e20566 Background: Immunotherapy (IO) is a viable therapeutic approach for non-small cell lung cancer (NSCLC). Despite the significant survival benefit of immune checkpoint inhibitors PD-1/PD-L1 in NSCLC, the objective response rate is not more than 50%, even with the combination of chemotherapy. There is heterogeneity in response to IO with patients with metastatic disease with long-term response, lasting years, and on the other hand, patients who fail to respond at all. While PD-L1 is a known biomarker to stratify IO responses, it is, at best, imperfect. There is thus an urgent need to discover new predictive multimodal biomarkers of response to IO in such a setting. Methods: A retrospective cohort of 187 patients (Memorial Sloan Kettering - MSKCC MIND cohort, training=140, validation=47) with advanced NSCLC, treated with PD-L1 blockade, were analyzed to develop a machine learning-based algorithm predictive of response to immunotherapy. Multimodal baseline data, including clinical parameters, PD-L1 expression, TMB, and baseline CT scan data, were used to predict treatment response status (responders vs. non-responders). For each patient, available tumor masks were used to generate peritumoral masks. 1 st , 2 nd, and higher-order statistics of extracted 8 Haralick radiomics textures and intensity gradients of intratumoral and peritumoral areas were used. A 2-step feature selection was performed, optimizing Gini impurity and Area Under the Curve (AUC) with Random Forest (RF) to select the best intratumoral and peritumoral radiomic features and their respective probabilities of predicting response in the training cohort using 5-fold cross-validation. These radiomics-based prediction probabilities were combined with clinical data like smoking status, age, TMB, and PD-L1 score to train another RF model to predict response. They were validated by computing the AUC of the ROC on the validation cohort. Results: The combination of clinical, intratumoral, and peritumoral features led to an impressive prediction of response to IO in NSCLC patients from our database, with an AUC=0.83. Peritumoral radiomics features alone led to an AUC=0.64 compared to intratumoral radiomics features of AUC=0.58, suggesting the importance of tumor microenvironment in the surrounding tumor areas in prediction of IO response. Conclusions: This proof-of-concept study suggests that machine learning applied to baseline multimodal data, including CT tumoral radiomics-based prediction probabilities and clinical variables, can help predict response to IO and improve patient selection. Prediction of IO response combining clinical parameters, intra- & peritumoral CT radiomics on validation cohort (n=47 patients). Features AUC Intratumoral CT radiomics 0.58 Peritumoral CT radiomics 0.64 Intra- + peritumoral CT radiomics + clinical parameters 0.83
Abstract. This study presents an estimation of the annual energy production (AEP) associated with active wake mixing (AWM) control strategies in a wind farm. To achieved this, we first conduct a series of high-fidelity large eddy simulations (LES) of a wind farm for various turbine layouts and control parameters. These simulations extend previous findings from two-turbine studies to a larger array of wind turbines, demonstrating the effectiveness of AWM in enhancing power generation, particularly in geometrically aligned wind farms situated in stable atmospheric boundary layers. The results indicate that while the conventional pulse method leads to the best performance for second-row turbines, the helix method leads to greater improvements in power generation for third-row turbines. Second, a framework for estimating the AEP associated with AWM strategies is developed within the FLOw Redirection and Induction in Steady-state (FLORIS) toolkit, using a new empirical Gaussian wake model. The FLORIS parameters are calibrated to the LES data and an optimization routine is established for determining the optimal use of AWM in a wind farm for maximizing AEP. AEP estimates are provided using Weibull data from the New York Bight for multiple turbine layouts and blade pitching amplitudes. Third, the AEP gains from wake mixing are compared to those from wake steering using the yaw optimization routines in FLORIS. The power performance is similar for both control methods, generally leading to power gains of 1 % to 3 % for the wind conditions where active wake mixing or steering is used, which translates to AEP gains that are mostly less than 1 % for the wind farm and control parameters considered in this study.
Objective To determine the prevalence and degree of abnormality of white matter hyperintensities (WMH) in a young and healthy population, with and without mild traumatic brain injury (mTBI). Design Retrospective review of high-resolution 3-Tesla 3D T2W FLAIR and SWI images. Setting A multicenter study of advanced neuroimaging biomarkers in mTBI. Participants 602 patients (413 mTBI and 189 controls) aged 15–50 years. Interventions (or Assessment of Risk Factors) All scans were independently reviewed by 2 board-certified blinded neuroradiologists. Outcome Measures The frequency and severity of WMH were identified; WMH was excessive if ≥5 objective, punctate white matter foci were present, lesions were >3mm or atypical in location. Main Results 32% of the mTBI patients and 40% of controls demonstrated at least 1 WMH. Of these patients, 34% of the mTBI group and 34% of the controls were deemed excessive. There was no difference in number of patients with at least 1 WMH (p=0.064) or excessive WMHs (p=0.313). Conclusions This is the largest study to date assessing the prevalence of WMH on high-resolution 3-Telsa 3D T2W-FLAIR in a youthful cohort of mTBI patients and controls. Interestingly, the prevalence of WMHs did not differ between groups and were observed in higher rates than reported in the current literature. The increased incidence rate is possibly attributable to the evolution of MRI technology. These results suggest an evidence-based need to reassess our understanding and interpretation of incidental findings in both a young and healthy population and in patients with mTBI.
