What the grant actually funds
Wayne State University School of Medicine has been awarded a five-year, $2.54 million grant from the National Cancer Institute, a development also covered by Radiology Business. Co-principal investigators Csaba Juhasz, PhD (professor of pediatrics, neurology and neurosurgery) and Otto Muzik, PhD (professor of pediatrics, neurology and radiology) are developing a PET tracer that targets the tryptophan-kynurenine pathway — a metabolic route glioma cells use to help evade the immune system.
The team had already shown that tracking this pathway with PET could identify infiltrative tumor tissue beyond what standard MRI picks up, using a carbon-11-based tracer. The problem was practical, not scientific: carbon-11 has a short half-life, so it requires an on-site cyclotron and specialized facilities, which limited how widely the technique could be tested or deployed. The new grant funds a fluorine-18 version of the tracer, developed with the Barbara Ann Karmanos Cancer Institute, whose longer half-life makes it far more workable for routine clinical PET.
"Our goal is to give physicians a clearer, highly precise map of active tumor tissue before and after treatment," Juhasz said. "By better targeting active cancer cells and distinguishing recurrence from treatment side effects, we aim to make treatments more effective and ultimately extend patient survival." Wael Sakr, MD, dean of the Wayne State University School of Medicine, added that "this NIH award recognizes the innovative work of Drs. Juhasz and Muzik and the strength of the multidisciplinary team they have assembled. Their research has the potential to provide clinicians with greater insight into the extent and activity of these tumors, supporting more precise and informed treatment decisions for patients with brain cancer."
Notice what the goal is not
Nobody involved is trying to build a better way to say "yes, there is a tumor here." On a contrast-enhanced MRI, a glioblastoma is rarely subtle — the mass is usually obvious the moment the images load. The grant is funding something narrower and harder: a way to quantify where the biologically active tumor actually ends, and to tell that apart from tissue that merely looks abnormal after treatment.
That distinction matters because two separate imaging problems get collapsed into "can you see the tumor," and both are reporting problems, not detection problems:
Infiltration beyond the visible margin
Glioblastoma cells migrate well past the contrast-enhancing region a radiologist circles on MRI. Standard sequences can't reliably quantify that invisible extension, which is exactly why a metabolic PET tracer — something that lights up active tumor biology rather than just structural distortion — is the tool being funded here, not a sharper MRI.
Recurrence vs. treatment effect
After surgery, radiation, and chemotherapy, new or enlarging contrast enhancement can mean the tumor is back, or it can mean the brain is reacting to treatment. Conventional MRI often cannot reliably tell these apart, and the two readings point to opposite next steps — more therapy versus watchful waiting.
A number a surgeon can act on
A radiology report that says "tumor present" is not actionable for a neurosurgeon deciding how much tissue to remove. A quantitative extent map — even an imperfect one — is. That gap between a qualitative impression and an actionable, structured measurement is the bottleneck this grant, and a wave of similar NIH-funded projects, are aimed at closing.
Why the reporting layer, not the scanner, decides survival
Extent of surgical resection is one of the few glioblastoma prognostic factors a hospital can actually influence, and the numbers behind it are substantial. A 2016 meta-analysis in JAMA Oncology pooling 37 studies and more than 41,000 patients found that gross total resection, compared with subtotal resection, reduced the relative risk of death at one year by about 38% and at two years by about 16%. Median survival with standard-of-care treatment still sits at roughly 12 to 15 months, which is precisely why every percentage point of safely resectable tumor matters.
But a surgeon can only resect what imaging tells them is there — and can only avoid resecting healthy tissue if imaging tells them, with some confidence, what isn't tumor. A 2018 review in Contrast Media & Molecular Imaging puts a number on how often that confidence is misplaced: pseudoprogression occurs in an estimated 10% to 30% of treated glioma patients, with most cases surfacing within three months of finishing treatment — the exact window when clinicians are deciding whether a therapy is failing. Get that reporting call wrong in either direction, and a patient either stops a treatment that was working or continues one that wasn't.
Detection reporting vs. quantitative extent reporting
The distinction shows up clearly when the two reporting styles are placed side by side:
| Dimension | Detection reporting | Quantitative extent reporting |
|---|---|---|
| Core question answered | Is there a mass? | Where exactly does active tumor end? |
| Typical output | Descriptive impression | Measured volume / metabolic map |
| Post-treatment scans | Enhancement noted | Recurrence vs. treatment-effect distinction |
| Clinical action supported | Refer for biopsy or follow-up | Guide resection margin or next-line therapy |
Where this connects beyond neuro-oncology
The Wayne State project sits in PET/MRI and surgical planning — outside CT, and outside where xAID's reporting tools operate today. But the underlying lesson generalizes across imaging: the expensive, underfunded part of most reporting workflows is rarely spotting an abnormality. It's turning that abnormality into a structured, quantitative, comparable measurement that a referring clinician can act on without re-reading the images themselves. That is the same thesis behind structured, foundation-model-based reporting in CT: the value isn't a flag that says "abnormal," it's a comprehensive, ready-to-sign draft with the measurements and comparisons already structured — reviewed in-house before it ever reaches the radiologist who signs it.
Frequently asked questions
What did the new NIH grant for glioblastoma imaging fund?
The National Cancer Institute awarded Wayne State University School of Medicine a five-year, $2.54 million grant to develop a fluorine-18 PET tracer targeting the tryptophan-kynurenine pathway, a metabolic process glioma cells use to evade the immune system. Co-principal investigators Csaba Juhasz, PhD, and Otto Muzik, PhD, aim to produce quantitative maps of tumor metabolic activity rather than a simple presence/absence read.
Why is detecting a glioblastoma not the hard part of imaging it?
A visible mass on MRI is rarely in doubt. The harder problems are quantifying how far tumor cells have infiltrated beyond the visible margin, and distinguishing true recurrence from pseudoprogression or radiation necrosis after treatment — findings that can look identical on conventional imaging but call for opposite clinical decisions.
How often is glioblastoma imaging misread as recurrence when it isn't?
Pseudoprogression — new or enlarging contrast enhancement that mimics tumor recurrence but is actually a treatment effect — has a reported incidence of roughly 10% to 30% in treated glioma patients, according to a 2018 review in Contrast Media & Molecular Imaging. Most cases appear within three months of finishing treatment, precisely when clinicians are deciding whether therapy is failing.
Does quantitative tumor-extent imaging actually change outcomes?
Extent of surgical resection is one of the few modifiable factors tied to glioblastoma survival. A 2016 JAMA Oncology meta-analysis of 37 studies and more than 41,000 patients found gross total resection cut relative one-year mortality risk by about 38% compared with subtotal resection. Imaging that quantifies tumor margins more precisely is what makes that more complete resection possible without removing healthy brain tissue.
Source: Wayne State University School of Medicine (August 2026), also covered by Radiology Business. Extent-of-resection figures from Brown et al., JAMA Oncology (2016). Pseudoprogression incidence from Zikou et al., Contrast Media & Molecular Imaging (2018). Median survival figure per peer-reviewed literature on glioblastoma outcomes. Figures are rounded as reported.