NeuroQuant® Brain Tumor: Revolutionizing Neuro-Oncology with AI-Driven Precision
Cortechs.ai’s FDA-cleared NeuroQuant Brain Tumor software represents a transformative innovation in neuro-oncology, enhancing radiologists’ ability to evaluate brain tumors with greater accuracy and efficiency. Leveraging artificial intelligence (AI), the software automates segmentation and volumetric analysis of high- and low-grade diffuse gliomas, both pre- and post-treatment, enabling precision-based longitudinal monitoring of tumor changes over time. Applied Radiology met with company representatives at RSNA 2024 to discuss the technology’s capabilities.
“The software calculates the volume of enhancing tumor tissue as well as the surrounding zone of FLAIR hyperintensity,” explains Suzie Bash, MD, a Medical Director at RadNet and Chief Medical Officer at Cortechs.ai. FLAIR hyperintensity, if present, may indicate vasogenic edema, gliosis, or neoplastic infiltration into adjacent brain tissues. The software distinguishes and quantifies the enhancing tumor tissue and surrounding FLAIR hyperintensity, as well as the central necrotic core and post-operative resection cavity.1 These measurements facilitate longitudinal tracking of these tumor subregions in addition to overall tumor size.
According to Dr Bash, “automated, AI-powered segmentation significantly enhances the accuracy and efficiency of longitudinal brain tumor assessment across multiple time points. Traditional manual measurements are labor-intensive, prone to inaccuracies, and subject to inter-reader variability due to complex tumor morphology and technical variations between studies. In contrast, automated tumor volume quantification provides precise, reproducible, and objective measurements for evaluating interval changes.”
Addressing Tumor Progression Vs. Treatment-Related Changes
“One of the most significant hurdles in brain tumor imaging is differentiating disease recurrence from treatment-induced pseudoprogression. To address this, NeuroQuant Brain Tumor incorporates patented Restriction Spectrum Imaging (RSI), an advanced diffusion-based deep learning method,” explains Dr Bash. “RSI helps distinguish true tumor progression, which appears ‘hot’ on RSI maps, from pseudoprogression, thereby serving as a valuable imaging biomarker for glioblastoma multiforme.”
Jeff Rudie, MD, PhD, a neuroradiologist at Scripps Health and assistant adjunct professor at UC San Diego, highlights a recent novel deep learning model his team developed integrating RSI with traditional structural MRI sequences (T1 pre-contrast, T1 post-contrast, T2, and FLAIR) for more precise identification of regions of active cellular tumor. The model augments the capability to differentiate true progression from pseudoprogression, as well as infiltrating tumor from edema and gliosis.2
“RSI alone is useful, but combining it with structural sequences improves our ability to detect progression and predict survival,” notes Dr Rudie. Future iterations of NeuroQuant Brain Tumor will incorporate this advanced combined RSI cellular tumor model.
Expanding Applications and Outcomes Impact
NeuroQuant Brain Tumor is expanding its capabilities this year to include meningiomas and metastases, which represent a significant proportion of central nervous system neoplasms, states Dr Rudie. He also notes that the software can be integrated into radiation treatment planning systems, offering enhanced precision in targeting lesions, ultimately improving patient outcomes.
“Early detection of metastases, even as small as one millimeter, through automated software, has the potential to significantly impact patient morbidity and mortality,” emphasizes Dr Bash.
NeuroQuant Brain Tumor was trained on data from 59 sites using multichannel input and multiclass output convolutional neural networks, underscoring its robust development process for brain tumor imaging, Dr Rudie says. This technology provides critical support for radiologists, oncologists, and neuro-oncologists by delivering an objective, quantitative analysis of evolving brain tumors.
By advancing the precision and efficiency of tumor imaging, NeuroQuant Brain Tumor exemplifies the potential of AI-driven tools to revolutionize cancer diagnosis, treatment, and longitudinal care.