variability. We sought to automate this score on preoperative computerized tomography (CT) scans by developing an artificial intelligence (AI) generated C-index score (AI C-Index). Subsequently, we aimed to evaluate its ability to predict meaningful oncologic and perioperative outcomes as compared to human-generated C-index scores (Human C-Index). METHODS: 300 patients with preoperative CT were identified from a cohort of 544 patients undergoing surgical extirpation for suspected renal cancer. A deep neural network approach was used to segment kidneys and tumors, and then programed to generate the measurements and calculate C-index score. Human C-index scores were calculated by medical personnel blinded to AI-scores. AI- and Human score agreement was assessed using linear regression and their predictive abilities for both oncologic and perioperative outcomes were assessed using logistic regression and compared with receiver operating characteristic (ROC) curve analyses. RESULTS: Of 300 tumors, 27% were high stage (HS), 37% high grade (HG), 63% underwent partial nephrectomy, and median tumor size was 4.2 cm. There was significant agreement between Human and AI C-Index on linear regression (R2=0.738, p<0.0001). Both AI and Human C-Index similarly predicted oncologic outcomes, with lower levels of either score associated with increased risk of malignant histology (H-score p=0.018, AI score p=0.014) HG disease (both p<0.0001), and HS disease (both p<0.0001). Lower levels of either score also predicted a radical nephrectomy rather than partial nephrectomy (both p<0.0001). AUC measurements (Figure 1) were similar but consistently superior for AI C-index. CONCLUSIONS: Automated AI C-index scores are comparable to Human C-index scores and predict a wide variety of meaningful patient-centered outcomes. Our results suggest that our AI C-index could be delivered automatically from a preoperative CT scan to a clinician and patient at the point of care to aid in decision making. Source of Funding: None © 2023 by American Urological Association Education and Research, Inc.FiguresReferencesRelatedDetails Volume 209Issue Supplement 4April 2023Page: e234 Advertisement Copyright & Permissions© 2023 by American Urological Association Education and Research, Inc.MetricsAuthor Information Andrew Wood More articles by this author Nicholas Heller More articles by this author Tarik Benidir More articles by this author Nour Abdallah More articles by this author Fabian Isensee More articles by this author Resha Tejpaul More articles by this author Chalairat Suk-Ouichai More articles by this author Caleb Curry More articles by this author Alex You More articles by this author Erick Remer More articles by this author Samuel Haywood More articles by this author Robert Abouassaly More articles by this author Steven Campbell More articles by this author Nikolaos Papanikolopoulos More articles by this author Christopher Weight More articles by this author Expand All Advertisement PDF downloadLoading ...

Recurrence After Operated Adhesive Small Bowel Obstruction: Development and Validation of a Predictive Model

ObjectiveRecurrence after operated adhesive small bowel obstruction (ASBO) is a threatening potentiality for patients and a therapeutic dilemma facing any surgeon. Yet, little is known about screening tools to identify patients at higher risk of recurrence. The present study aimed to determine the risk factors of recurrence in patients operated on for ASBO and to construct a reliable predictive model capable of quantifying the risk of recurrence, which will be a practical tool to guide individualized patient management. MethodsWe developed a best-fit nomogram using data from a training cohort of 454 patients with ASBO treated surgically at the Affiliated Hospital of Qingdao University from 2012 through 2021. The nomogram was developed based on variables independently associated with recurrent ASBO screened via multivariate logistic regression. This model was validated using data from an independent internal cohort of 194 patients at the same institution and an external cohort of 196 patients at Qilu Hospital of Shandong University. Both internal and external validation was performed with a bootstrap resampling method (1000 iterations). The predictive accuracy of the nomogram was evaluated using Harrell’s concordance index (C index) and calibration curve. ResultsEight-one out of 648 (12.5%) patients experienced at least one ASBO relapse with a median follow-up of 37 (interquartile range, 18–70) months. Univariate and multivariate analyses identified the following independent predictive factors associated with recurrence: diabetes mellitus (P = 0.004), preoperative albumin levels (P = 0.002), omentectomy (P = 0.002), matted adhesions (P = 0.036), and the approach of surgery (P = 0.008). Incorporating these five predictors, our nomogram predicted recurrent ASBO with C-index scores of 0.932 (95% CI 0.867–0.996) in the training cohort, 0.874 (95% CI 0.706–1) in the internal validation cohort, and 0.852 (95% CI 0.667–0.920) in the external validation cohort. The predictive model showed a very good fitting degree. ConclusionsThe development of a practical, easy-to-use nomogram for calculating the risk of recurrence in patients with ASBO treated surgically will enable physicians to tailor therapeutic strategies and monitor disease in advance.

Predictive accuracy of computer-generated C-index nephrometry scores compared with human-generated scores in predicting oncologic and perioperative outcomes.

623 Background: The Centrality index (C-index) score is a validated nephrometry scoring system that requires precise measurements and mathematical calculations of cross sectional imaging. Like other nephrometry scores, its implementation has been slowed by required time investment and interobserver variability. We sought to automate this score on preoperative computerized tomography scans by developing an artificial intelligence-generated C-index score. We then aimed to evaluate its ability to predict meaningful oncologic and perioperative outcomes as compared to human-generated C-index nephrometry scores. Methods: 300 patients with preoperative computerized tomography with early arterial contrast phase were identified from a cohort of 544 consecutive patients undergoing surgical extirpation for suspected renal cancer. A deep neural network approach was used to automatically segment kidneys and tumors, and then programed to generate the measurements and calculate C-index score. Human C-index scores were independently calculated by medical personnel blinded to AI-scores. AI- and Human score agreement was assessed using bivariate linear regression correlation and their predictive abilities for both oncologic and perioperative outcomes were assessed using logistic regression and compared with receiver operating characteristic (ROC) curve analyses with measurements of areas under the curve (AUC). Results: Median age was 60 years (IQE 51–68), and 40% were female. Median tumor size was 4.2 cm and 91.3% had malignant tumors. 27% were high stage, 37% high grade, and 63% underwent partial nephrectomy. There was significant agreement between Human scores and AI-scores on linear regression analysis (R2 = 0.738, p &lt;0.0001). Both AI- and Human generated C-index scores similarly predicted meaningful oncologic outcomes, with lower levels of either C-index score associated with increased risk of malignant histology (H-score p = 0.018, AI score p =0.014) high-grade disease (both p &lt;0.0001), and high stage disease (both p &lt;0.0001). Lower levels of either AI or human generated C-index scores also predicted a radical nephrectomy rather than partial nephrectomy surgical approach (both p &lt;0.0001). AUC measurements (Table) were similar but consistently superior for AI generated C-index scores. Conclusions: Fully automated AI-generated C-index scores are comparable to human-generated C-index scores and predict a wide variety of meaningful patient-centered outcomes. Once validated in additional populations, our results suggest that our AI generated C-index could be delivered automatically from a preoperative CT scan to a clinician and patient at the point of care to aid in decision making. [Table: see text]

Immunogenic cell death-related long noncoding RNA influences immunotherapy against lung adenocarcinoma.

Lung adenocarcinoma (LUAD) is the leading cause of cancer-related deaths, accounting for over a million deaths worldwide annually. Immunogenic cell death (ICD) elicits an adaptive immune response. However, the role of ICD-related long noncoding RNAs (lncRNAs) in LUAD is unknown. In this study, we investigated the characteristics of the tumor microenvironment in LUAD, the prognostic significance of ICD-related lncRNAs, and the half-maximal inhibitory concentration (IC50) of possible chemotherapeutic drugs. We sorted prognostic lncRNAs using univariate Cox regression and constructed a risk signature based on them. We then confirmed the model's accuracy and generated a nomogram. Additionally, we performed immune microenvironment analysis, somatic mutation calculation, Tumor Immune Dysfunction and Exclusion (TIDE) analysis, and anticancer pharmaceutical IC50 prediction. Least absolute shrinkage and selection operator Cox regression identified 27 prognostic lncRNAs related to ICD, and a unique risk signature using 10 ICD-related lncRNAs was constructed. The risk score was confirmed to be a reliable predictor of survival, with the highest c-index score. The signature had a remarkable predictive performance with clinical applicability and could accurately predict the overall survival in LUAD. Furthermore, the lncRNA signature was closely associated with immunocyte invasion. We also analyzed the correlation between the risk score, tumor-infiltrating immune cells, and prognosis and identified high immune and ESTIMATE scores in low-risk patients. Moreover, we observed elevated checkpoint gene expression and low TIDE scores in high-risk patients, indicating a good immunotherapy response. Finally, high-risk patients were shown to be susceptible to anticancer medications. Therefore, our unique risk signature comprising 10 ICD-related lncRNAs was demonstrated to indicate the characteristics of the tumor-immune microenvironment in LUAD, predict patients' overall survival, and guide individualized treatment.

Histidine-rich glycoprotein as a novel predictive biomarker of postoperative complications in intensive care unit patients: a prospective observational study

BackgroundDecrease in histidine-rich glycoprotein (HRG) was reported as a cause of dysregulation of the coagulation-fibrinolysis and immune systems, leading to multi-organ failure, and it may be a biomarker for sepsis, ventilator-associated pneumonia, preeclampsia, and coronavirus disease 2019. However, the usefulness of HRG in perioperative management remains unclear. This study aimed to assess the usefulness of HRG as a biomarker for predicting postoperative complications.MethodsThis was a single-center, prospective, observational study of 150 adult patients who were admitted to the intensive care unit after surgery. Postoperative complications were defined as those having a grade II or higher in the Clavien–Dindo classification, occurring within 7 days after surgery. The primary outcome was HRG levels in the patients with and without postoperative complications. The secondary outcome was the ability of HRG, white blood cell, C-reactive protein, procalcitonin, and presepsin to predict postoperative complications. Data are presented as number and median (interquartile range).ResultsThe incidence of postoperative complications was 40%. The HRG levels on postoperative day 1 were significantly lower in patients who developed postoperative complications (n = 60; 21.50 [18.12–25.74] µg/mL) than in those who did not develop postoperative complications (n = 90; 25.46 [21.05–31.63] µg/mL). The Harrell C-index scores for postoperative complications were HRG, 0.65; white blood cell, 0.50; C-reactive protein, 0.59; procalcitonin, 0.73; and presepsin, 0.73. HRG was independent predictor of postoperative complications when adjusted for age, the presence of preoperative cardiovascular comorbidities, American Society of Anesthesiologists Physical Status Classification, operative time, and the volume of intraoperative bleeding (adjusted hazard ratio = 0.94; 95% confidence interval, 0.90–0.99).ConclusionsThe HRG levels on postoperative day 1 could predict postoperative complications. Hence, HRG may be a useful biomarker for predicting postoperative complications.