Modernizing Neuro-Oncology: The Impact of Imaging, Liquid Biopsies, and AI on Diagnosis and Treatment

Abstract

Advances in neuro-oncology have transformed the diagnosis and management of brain tumors, which are among the most challenging malignancies due to their high mortality rates and complex neurological effects. Despite advancements in surgery and chemoradiotherapy, the prognosis for glioblastoma multiforme (GBM) and brain metastases remains poor, underscoring the need for innovative diagnostic strategies. This review highlights recent advancements in imaging techniques, liquid biopsies, and artificial intelligence (AI) applications addressing current diagnostic challenges. Advanced imaging techniques, including diffusion tensor imaging (DTI) and magnetic resonance spectroscopy (MRS), improve the differentiation of tumor progression from treatment-related changes. Additionally, novel positron emission tomography (PET) radiotracers, such as ¹⁸F-fluoropivalate, ¹⁸F-fluoroethyltyrosine, and ¹⁸F-fluluciclovine, facilitate metabolic profiling of high-grade gliomas. Liquid biopsy, a minimally invasive technique, enables real-time monitoring of biomarkers such as circulating tumor DNA (ctDNA), extracellular vesicles (EVs), circulating tumor cells (CTCs), and tumor-educated platelets (TEPs), enhancing diagnostic precision. AI-driven algorithms, such as convolutional neural networks, integrate diagnostic tools to improve accuracy, reduce interobserver variability, and accelerate clinical decision-making. These innovations advance personalized neuro-oncological care, offering new opportunities to improve outcomes for patients with central nervous system tumors. We advocate for future research integrating these tools into clinical workflows, addressing accessibility challenges, and standardizing methodologies to ensure broad applicability in neuro-oncology.

Keywords: artificial intelligence; convolutional neural networks; deep learning; glioblastoma; liquid biopsy; neuro-oncology; positron emission tomography; pseudoprogression; radiomics; transfer learning.

Figure 1

Post-therapy MRI challenges in differentiating tumor progression and treatment effects. The main obstacles in assessing brain tumors using MRI post-treatment are presented as follows: (A) Pseudoprogression: Post-radiation inflammation can cause temporary tumor enlargement on imaging, mimicking actual tumor progression. Radiation therapy triggers increased vascular permeability and inflammatory responses, causing this phenomenon, which typically resolves within 3–4 months. Misinterpretation as true progression can complicate diagnosis and delay appropriate management. (B) Radiation Necrosis: A delayed complication of radiation therapy, radiation necrosis manifests as cell death, increased vascular permeability, and cerebral edema, usually appearing 1–2 years post-treatment. These pathological changes can create imaging findings that resemble tumor recurrence, complicating diagnosis and treatment planning. Differentiating radiation necrosis from true tumor progression often requires advanced imaging techniques or complementary diagnostics. (C) Contrast and the Blood-Brain Barrier (BBB): The BBB, consisting of endothelial tight junctions, pericytes, and astrocytic foot processes, regulates the passage of substances into the brain. Common MRI contrast agents like gadolinium may struggle to penetrate the intact BBB, particularly in deep-seated or partially infiltrative tumors. This limitation reduces imaging sensitivity and may obscure tumor boundaries, necessitating alternative or adjunctive diagnostic methods.

Figure 2

Overview of liquid biopsy components and applications in neuro-oncology. The process and key components of liquid biopsy, a minimally invasive diagnostic tool for neuro-oncology, are illustrated, highlighting its potential for early detection and patient management. (1) Tumor Origin: Tumors within the central nervous system (CNS) shed molecular and cellular components into surrounding biological fluids, such as blood and cerebrospinal fluid (CSF). These materials are subsequently released into circulation. (2) Sample Collection: Liquid biopsy involves the collection of patient blood or CSF through minimally invasive techniques. This allows for the isolation of tumor-derived biomarkers without requiring direct tissue sampling. (3) Key Biomarkers: Free Circulating DNA (cfDNA): Tumor cells release DNA fragments into the bloodstream, carrying tumor-specific mutations or methylation patterns. Extracellular Vesicles (EVs): Tumor cells release nanoparticles containing proteins, RNA, and DNA that reflect tumor biology and progression. Circulating Tumor Cells (CTCs): Intact tumor cells that escape from the primary tumor into circulation, providing direct cellular information about the cancer. Tumor-Educated Platelets (TEPs): Tumor cells reprogram platelets to reflect tumor-specific RNA and protein signatures, aiding in tumor detection and characterization. (4) Clinical Applications: The liquid biopsy components have a wide range of diagnostic and prognostic uses, including: Early Detection: Identifying tumors before clinical symptoms arise. Monitoring Tumor Progression: Tracking changes in tumor biology over time. Distinguishing Treatment-Related Changes: Differentiating true tumor progression from pseudoprogression or radiation necrosis. Predicting Patient Outcomes: Assessing tumor aggressiveness and potential treatment responses. Guiding Treatment: Informing personalized therapeutic strategies based on real-time tumor monitoring.

Figure 3

Hierarchical framework and applications of artificial intelligence in neuro-oncology diagnostics. The hierarchical relationship between artificial intelligence (AI), machine learning (ML), deep learning (DL), convolutional neural networks (CNNs), transformer-based AI (TBAI), and transfer learning (TL) is represented, focusing on their applications in neuro-oncology diagnostics. AI forms the broadest layer, encompassing techniques for analyzing complex neuro-oncology datasets to identify tumors, assess treatment responses, and predict patient outcomes. ML represents a subset of AI that uses algorithms to detect patterns in data, such as MRI scans or biomarker profiles, to aid in tumor diagnosis and treatment monitoring. DL, a specialized subset of ML, leverages large neuro-oncology datasets to make specific predictions, such as identifying glioma subtypes. CNNs, a distinct architecture within DL, extract detailed imaging features from scans (e.g., MRI, PET) to improve tumor detection, grading, and monitoring. TBAI is another subset of DL that utilizes an attention mechanism to identify the most crucial information, excelling in the context of large datasets and a variety of data to improve predictions about prognosis and tumor progression. TL is depicted as a technique that intersects these frameworks, adapting pre-trained models for specific neuro-oncology tasks, such as distinguishing glioblastoma from brain metastases. The arrows originating from TL represent its role in optimizing the performance of ML, DL, CNN, and TBAI: To ML: TL adapts existing ML algorithms for tailored diagnostic applications. To DL: TL leverages pre-trained DL models to enhance predictions with limited data. To CNNs: TL optimizes CNN architectures for specialized imaging tasks, improving accuracy and efficiency. To TBAI: TL allows TBAI to be focused on domain-specific tasks based on available data types.