Objectivity The article demonstrates strong objectivity by adopting a balanced "current status-issues" framework, without one-sidedly overstating research progress on glioblastoma multiforme (GBM). It systematically outlines the application value of imaging technologies such as MRI (with T1-weighted contrast enhancement for visualizing tumor necrosis) and PET (with 18F-FET for distinguishing recurrence from radiation necrosis), while also clearly pointing out limitations like the poor soft tissue contrast of CT and the need for a nearby cyclotron for 11C-methylmethionine (11C-MET). When elaborating on the standard "surgery + radiotherapy + temozolomide (TMZ)" regimen, it simultaneously highlights the therapeutic bottleneck that the 5-year survival rate remains only 4.6%-5.6%. Literature screening covers multiple databases including MEDLINE, with studies up to February 2025 included in accordance with PRISMA guidelines, avoiding selective inclusion of positive results. It also maintains a neutral stance when discussing controversies in AI applications (e.g., low proficiency of doctors in operating AI tools) and does not shy away from technical flaws. Credibility The credibility is constrained by the peer review results (with scientific quality, innovation, and other indicators all rated Grade D), but there are still supporting elements. The authors are affiliated with the Department of Medical Imaging and Radiation Sciences, University of Sherbrooke, Canada; the corresponding author is a professor and contact information is provided. Literature citations are standardized, including DOI and PMID, and Tables 1-3 clearly summarize key data such as imaging tracers and chemotherapeutic drugs, with a reproducible PRISMA process. However, the single-blind review (involving only one expert from Japan) may lead to a limited perspective, and the article does not mention data replication and verification, which weakens part of its credibility. Scientific Quality As a review article, its scientific quality is moderate. Its strengths lie in a clear structure, progressing logically from "epidemiology-imaging-treatment-AI," with each conclusion supported by literature (e.g., citing Zheng et al.’s AI study using 40 million samples). It also compares the advantages and disadvantages of different imaging modalities and treatment regimens, providing references for clinical practice. Nevertheless, there are shortcomings: it fails to quantify key indicators (such as AI segmentation accuracy), describes research data on emerging nanodrugs and PARP inhibitors in a vague manner, and does not analyze diagnostic and therapeutic differences among different molecular subtypes (e.g., IDH-mutant GBM), resulting in insufficient scientific depth. Resolved Issues The article clarifies regional differences in GBM epidemiology, applicable scenarios of mainstream imaging and treatment regimens, and constructs an application framework for AI in GBM, providing systematic knowledge for beginners. It also identifies core bottlenecks such as the blood-brain barrier and difficulties in early detection, defining the research scope. Unresolved and Existing Issues It does not propose specific biomarkers for early detection (e.g., circulating tumor cells), fails to quantify data on the generalizability of AI models, provides ambiguous solutions for multi-center collaboration and data privacy protection, and does not discuss personalized diagnosis and treatment strategies for different molecular subtypes. Future Solutions The authors mention that future efforts should focus on strengthening structured validation and ethical oversight of AI. Specifically, multi-center collaboration can be conducted to establish a standardized imaging database and carry out prospective AI trials; subtype-specific treatment models can be developed by integrating genomic data; and early screening schemes combining "AI + biomarkers" can be explored to fill current research gaps.