PURPOSE: NRG-RTOG0617 demonstrated a detrimental effect of uniform high-dose radiation in stage III non–small cell lung cancer. NRG-RTOG1106/ECOG-ACRIN6697 (ClinicalTrials.gov identifier: NCT01507428 ), a randomized phase II trial, studied whether midtreatment ¹⁸ F-fluorodeoxyglucose positron emission tomography/computed tomography (FDG-PET/CT) can guide individualized/adaptive dose-intensified radiotherapy (RT) to improve and predict outcomes in patients with this disease. MATERIALS AND METHODS: Patients fit for concurrent chemoradiation were randomly assigned (1:2) to standard (60 Gy/30 fractions) or FDG-PET–guided adaptive treatment, stratified by substage, primary tumor size, and histology. All patients had midtreatment FDG-PET/CT; adaptive arm patients had an individualized, intensified boost RT dose to residual metabolically active areas. The primary therapeutic end point was 2-year centrally reviewed freedom from local-regional progression (FFLP), defined as no progression in or near the planning target volume and/or regional nodes. FFLP was analyzed on a modified intent-to-treat population at a one-sided Z -test significance level of 0.15. The primary imaging end point was centrally reviewed change in SUV _peak from baseline to midtreatment; its association with FFLP was assessed using the two-sided Wald test on the basis of Cox regression. RESULTS: Of 138 patients enrolled, 127 were eligible. Adaptive-arm patients received a mean 71 Gy in 30 fractions, with mean lung dose 17.9 Gy. There was no significant difference in centrally reviewed 2-year FFLP (59.5% and 54.6% in standard and adaptive arms; P = .66). There were no significant differences in protocol-specified grade 3 toxicities, survival, or progression-free survival ( P > .4). Median SUV _peak and metabolic tumor volume (MTV) in the adaptive arm decreased 49% and 54%, from pre-RT to mid-RT PET. However, ΔSUV _peak and ΔMTV were not associated with FFLP (hazard ratios, 0.997; P = .395 and .461). CONCLUSION: Midtreatment PET-adapted RT dose escalation as given in this study was safe and feasible but did not improve efficacy outcomes.

Objective: To investigate the comparative efficacy and safety of the low-dose versus standard-dose alteplase using real-world acute stroke registry data from Asian countries. Methods: Individual participant data were obtained from nine acute stroke registries from China, Japan, Philippines, Singapore, South Korea, and Taiwan between 2005 and 2018. Inverse probability of treatment weight was used to remove baseline imbalances between those receiving low-dose versus standard-dose alteplase. The primary outcome was death or disability defined by modified Rankin Scale scores of 2 to 6 at 90 days. Secondary outcomes were symptomatic intracerebral hemorrhage and death. Generalized linear mixed models with the individual registry as a random intercept were performed to determine associations of treatment with low-dose alteplase and outcomes. Results: Of the 6250 patients (mean age 66 years, 36% women) included in these analyses, 1610 (24%) were treated with low-dose intravenous alteplase. Clinical outcomes for low-dose alteplase were not significantly different to those for standard-dose alteplase, adjusted odds ratios for death or disability: 1.00 (0.85–1.19) and symptomatic intracerebral hemorrhage 0.87 (0.63–1.19), except for lower death with borderline significance, 0.77 (0.59–1.01). Conclusions: The present analyses of real-world Asian acute stroke registry data suggest that low-dose intravenous alteplase has overall comparable efficacy for functional recovery and greater potential safety in terms of reduced mortality, to standard-dose alteplase for the treatment of acute ischemic stroke.

Clinical informatics, decision support for health professionals, electronic health records, and ehealth infrastructures.

De-identification of medical images intended for research is a core requirement for data sharing initiatives, particularly as the demand for data for artificial intelligence (AI) applications grows. The Center for Biomedical Informatics and Information Technology (CBIIT) of the United States National Cancer Institute (NCI) convened a two half-day virtual workshop with the intent of summarizing the state of the art in de-identification technology and processes and exploring interesting aspects of the subject. This paper summarizes the highlights of the second day of the workshop, the recordings and presentations of which are publicly available for review. The topics covered included pathology whole slide image de-identification, de-facing, the role of AI in image de-identification, and the NCI Medical Image De-Identification Initiative (MIDI) datasets and pipeline.

Feature dimensionality reduction plays an important role in radiomic studies with a large number of features. However, conventional radiomic approaches may suffer from noise, and feature dimensionality reduction techniques are not equipped to utilize latent supervision information of patient data under study, such as differences in patients, to learn discriminative low dimensional representations. To achieve robustness to noise and feature dimensionality reduction with improved discriminative power, we develop a robust collaborative clustering method to simultaneously cluster patients and radiomic features into distinct groups respectively under adaptive sparse regularization. Our method is built upon matrix tri-factorization enhanced by adaptive sparsity regularization for simultaneous feature dimensionality reduction and denoising. Particularly, latent grouping information of patients with distinct radiomic features is learned and utilized as supervision information to guide the feature dimensionality reduction, and noise in radiomic features is adaptively isolated in a Bayesian framework under a general assumption of Laplacian distributions of transform-domain coefficients. Experiments on synthetic data have demonstrated the effectiveness of the proposed approach in data clustering, and evaluation results on an FDG-PET/CT dataset of rectal cancer patients have demonstrated that the proposed method outperforms alternative methods in terms of both patient stratification and prediction of patient clinical outcomes.