Training Cohort Research Articles

A Prognostic Model Based on the Log Odds Ratio of Positive Lymph Nodes Predicts Prognosis of Patients with Rectal Cancer.

This study aimed to compare the prognostic value of rectal cancer by comparing different lymph node staging systems, and a nomogram was constructed based on superior lymph node staging. Overall, 8700 patients with rectal cancer was obtained from the Surveillance, Epidemiology, and End Results (SEER) database between 2010 and 2015. The area under the curve (AUC), the C index, and the Akaike informativeness criteria (AIC) were used to examine the predict ability of various lymph node staging methods. Prognostic indicators were assessed using univariate and multivariate COX regression, and further correlation nomograms were created after the data were randomly split into training and validation cohorts. To evaluate the effectiveness of the model, the C index, calibration curves, decision curves (DCA), and receiver operating characteristic curve (ROC) were used. We ran Kaplan-Meier survival analyses to look for variations in risk classification. While compared to the N-stage positive lymph node ratio (LNR), the log odds ratio of positive lymph nodes (LODDS) had the highest predictive effectiveness. Multifactorial COX regression analyses were used to create nomograms for overall survival (OS) and cancer-specific survival (CSS). The C indices of OS and CSS for this model were considerably higher than those for TNM staging in the training cohort. The created nomograms demonstrated good efficacy based on ROC, rectification, and decision curves. Kaplan-Meier survival analysis revealed notable variations in patient survival across various patient strata. Compared to AJCC staging, the LODDS-based nomograms have a more accurate predictive effectiveness in predicting OS and CSS in patients with rectal cancer.

Classification of multiple primary lung cancer in patients with multifocal lung cancer: assessment of a machine learning approach using multidimensional genomic data

BackgroundMultiple primary lung cancer (MPLC) is an increasingly well-known clinical phenomenon. However, its molecular characterizations are poorly understood, and still lacks of effective method to distinguish it from intrapulmonary metastasis (IM). Herein, we propose an identification model based on molecular multidimensional analysis in order to accurately optimize treatment.MethodsA total of 112 Chinese lung cancers harboring at least two tumors (n = 270) were enrolled. We retrospectively selected 74 patients with 121 tumor pairs and randomly divided the tumor pairs into a training cohort and a test cohort in a 7:3 ratio. A novel model was established in training cohort, optimized for MPLC identification using comprehensive genomic profiling analyzed by a broad panel with 808 cancer-related genes, and evaluated in the test cohort and a prospective validation cohort of 38 patients with 112 tumors.ResultsWe found differences in molecular characterizations between the two diseases and rigorously selected the characterizations to build an identification model. We evaluated the performance of the classifier using the test cohort data and observed an 89.5% percent agreement (PA) for MPLC and a 100.0% percent agreement for IM. The model showed an excellent area under the curve (AUC) of 0.947 and a 91.3% overall accuracy. Similarly, the assay achieved a considerable performance in the independent validation set with an AUC of 0.938 and an MPLC predictive value of 100%. More importantly, the MPLC predictive value of the classification achieved 100% in both the test set and validation cohort. Compared to our previous mutation-based method, the classifier showed better κ consistencies with clinical classification among all 112 patients (0.84 vs. 0.65, p <.01).ConclusionThese data provide novel evidence of MPLC-specific genomic characteristics and demonstrate that our one-step molecular classifier can accurately classify multifocal lung tumors as MPLC or IM, which suggested that broad panel NGS may be a useful tool for assisting with differential diagnoses.

PSMA-positive prostatic volume prediction with deep learning based on T2-weighted MRI.

High PSMA expression might be correlated with structural characteristics such as growth patterns on histopathology, not recognized by the human eye on MRI images. Deep structural image analysis might be able to detect such differences and therefore predict if a lesion would be PSMA positive. Therefore, we aimed to train a neural network based on PSMA PET/MRI scans to predict increased prostatic PSMA uptake based onthe axial T2-weighted sequence alone. All patients undergoing simultaneous PSMA PET/MRI for PCa staging or biopsy guidance between April 2016 and December 2020 at our institution were selected. To increase the specificity of our model, the prostatic beds on PSMA PET scans were dichotomized in positive and negative regions using an SUV threshold greater than 4 to generate a PSMA PET map. Then, a C-ENet was trained on the T2 images of the training cohort to generate a predictive prostatic PSMA PET map. One hundred and fifty-four PSMA PET/MRI scans were available (133 [68Ga]Ga-PSMA-11 and 21 [18F]PSMA-1007). Significant cancer was present in 127 of them. The whole dataset was divided into a training cohort (n = 124) and a test cohort (n = 30). The C-ENet was able to predict the PSMA PET map with a dice similarity coefficient of 69.5 ± 15.6%. Increased prostatic PSMA uptake on PET might be estimated based on T2 MRI alone. Further investigation with larger cohorts and external validation is needed to assess whether PSMA uptake can be predicted accurately enough to help in the interpretation of mpMRI.

Cholangiocarcinoma combined with biliary obstruction: an exosomal circRNA signature for diagnosis and early recurrence monitoring

Cholangiocarcinoma (CCA) is a highly malignant biliary tract cancer with currently suboptimal diagnostic and prognostic approaches. We present a novel system to monitor CCA using exosomal circular RNA (circRNA) via serum and biliary liquid biopsies. A pilot cohort consisting of patients with CCA-induced biliary obstruction (CCA-BO, n = 5) and benign biliary obstruction (BBO, n = 5) was used to identify CCA-derived exosomal circRNAs through microarray analysis. This was followed by a discovery cohort (n = 20) to further reveal a CCA-specific circRNA complex (hsa-circ-0000367, hsa-circ-0021647, and hsa-circ-0000288) in both bile and serum exosomes. In vitro and in vivo studies revealed the three circRNAs as promoters of CCA invasiveness. Diagnostic and prognostic models were established and verified by two independent cohorts (training cohort, n = 184; validation cohort, n = 105). An interpreter-free diagnostic model disclosed the diagnostic power of biliary exosomal circRNA signature (Bile-DS, AUROC = 0.947, RR = 6.05) and serum exosomal circRNA signature (Serum-DS, AUROC = 0.861, RR = 4.04) compared with conventional CA19-9 (AUROC = 0.759, RR = 2.08). A prognostic model of CCA undergoing curative-intent surgery was established by calculating early recurrence score, verified with bile samples (Bile-ERS, C-index=0.783) and serum samples (Serum-ERS, C-index = 0.782). These models, combined with other prognostic factors revealed by COX-PH model, enabled the establishment of nomograms for recurrence monitoring of CCA. Our study demonstrates that the exosomal triple-circRNA panel identified in both bile and serum samples serves as a novel diagnostic and prognostic tool for the clinical management of CCA.

Abstract PO5-24-10: Clinical significance and immune landscape analyses of the immune related genes based prognostic signature for HER2 positive breast cancer

Abstract Objective: To date, significant breakthrough has yet to be achieved in immunotherapy for HER2-positive breast cancer (HPBC). In this study, we aimed to construct a risk model for predicting prognosis and therapy response and evaluating the tumor immune landscape in HPBC patients. Methods: Differentially expressed immune-related genes (IRGs) were used as input to weighted correlation network analysis (WGCNA) to identify modules of highly correlated gene. An optimal prognostic TPBC-IRGs signature were constructed by least absolute shrinkage and selection operator (LASSO) cox regression. Survival analysis and ROC curves were analyzed to identify the predictive value in a training cohort and external validation cohorts. With three GEO datasets and samples from our center, the anti-HER2 therapeutic efficacy in HPBC patients was predicted through this prognostic model. Additionally, the immunotherapy outcome was forecasted using IMvigor210 dataset. We also investigated the connection between high and low-risk populations and immune checkpoints and analyzed immune infiltration levels of these populations with several algorithms. Ultimately, we develop a combined model and nomogram by integrating the risk score with clinical factors. Results: 11 key genes associated with overall survival (OS) in the blue module were selected for further analysis. A 3-gene signature (CX3CL1, RARB, TANK) based prognostic risk model was developed. The Kaplan-Meier survival curve revealed that patients with high-risk score had shorter OS compared with the low-risk group. The AUC of the training cohort at 1-, 3- and 5-year were 0.718, 0.784 and 0.767. The performance of the risk model was validated with GSE25066 and GSE55348. It’s showed that the risk scores of HPBC who were resistant to trastuzumab were significantly higher than those of patients who were sensitive to trastuzumab. The results indicated that the levels of immune checkpoints expression (such as PD-1, CTLA4 and TIGIT) and immune infiltration were higher in low-risk group than in high-risk group. The combined model showed better prognostic prediction accuracy compared to the clinical model or gene signature alone (1-year AUC=0.910;3-year AUC=0.832;5-year AUC=0.882). Conclusion: The immune gene-based risk profile and combined prognostic model in this study may hold promise for clinical use in predicting the prognosis and treatment efficacy of HPBC patients. figure3. Construction of HPBC-IRGs signature based prognostic model. Citation Format: Xiaofen Li, Wenfen Fu, Yushuai Yu, Jie Zhang, Chuangui Song. Clinical significance and immune landscape analyses of the immune related genes based prognostic signature for HER2 positive breast cancer [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO5-24-10.

Deciphering the role of apoptosis signature on the immune dynamics and therapeutic prognosis in breast cancer: Implication for immunotherapy

Background: In breast cancer oncogenesis, the precise role of cell apoptosis holds untapped potential for prognostic and therapeutic insights. Thus, it is important to develop a model predicated for breast cancer patients’ prognosis and immunotherapy response based on apoptosis-related signature.Methods: Our approach involved leveraging a training dataset from The Cancer Genome Atlas (TCGA) to construct an apoptosis-related gene prognostic model. The model’s validity was then tested across several cohorts, including METABRIC, Sun Yat-sen Memorial Hospital Sun Yat-sen University (SYSMH), and IMvigor210, to ensure its applicability and robustness across different patient demographics and treatment scenarios. Furthermore, we utilized Quantitative Polymerase Chain Reaction (qPCR) analysis to explore the expression patterns of these model genes in breast cancer cell lines compared to immortalized mammary epithelial cell lines, aiming to confirm their differential expression and underline their significance in the context of breast cancer.Results: Through the development and validation of our prognostic model based on seven apoptosis-related genes, we have demonstrated its substantial predictive power for the survival outcomes of breast cancer patients. The model effectively stratified patients into high and low-risk categories, with high-risk patients showing significantly poorer overall survival in the training cohort and across all validation cohorts. Importantly, qPCR analysis confirmed that the genes constituting our model indeed exhibit differential expression in breast cancer cell lines when contrasted with immortalized mammary epithelial cell lines.Conclusion: Our study establishes a groundbreaking prognostic model using apoptosis-related genes to enhance the precision of breast cancer prognosis and treatment, particularly in predicting immunotherapy response.

Nomograms for predicting the overall and cancer-specific survival in patients with non-urothelial carcinoma of the bladder: A population-based study.

Non-urothelial carcinoma of the bladder (NUCB) is a relatively rare condition, with limited comprehensive studies conducted to date. This research aims to establish nomograms for forecasting overall survival (OS) and cancer-specific survival (CSS) in NUCB patients. It utilizes data of 2,522 patients from the Surveillance, Epidemiology, and End Results (SEER) database spanning from 2004 to 2015. The effectiveness of these nomograms was assessed using receiver operating characteristic (ROC) curves, calibration curves and decision curve analysis (DCA). Key independent predictors for OS included age, race, marital status, histological variants, grade, T stage, N stage, M stage, radical cystectomy and chemotherapy administration. For CSS, the predictors were similar, encompassing age, sex, marital status, histological variants, grade, T stage, N stage, M stage, radical cystectomy and chemotherapy. The nomograms showed strong predictive accuracy. In the training cohort, the area under the curve (AUC) values were 0.796 (OS) and 0.799 (CSS) at 1-year, 0.807 (OS) and 0.824 (CSS) at 3-year, and 0.807 (OS) and 0.827 (CSS) at 5-year intervals. In the validation cohort, AUC values were 0.798 (OS) and 0.798 (CSS) at 1-year, 0.810 (OS) and 0.826 (CSS) at 3-year, and 0.811 (OS) and 0.825 (CSS) at 5-year intervals, consistently around 0.8. Calibration curves indicated high congruence between predicted and actual probabilities of OS and CSS, while DCA demonstrated the models' substantial clinical utility. Overall, this study successfully developed and validated prognostic nomograms for NUCB, capable of accurately predicting OS and CSS at 1-, 3-, and 5-years, thereby offering valuable support in clinical decision-making and the design of clinical trials.

Abstract PO1-07-07: Radiomics features for distinguishing true recurrence versus new primary tumor following breast-conserving treatment

Abstract Background Despite modern surgical and irradiation techniques, ipsilateral breast tumor recurrence (IBTR) accounts for 5-15% of all cancer recurrences in women treated with breast conservative treatment. Historically, the methods to define true recurrence (TR) and new primary tumor (NP) for IBTR mainly rely on clinical and pathological criteria, limiting the accuracy of the discerption and causing misclassification. This study aimed to develop a preoperative, noninvasive model for distinguishing IBTR by integrating clinicopathological review with paired dynamic contrast-enhanced breast magnetic resonance imaging (DCE-MRI) at diagnosis and following in-breast recurrence. MethodsWe retrospectively extracted radiomics features from MRI to develop a radiomic cohort of IBTR (n =46), among which all patients underwent paired preoperative DCE-MRI. This radiomic cohort was divided into a training cohort (n =27) and a validation cohort (n =19) with stratified random sampling. Classification of IBTR as TR or NP on the basis of tumor location, histologic subtype, estrogen receptor, and HER2 status was set as the gold standard. The least absolute shrinkage and selection operator (LASSO) regression and logistic regression were utilized to perform radiomics feature selection and model training, respectively. The clinical utility of the model was determined via decision curve analysis (DCA). Results We selected three radiomics features (first-order feature Kurtosis from the first postcontrast phase on IBTR DCE-MRI, GLCM feature Imc2 from the first postcontrast phase on IBTR DCE-MRI, and the delta feature Correlation from the precontrast phase of the change between IBTR and primary tumor DCE-MRI) to develop an IBTR-classification predicting radiomic model, which performed well in the validation cohort (AUC 0.867, 95% confidence interval (CI) 0.694-1). Further investigation for sensitivity (73.3%) and specificity (100%) verified a favorable concordance between the radiomic classification and the conventional standard, with a diagnostic accuracy of 79%. Conclusions: Our study demonstrated the feasibility of the radiomics model in predicting IBTR classification and provided preoperative information about the nature of “recurrence”. This might have important implications in surgical approaches and multidisciplinary care for IBTR. Further efforts are needed to improve the reproducibility of radiomics features and models in multiple centers. Citation Format: Feilin Qu, Guan-Hua Su, Jin-Hui Li, Chao You, Zhi-Ming Shao. Radiomics features for distinguishing true recurrence versus new primary tumor following breast-conserving treatment [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO1-07-07.

Abstract PO4-02-12: MDA iLobular Prognostic Tool: A Novel Approach for Risk Stratification in Invasive Lobular Carcinoma

Abstract Background: Invasive lobular carcinoma (ILC) is a distinct subtype of breast cancer that differs from the more common invasive ductal carcinoma (IDC) by its unique characteristics that influence prognosis. Recent studies have shown that ILC is associated with worse long-term outcomes compared to IDC. Risk stratification is essential in breast cancer for tailoring treatment plans to the individual patient’s needs, maximizing the chances of successful outcomes and minimizing unnecessary interventions. Although there are multiple tools available for breast cancer, there is currently not a risk stratification tool that is specific for ILC. The MDA iLobular prognostic tool is the first risk stratification tool that has been developed specifically for ILC patients. Methods: We retrospectively searched for patients treated at MD Anderson Cancer Center with a diagnosis of stage I-III ILC in our prospectively collected electronic database. The study focused on two primary endpoints: Overall Survival (OS) and Distant Recurrence Free Survival (DRFS). We utilized univariate and multivariate Cox Proportional Hazard (PH) regression models to assess the statistical significance of all variables. The univariate Cox analysis identified prognostic factors, which were further analyzed using backward and stepwise multivariate Cox proportional hazards regression analysis. This process helped identify statistically significant prognostic factors that were included in the final multivariate models. We estimated hazard ratios and 95% confidence intervals for each factor, considering a P-value of < 0.05 as statistically significant. To evaluate the performance of the fitted multivariate Cox PH regression models for OS and DRFS, we randomly divided two-thirds of the data points into a training dataset, while the remaining one-third constituted the test dataset. We assessed the discrimination capacity of each model using Harrell's C-index. Results: The study included a total of 4,216 female patients, with a median age of 56 years. The median pathological tumor size was 2 mm, and the median number of lymph nodes was one. Among these patients, 1,376 were pre-menopausal, while 2,837 were post-menopausal. The training cohort was a subset of 2,950 patients and the test cohort was a subset of 1,266 patients. After evaluating various prognostic models, we identified the model with the highest prognostic accuracy for OS and DRFS. This selected model demonstrated a Harrell's C-index of 0.704 for OS and 0.718 for DRFS on the training dataset and a Harrell's C-index of 0.702 for OS and 0.671 for DRFS on the test dataset. The model incorporated several covariates, including age at the time of diagnosis, number of lymph nodes, pathological tumor size (mm), ER status, tumor grade, ILC histology, and the presence or absence of concomitant LCIS (Table 1). Conclusion: In conclusion, the MDA iLobular prognostic tool represents a significant advancement as the first dedicated tool for risk stratifying ILC, providing valuable guidance for tailoring therapy to individual ILC patients based on their specific risk profiles. Table 1. Multivariate Cox proportional hazard model parameter estimates for OS and DRFS Citation Format: Jason Mouabbi, Sarah Pasyar, Roland Bassett, Akshara Singareeka Raghavendra, Amy Hassan, Rachel Layman, Debu Tripathy. MDA iLobular Prognostic Tool: A Novel Approach for Risk Stratification in Invasive Lobular Carcinoma [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO4-02-12.

Abstract PO5-15-10: Plasticity marker FOXC1 expression accurately predicts efficacy of Adjuvant Tamoxifen + Chemotherapy in reducing all-cause mortality in ER+LN- Breast Cancer: Validation in the SCAN-B Prospective Study (NCT02306096)

Abstract INTRODUCTION: Despite the development and validation of multimarker gene panels like OncotypeDx®, Mammaprint® and EndoPredict®, an affordable and globally accessible biomarker approach that can predict increased risk of all-cause mortality in patients diagnosed with ER+LN- breast cancer, as well as the efficacy of adjuvant endocrine + chemotherapy in mitigating such elevated risk, continues to be an unmet medical need. Lack of such a pragmatic solution continues to be the cause of preventable recurrent/metastatic disease in patients diagnosed with ER+LN- breast cancer in resource-challenged settings around the world. We hypothesized that a predictive biomarker strategy that measures expression of the plasticity marker FOXC1, in combination with clinical parameters like tumor size (TS) and tumor grade (TG), may offer equivalent results to that of the above multimarker gene panels at a significantly lower economic cost without compromising accuracy of prediction results. METHODS: Pre-treatment tumor RNA data obtained from a training cohort (compendium of gene expression microarray datasets, n=2857) and a single large validation cohort (SCAN-B Prospective Multicenter Observational Study, n=3520) for patients diagnosed with ER+LN- breast cancer were analyzed for FOXC1 expression, tumor size (TS) and tumor grade (TG) and correlated with Recurrence-Free Survival (RFS) and Overall Survival (OS). Optimized biomarker cutoff values based on model area-under-curve were leave-one-out cross validated and Risk-of-Recurrence (ROR) prediction algorithm derived utilizing the (compendium) training dataset. The unmodified strategy was then validated in the independent (SCAN-B) validation dataset. RESULTS: A predetermined High ROR score calculated using FOXC1 expression, TS and TG (trained using the compendium dataset) predicted efficacy of adjuvant endocrine + chemotherapy over that of adjuvant endocrine therapy alone in ER+LN- patients in terms of statistically significant reduction in all-cause mortality at 8-years post-diagnosis in the SCAN-B validation dataset (4.66% vs 24.06%, n=326, OR=6.48, 95%CI [2.97-14.11], p< 0.0001). OncotypeDx® (5.7% vs 14.4%, n=762, OR=2.79, 95%CI [1.37-5.67], p=0.002), Mammaprint® (5.1% vs 18.4%, n=665, OR= 95%CI [2.04-8.43], p< 0.0001) and Endopredict® (4.8% vs 15.2%, n=923, OR=3.58 95%CI [1.78-7.21], p=0.0002) were also statistically significant predictors of the same. CONCLUSION: Pre-treatment tumor FOXC1 mRNA or protein expression (assessed using qRT-PCR or routine immunohistochemistry (IHC), respectively, when combined with TS and TG presents a unique and economical alternative solution to multimarker gene panel tests like OncotypeDx®, Mammaprint® or Endopredict®, for guiding therapy of patients diagnosed with ER+LN- breast cancer in resource challenged settings. Such an approach to identify elevated risk of recurrence in patients diagnosed with ER+LN- breast cancer and prevent the same by guiding adjuvant endocrine + chemotherapy decisions, could help to extend recurrence-free and overall survival. Such an approach merits testing in real world ER+LN- patient cohorts in resource-challenged settings to help support implementation of this FOXC1-driven predictive biomarker strategy in the clinic. Citation Format: Partha Ray, Tania Ray, Clive Taylor, Robert Hussa. Plasticity marker FOXC1 expression accurately predicts efficacy of Adjuvant Tamoxifen + Chemotherapy in reducing all-cause mortality in ER+LN- Breast Cancer: Validation in the SCAN-B Prospective Study (NCT02306096) [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO5-15-10.

Developing and validating multi-omics prediction models for late patient-reported dysphagia in head and neck radiotherapy.

Background and purpose. To investigate models developed using radiomic and dosiomic (multi-omics) features from planning and treatment imaging for late patient-reported dysphagia in head and neck radiotherapy.Materials and methods. Training (n=64) and testing (n=23) cohorts of head and neck cancer patients treated with curative intent chemo-radiotherapy with a follow-up time greater than 12 months were retrospectively examined. Patients completed the MD Anderson Dysphagia Inventory and a composite score ≤60 was interpreted as patient-reported dysphagia. A chart review collected baseline dysphagia and clinical factors. Multi-omic features were extracted from planning and last synthetic CT images using the pharyngeal constrictor muscle contours as a region of interest. Late patient-reported dysphagia models were developed using a random forest backbone, with feature selection and up-sampling methods to account for the imbalanced data. Models were developed and validated for multi-omic feature combinations for both timepoints.Results. A clinical and radiomic feature model developed using the planning CT achieved good performance (validation: sensitivity=80±27% / balanced accuracy=71±23%, testing: sensitivity=80±10% / balanced accuracy=73±11%). The synthetic CT models did not show improvement over the plan CT multi-omics models, with poor reliability of the radiomic features on these images. Dosiomic features extracted from the synthetic CT showed promise in predicting late patient-reported dysphagia.Conclusion. Multi-omics models can predict late patient-reported dysphagia in head and neck radiotherapy patients. Synthetic CT dosiomic features show promise in developing successful models to account for changes in delivered dose distribution. Multi-center or prospective studies are required prior to clinical implementation of these models.

Leveraging artificial intelligence to identify high-risk patients for postoperative sore throat: An observational study.

Postoperative sore throat (POST) is a prevalent complication after general anesthesia and targeting high-risk patients helps in its prevention. This study developed and validated a machine learning model to predict POST. A total number of 834 patients who underwent general anesthesia with endotracheal intubation were included in this study. Data from a cohort of 685 patients was used for model development and validation, while a cohort of 149 patients served for external validation. The prediction performance of random forest (RF), neural network (NN), and extreme gradient boosting (XGBoost) models was compared using comprehensive performance metrics. The Local Interpretable Model-Agnostic Explanations (LIME) methods elucidated the best-performing model. POST incidences across training, validation, and testing cohorts were 41.7%, 38.4%, and 36.2%, respectively. Five predictors were age, sex, endotracheal tube cuff pressure, endotracheal tube insertion depth, and the time interval between extubation and the first drinking of water after extubation. After incorporating these variables, the NN model demonstrated superior generalization capabilities in predicting POST when compared to the XGBoost and RF models in external validation, achieving an area under the receiver operating characteristic curve (AUROC) of 0.81 (95% CI 0.74-0.89) and a precision-recall curve (AUPRC) of 0.77 (95% CI 0.66-0.86). The model also showed good calibration and clinical usage values. The NN model outperforms the XGBoost and RF models in predicting POST, with potential applications in the healthcare industry for reducing the incidence of this common postoperative complication.

A novel model of artificial intelligence based automated image analysis of CT urography to identify bladder cancer in patients investigated for macroscopic hematuria.

To evaluate whether artificial intelligence (AI) based automatic image analysis utilising convolutional neural networks (CNNs) can be used to evaluate computed tomography urography (CTU) for the presence of urinary bladder cancer (UBC) in patients with macroscopic hematuria. Our study included patients who had undergone evaluation for macroscopic hematuria. A CNN-based AI model was trained and validated on the CTUs included in the study on a dedicated research platform (Recomia.org). Sensitivity and specificity were calculated to assess the performance of the AI model. Cystoscopy findings were used as the reference method. The training cohort comprised a total of 530 patients. Following the optimisation process, we developed the last version of our AI model. Subsequently, we utilised the model in the validation cohort which included an additional 400 patients (including 239 patients with UBC). The AI model had a sensitivity of 0.83 (95% confidence intervals [CI], 0.76-0.89), specificity of 0.76 (95% CI 0.67-0.84), and a negative predictive value (NPV) of 0.97 (95% CI 0.95-0.98). The majority of tumours in the false negative group (n = 24) were solitary (67%) and smaller than 1 cm (50%), with the majority of patients having cTaG1-2 (71%). We developed and tested an AI model for automatic image analysis of CTUs to detect UBC in patients with macroscopic hematuria. This model showed promising results with a high detection rate and excessive NPV. Further developments could lead to a decreased need for invasive investigations and prioritising patients with serious tumours.

Predicting Osteoporosis and Osteopenia by Fusing Deep Transfer Learning Features and Classical Radiomics Features Based on Single-Source Dual-energy CT Imaging

Rationale and ObjectivesTo develop and validate a predictive model for osteoporosis and osteopenia prediction by fusing deep transfer learning (DTL) features and classical radiomics features based on single-source dual-energy computed tomography (CT) virtual monochromatic imaging. MethodsA total of 606 lumbar vertebrae with dual-energy CT imaging and quantitative CT (QCT) evaluation were included in the retrospective study and randomly divided into the training (n=424) and validation (n=182) cohorts. Radiomics features and DTL features were extracted from 70-keV monochromatic CT images, followed by feature selection and model construction, radiomics and DTL features models were established. Then, we integrated the selected two types of features into a features fusion model. We developed a two-level classifier for the hierarchical pairwise classification of each vertebra. All the vertebrae were first classified into osteoporosis and non-osteoporosis groups, then non-osteoporosis group was classified into osteopenia and normal groups. QCT was used as reference. The predictive performance and clinical usefulness of three models were evaluated and compared. ResultsThe area under the curve (AUC) of the features fusion, radiomics and DTL models for the classification between osteoporosis and non-osteoporosis were 0.981, 0.999, 0.997 in the training cohort and 0.979, 0.943, 0.848 in the validation cohort. Furthermore, the AUCs of the abovementioned models for the differentiation between osteopenia and normal were 0.994, 0.971, 0.996 in the training cohort and 0.990, 0.968, 0.908 in the validation cohort. The overall accuracy of the abovementioned models for two-level classifications was 0.979, 0.955, 0.908 in the training cohort and 0.918, 0.885, 0.841 in the validation cohort. Decision curve analysis (DCA) showed that all models had high clinical value. ConclusionThe feature fusion model can be used for osteoporosis and osteopenia prediction with improved predictive ability over a radiomics model or a DTL model alone.

Clinical study on forecasting the prognosis of patients with cerebellar hemorrhage based on CT radiomics models

Background and purposeWe created and verified a clinical–radiomics nomogram to forecast the prognosis of patients with cerebellar hemorrhage. Materials and MethodsEighty-eight patients with cerebellar hemorrhage from two different hospitals were retrospectively analyzed (62 cases from first hospital (training cohort) and 27 cases from second hospital (external validation cohort)). All patients underwent computed tomography scanning within 72 hours of the onset of cerebellar hemorrhage and were split into two categories based on their prognosis. Clinically independent risk factors were identified using the independent samples t-test and binary logistic regression analysis. Radiomic features were obtained using the SelectkBest, variance threshold, least absolute shrinkage, and selection operator. Radiomics scores (rad-scores) were produced from eight radiomics features of perihematomal edema (PHE) and hematoma associated with cerebellar hemorrhage prognosis. Logistic regression analyses of the rad-scores against clinical risk factors yielded clinical–radiomics column line plots. Predictive performance of the column line plots was validated using decision curve analysis analyses in both the training and external validation cohorts. ResultsTo construct rad-scores, eight radiomics traits and four clinical risk factors were combined to create clinical–radiomics models. In both the training cohort (area under the curve [AUC], 0.96 versus vs. 0.87/0.84) and the external validation cohort (AUC: 0.93 vs. 0.74/0.69), the clinical–radiomics nomogram achieved superior performance to the clinical and radiomics nomogram models. Clinical–radiomics column line plot demonstrated satisfactory calibration and clinical advantages in both cohorts. ConclusionsRadiomics features of PHE and hematoma can predict cerebellar hemorrhage prognosis. Clinical–radiomics models can help clinicians make individualized clinical treatment decisions and efficacy assessments for patients with cerebellar hemorrhage, help physicians prepare for possible deterioration, and promptly communicate treatment goals and efficacy expectations to patients and their families.

Radiomics Model and Deep Learning Model Based on T1WI Image for Acute Lymphoblastic Leukemia Identification

ObjectiveThis study aimed to develop highly precise radiomics and deep learning models to accurately detect acute lymphoblastic leukemia (ALL) using a T1WI image. MethodsA total of 604 brain magnetic resonance data of ALL group and normal children (NC) group. Two radiologists independently retrieved radiomics features after manually delineating the area of interest along the clivus at the median sagittal position of T1WI. According to the 9:1 ratio, all samples were randomly divided into the training cohort and the testing cohort. the support vector machine was then used to classify the radiomics model using the features that had a correlation coefficient of greater than 0.99 in the training cohort.the Efficientnet-B3 network model received the training set images to create a deep learning model. The sensitivity, specificity, and area under the ROC curve were calculated in order to evaluate the diagnostic efficacy of the different models after the validation of two aforementioned models in the testing cohort. ResultsThe deep learning model had a higher AUC value of 0.981 than the radiomics model's value of 0.962 in the testing cohort. Delong's test showed no statistical difference between the two models (P>0.05).The accuracy/sensitivity/specificity/negative predictive value/positive predictive value achieved 0.9180/0.9565/0.8947/0.9714/0.8462 for the radiomics model and 0.9344/0.8696/0.9737/0.9250/0.9524 for deep learning model. ConclusionThe deep learning and radiomics models showed high AUC values in the training and test cohorts. They also exhibited good diagnostic efficacy for predicting ALL.

Late-xerostomia prediction model based on 18F-FDG PET image biomarkers of the main salivary glands

Background and purposeRecently, a comprehensive xerostomia prediction model was published, based on baseline xerostomia, mean dose to parotid glands (PG) and submandibular glands (SMG). Previously, PET imaging biomarkers (IBMs) of PG were shown to improve xerostomia prediction. Therefore, this study aimed to explore the potential improvement of the additional PET-IBMs from both PG and SMG to the recent comprehensive xerostomia prediction model (i.e., the reference model). Materials and methodsTotally, 540 head and neck cancer patients were split into training and validation cohorts. PET-IBMs from the PG and SMG, were selected using bootstrapped forward selection based on the reference model. The IBMs from both the PG and SMG with the highest selection frequency were added to the reference model, resulting in a PG-IBM model and a SMG-IBM model which were combined into a composite model. Model performance was assessed using the area under the curve (AUC). Likelihood ratio test compared the predictive performance between the reference model and models including IBMs. ResultsThe final selected PET-IBMs were 90th percentile of the PG SUV and total energy of the SMG SUV. The additional two PET-IBMs in the composite model improved the predictive performance of the reference model significantly. The AUC of the reference model and the composite model were 0.67 and 0.69 in the training cohort, and 0.71 and 0.73 in the validation cohort, respectively. ConclusionThe composite model including two additional PET-IBMs from PG and SMG improved the predictive performance of the reference xerostomia model significantly, facilitating a more personalized prediction approach.

Deep Neural Network for the Prediction of KRAS, NRAS and BRAF Genotypes in left-sided colorectal cancer based on histopathologic images

BackgroundThe KRAS, NRAS, and BRAF genotypes are critical for selecting targeted therapies for patients with metastatic colorectal cancer (mCRC). Here, we aimed to develop a deep learning model that utilizes pathologic whole-slide images (WSIs) to accurately predict the status of KRAS, NRAS, and BRAFV600E. Methods129 patients with left-sided colon cancer and rectal cancer from the Third Affiliated Hospital of Sun Yat-sen University were assigned to the training and testing cohorts. Utilizing three convolutional neural networks (ResNet18, ResNet50, and Inception v3), we extracted 206 pathological features from H&E-stained WSIs, serving as the foundation for constructing specific pathological models. A clinical feature model was then developed, with carcinoembryonic antigen (CEA) identified through comprehensive multiple regression analysis as the key biomarker. Subsequently, these two models were combined to create a clinical-pathological integrated model, resulting in a total of three genetic prediction models. Result103 patients were evaluated in the training cohort (1,782,302 image tiles), while the remaining 26 patients were enrolled in the testing cohort (489,481 image tiles). Compared with the clinical model and the pathology model, the combined model which incorporated CEA levels and pathological signatures, showed increased predictive ability, with an area under the curve (AUC) of 0.96 in the training and an AUC of 0.83 in the testing cohort, accompanied by a high positive predictive value (PPV 0.92). ConclusionThe combined model demonstrated a considerable ability to accurately predict the status of KRAS, NRAS, and BRAFV600E in patients with left-sided colorectal cancer, with potential application to assist doctors in developing targeted treatment strategies for mCRC patients, and effectively identifying mutations and eliminating the need for confirmatory genetic testing.

Development and validation of a deep learning radiomics model with clinical-radiological characteristics for the identification of occult peritoneal metastases in patients with pancreatic ductal adenocarcinoma.

Occult peritoneal metastases (OPM) in patients with pancreatic ductal adenocarcinoma (PDAC) are frequently overlooked during imaging. The authors aimed to develop and validate a computed tomography (CT)-based deep learning-based radiomics (DLR) model to identify OPM in PDAC before treatment. This retrospective, bicentric study included 302 patients with PDAC (training: n =167, OPM-positive, n =22; internal test: n =72, OPM-positive, n =9: external test, n =63, OPM-positive, n =9) who had undergone baseline CT examinations between January 2012 and October 2022. Handcrafted radiomics (HCR) and DLR features of the tumor and HCR features of peritoneum were extracted from CT images. Mutual information and least absolute shrinkage and selection operator algorithms were used for feature selection. A combined model, which incorporated the selected clinical-radiological, HCR, and DLR features, was developed using a logistic regression classifier using data from the training cohort and validated in the test cohorts. Three clinical-radiological characteristics (carcinoembryonic antigen 19-9 and CT-based T and N stages), nine HCR features of the tumor, 14 DLR features of the tumor, and three HCR features of the peritoneum were retained after feature selection. The combined model yielded satisfactory predictive performance, with an area under the curve (AUC) of 0.853 (95% CI: 0.790-0.903), 0.845 (95% CI: 0.740-0.919), and 0.852 (95% CI: 0.740-0.929) in the training, internal test, and external test cohorts, respectively (all P <0.05). The combined model showed better discrimination than the clinical-radiological model in the training (AUC=0.853 vs. 0.612, P <0.001) and the total test (AUC=0.842 vs. 0.638, P <0.05) cohorts. The decision curves revealed that the combined model had greater clinical applicability than the clinical-radiological model. The model combining CT-based DLR and clinical-radiological features showed satisfactory performance for predicting OPM in patients with PDAC.

The added value of ultrasound imaging biomarkers to clinicopathological factors for the prediction of high-risk Oncotype DX recurrence scores in patients with breast cancer.

The Oncotype DX (ODX) recurrence score (RS), a 21-gene assay, has been proven to recognize patients at high risk of recurrence (RS ≥26) who would benefit from chemotherapy. However, it has limited availability and high costs. Our study thus aimed to identify ultrasound (US) imaging biomarkers and develop a prediction model for identifying patients with a high ODX RS. In this retrospective study, consecutive patients with T1-3N0-1M0 breast cancer who were hormone receptor positive and human epidermal growth factor receptor 2 (HER2) negative who had an available ODX RS were reviewed. Patients treated from May 2012 and December 2015 were placed into a training cohort, and those treated from January 2016 to January 2017 were placed in a validation cohort. Clinicopathologic data were collected, and preoperative US scans were analyzed. Univariable and multivariable regression analyses were performed to evaluate the independent predictors for a high-risk of breast cancer in the training cohort, and a nomogram was developed and evaluated with the area under the receiver operating characteristic curve (AUC), calibration curve, and decision curve analysis (DCA). A total of 363 patients were in the training cohort and 160 in the validation cohort, with the proportion with a high RS (RS 26-100) being 14% and 13.1%, respectively. Echogenic halo, enhanced posterior echo, low level of progesterone receptor (PR), and high Ki-67 index were identified as independent risk factors for high RS (all P values <0.05). The nomogram was constructed based on the combined model, which showed a better discrimination ability than did the clinicopathological model [combined model: AUC =0.95, 95% confidence interval (CI): 0.93-0.97; clinicopathological model: AUC =0.89, 95% CI: 0.86-0.92; P=0.001] and greater clinical benefit according to DCA. Furthermore, the nomogram was found to be effective in the validation cohort (AUC =0.90, 95% CI: 0.84-0.94), especially in patients with stage T1N0M0 disease (AUC =0.91, 95% CI: 0.84-0.95). US features may serve as valuable imaging biomarkers for the prediction of high recurrence risk in patients with T1-3N0-1M0 breast cancer and hormone receptor (HR)-positive and HER2-negative status. A nomogram incorporating PR status, Ki-67 index, and US imaging biomarkers showed a good discrimination ability in the early selection of patients at high risk of recurrence, especially in those with stage T1N0M0 disease.