Neuro-oncology advances最新文献

Background: This research aims to improve glioblastoma survival prediction by integrating MR images, clinical, and molecular-pathologic data in a transformer-based deep learning model, addressing data heterogeneity and performance generalizability.

Methods: We propose and evaluate a transformer-based nonlinear and nonproportional survival prediction model. The model employs self-supervised learning techniques to effectively encode the high-dimensional MRI input for integration with nonimaging data using cross-attention. To demonstrate model generalizability, the model is assessed with the time-dependent concordance index (Cdt) in 2 training setups using 3 independent public test sets: UPenn-GBM, UCSF-PDGM, and Rio Hortega University Hospital (RHUH)-GBM, each comprising 378, 366, and 36 cases, respectively.

Results: The proposed transformer model achieved a promising performance for imaging as well as nonimaging data, effectively integrating both modalities for enhanced performance (UCSF-PDGM test-set, imaging Cdt 0.578, multimodal Cdt 0.672) while outperforming state-of-the-art late-fusion 3D-CNN-based models. Consistent performance was observed across the 3 independent multicenter test sets with Cdt values of 0.707 (UPenn-GBM, internal test set), 0.672 (UCSF-PDGM, first external test set), and 0.618 (RHUH-GBM, second external test set). The model achieved significant discrimination between patients with favorable and unfavorable survival for all 3 datasets (log-rank P 1.9 × 10^-8, 9.7 × 10^-3, and 1.2 × 10^-2). Comparable results were obtained in the second setup using UCSF-PDGM for training/internal testing and UPenn-GBM and RHUH-GBM for external testing (Cdt 0.670, 0.638, and 0.621).

Conclusions: The proposed transformer-based survival prediction model integrates complementary information from diverse input modalities, contributing to improved glioblastoma survival prediction compared to state-of-the-art methods. Consistent performance was observed across institutions supporting model generalizability.

Background: Ensuring equitable access to treatments and therapies in the constantly evolving field of neuro-oncology is an imperative global health issue. With its unique demographic, cultural, socioeconomic, and infrastructure characteristics, Sub-Saharan Africa faces distinct challenges. This literature review highlights specific barriers to neuro-oncology care in the region and explores potential opportunities for enhancing access.

Methods: Predetermined keyword searches were employed to screen titles and abstracts using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses framework. Inclusion criteria were studies published between January 1, 2003, and June 20, 2023, specifically addressing the capacity and challenges of neuro-oncology in the Sub-Saharan African region. The data sources queried were PubMed and Google Scholar. Systematic reviews and meta-analyses were deliberately excluded. All authors conducted independent screening and structured data extraction meticulously.

Results: Our paper identified multiple challenges that impede access to quality treatment for brain tumors. These include constrained resources, insufficient training of healthcare professionals, certain cultural beliefs, and a general lack of awareness about brain tumors, all contributing to delayed diagnosis and treatment. Furthermore, the lack of detailed data on the incidence and prevalence of primary central nervous system tumors impairs the accurate assessment of disease burden and precise identification of areas requiring improvement. However, we discovered that ongoing research, advocacy, enhanced training, mentorship, and collaborative efforts present valuable opportunities for substantial progress in neuro-oncology access.

Conclusions: While we provide a glimpse of the current state, we hope these results will help stimulate dialogue and catalyze initiatives to surmount highlighted obstacles and improve neuro-oncology outcomes across Sub-Saharan Africa.

[This corrects the article DOI: 10.1093/noajnl/vdae072.].