NetraMark Partners with Academic Medical Center to Apply AI Platform for Glioblastoma Trial Optimization
核心洞察
NetraMark Holdings Inc. (搜索) announced a new collaboration with a leading U.S. academic medical center to apply its NetraAI (搜索) platform to glioblastoma (搜索) datasets for identifying patient subgroups.
The partnership aims to address the 90% failure rate in glioblastoma (搜索) clinical trials by analyzing cerebrospinal fluid (搜索) proteomic data to develop therapeutic decision support tools.
NetraAI (搜索)'s explainable AI technology will focus on six key objectives including disease recurrence markers, treatment response mechanisms, and biomarker discovery to optimize future trial designs.
NetraMark Holdings Inc. (搜索) has announced a strategic collaboration with a leading U.S. academic medical center to apply artificial intelligence technology to one of oncology's most challenging cancers. The partnership will leverage NetraMark's NetraAI (搜索) platform to analyze glioblastoma (搜索) datasets with the goal of identifying patient subgroups that could inform more successful clinical trial designs.
The collaboration addresses a critical need in glioblastoma (搜索) research, where clinical trials face a failure rate exceeding 90%. This high failure rate stems from heterogeneous patient populations, limited predictive biomarkers, and poorly defined inclusion criteria. Glioblastoma remains an aggressive and difficult-to-treat cancer with a median survival of just 15 months and fewer than 7% of patients surviving beyond five years.
AI-Driven Analysis of Proteomic Data
Under the collaboration agreement, NetraMark gains access to clinical and biomarker datasets through a license to proprietary know-how. The company's NetraAI (搜索) platform will analyze longitudinal cerebrospinal fluid (搜索) (CSF) proteomic datasets generated on the SomaLogic (搜索) platform, along with related de-identified materials provided by the academic partner.
NetraAI (搜索) will segment the data into explainable and unexplainable subpopulations, seeking to deliver insights in the form of NetraPersonas—collections of patients, variables, and statistical evidence that can be leveraged to optimize future trial enrichment strategies.
"We are excited to collaborate with a leading academic medical center on this challenging area of oncology," said Josh Spiegel, President of NetraMark. "By applying NetraAI (搜索) to high-quality glioblastoma (搜索) datasets, we aim to identify explainable patient subgroups that can guide treatment strategies, accelerate biomarker discovery, and support the design of more successful clinical trials."
Six-Pronged Research Approach
The collaboration will pursue six specific research objectives designed to generate novel insights and hypotheses from the proteomic datasets:
Tumor Classification: The analysis will identify markers distinguishing glioblastoma (搜索) samples from non-tumor controls and differentiate glioblastoma from supratentorial brain metastases (搜索) to refine glioma-specific biomarkers.
Disease Progression: Researchers will contrast primary and recurrent gliomas to identify molecular markers of recurrence and tumor evolution.
Treatment Impact Assessment: The study will evaluate paired pre- and post-surgical resection samples to understand molecular changes from surgery, analyze pre- and post-chemoradiation samples to assess therapy response and resistance mechanisms, and explore pre- and post-immunotherapy lumbar CSF samples to identify molecular effects of immunotherapy.
Explainable AI Technology
NetraAI (搜索) distinguishes itself from other AI-based methods through focus mechanisms that separate small datasets into explainable and unexplainable subsets. Unexplainable subsets represent collections of patients that can lead to suboptimal overfit models and inaccurate insights due to poor correlations with the variables involved.
The platform uses explainable subsets to derive insights and hypotheses, including factors that influence treatment and placebo responses, as well as adverse events, potentially increasing the chances of clinical trial success. Many other AI methods lack these focus mechanisms and assign every patient to a class, often leading to overfitting that can obscure critical information.
NetraMark's technology employs a novel topology-based algorithm that parses patient datasets into subsets of people strongly related according to several variables simultaneously. This approach allows the company to work with smaller datasets while accurately segmenting diseases into different types and classifying patients for sensitivity to drugs and treatment efficacy.
The collaboration represents an important opportunity in applying explainable AI to a significant challenge in oncology, with the potential to help trial sponsors design more efficient, targeted, and successful glioblastoma (搜索) studies.