Add BrightSurf on Google Email

Comparing AI anatomy segmentation models when ground truth is missing

A recent study introduces a practical framework for comparing AI-based anatomy segmentation models in the absence of expert reference annotations. The work focuses on chest CT scans from the National Lung Screening Trial dataset and evaluates how consistently different open-source models label anatomical structures. Key findings includ...

SourceSPIE--International Society for Optics and Photonics·JournalJournal of Medical Imaging·DateMay 13, 2026

Novel theranostic approach for radioimmunotherapy achieves curative responses in colorectal cancer tumors

A new pretargeted radioimmunotherapy technique has been shown to be effective and safe in eradicating tumors from a preclinical colorectal cancer model. The multi-step theranostic approach delivers alpha-emitting radiation directly to tumors while limiting exposure to healthy tissues.

SourceSociety of Nuclear Medicine and Molecular Imaging·JournalJournal of Nuclear Medicine·DateApr 30, 2026

Cardiac CT scans see the future: Visualizing "invisible" heart risks

Researchers at Kumamoto University discovered that combining cardiac Computed Tomography (CT) scan markers can identify patients at highest risk for future heart failure and death. The 'delayed phase' scan detects localized scarring and subtle damage throughout the heart muscle, providing a synergistic view of heart health.

SourceKumamoto University·JournalEuropean Heart Journal - Cardiovascular Imaging·TypeObservational study·DateApr 21, 2026

Markers of lymphoma cancer relapse identified

A new study from the University of Missouri identified over 10 genetic or molecular markers that predict follicular lymphoma relapse early, allowing for targeted surveillance testing. This could improve patient outcomes, reduce unnecessary imaging tests, and lower healthcare costs.

SourceUniversity of Missouri-Columbia·JournalAmerican Journal of Clinical Oncology·TypeData/statistical analysis·DateMar 31, 2026

Lipid identification: Chemical fingerprint instead of dye

A new microscopy method can distinguish lipid species in living cells using mid-infrared illumination and optoacoustic detection, producing a unique spectral fingerprint. This approach eliminates the need for chemical labels, reducing stress on cells and enabling real-time lipid mapping.

The Journal of Nuclear Medicine Ahead-of-Print Tip Sheet: March 27, 2026

Recent studies published in The Journal of Nuclear Medicine have made significant advancements in the field. Researchers have developed a new tracer to detect active collagen turnover following myocardial infarction and evaluated folate-based radioconjugates for ovarian cancer imaging. Additionally, targeted radiotherapy shows promise ...

SourceSociety of Nuclear Medicine and Molecular Imaging·JournalJournal of Nuclear Medicine·DateMar 27, 2026

University of Houston BRAIN Center finds exposure to nature associated with reductions in negative emotions

A systematic review and meta-analysis of 33 studies found that nature exposure is associated with reductions in negative emotions and increases in positive emotions. Experts recommend integrating nature into urban design to promote brain health and treat mental illnesses.

SourceUniversity of Houston·JournalInternational Journal of Environmental Research and Public Health·DateMar 24, 2026

The Journal of Nuclear Medicine Ahead-of-Print Tip Sheet: March 20, 2026

The study investigated the link between heart attack-induced inflammation and cognitive decline, as well as the potential of PET-based metabolic tumor volume to predict CAR T-cell therapy outcomes. Additionally, researchers combined advanced imaging with genetic profiling to explore hidden metabolic patterns in recurrent brain cancer.

SourceSociety of Nuclear Medicine and Molecular Imaging·JournalJournal of Nuclear Medicine·DateMar 19, 2026

Can a specialized AI model steer doctors toward the right scan?

A specialized AI model, AMIR-GPT, has been developed to improve radiology guideline alignment, outperforming general purpose models in 33.3% of test responses. However, the model's performance varied across performance bands, and qualitative review revealed limitations, such as omissions and deviations from standard recommendations.

SourceIntelligent Medicine·JournalIntelligent Medicine·TypeComputational simulation/modeling·DateMar 17, 2026

IEEE researchers achieve 20x signal boost in cerebral blood flow monitoring with next-generation interferometric diffusing wave spectroscopy

Researchers optimize interferometric diffusing wave spectroscopy technique to boost weak optical field returning from the brain, achieving over 20x signal to noise ratio. The novel approach provides higher brain sensitivity compared to DCS-inspired approaches and is approximately two orders of magnitude less expensive.

SourceInstitute of Electrical and Electronics Engineers·JournalIEEE Journal of Selected Topics in Quantum Electronics·TypeExperimental study·DateMar 11, 2026

Chinese Neurosurgical Journal study develops radiomics model to predict secondary decompressive craniectomy

A Chinese Neurosurgical Journal study developed a radiomics-based machine learning model to identify high-risk patients with traumatic brain injury who require emergency decompressive surgery. The model accurately distinguished patients who later required secondary surgery, suggesting its potential to complement clinical judgment.

SourceChinese Neurosurgical Journal·JournalChinese Neurosurgical Journal·TypeObservational study·DateMar 10, 2026

Precision tumor imaging with a fluorescence probe and engineered enzymes

Researchers developed a bioorthogonal fluorescence probe and matching reporter enzyme that selectively activates at targeted tumor sites, enabling high-contrast tumor visualization with minimal background. This technology has potential for improving cancer surgery outcomes and may also be adapted for targeted drug delivery.

SourceUniversity of Tokyo·JournalJournal of the American Chemical Society·TypeExperimental study·DateMar 1, 2026

Targeted PET/CT predicts early treatment response in rheumatoid arthritis patients

A new targeted PET/CT tracer can detect treatment response in rheumatoid arthritis patients as little as four weeks after treatment initiation. The imaging technique uses macrophages, a type of white blood cell, as a biomarker for disease activity, potentially allowing non-responders to pursue more effective therapies.

SourceSociety of Nuclear Medicine and Molecular Imaging·JournalJournal of Nuclear Medicine·DateFeb 27, 2026

The Journal of Nuclear Medicine ahead-of-print tip sheet: February 27, 2026

Researchers have developed a new PSMA-targeted PET tracer that shows promise in early human studies, providing strong tumor uptake and favorable radiation dosimetry. Additionally, advanced dynamic PET imaging has been used to track heart amyloid buildup more precisely, detecting significant reductions after six months of tafamidis ther...

SourceSociety of Nuclear Medicine and Molecular Imaging·JournalJournal of Nuclear Medicine·DateFeb 27, 2026

The Journal of Nuclear Medicine ahead-of-print tip sheet: February 20, 2026

Researchers have made significant advancements in nuclear medicine by developing a new reporting framework for PET/CT scans, using targeted radiation therapy to treat sarcoma, and investigating how genetic mutations affect prostate cancer treatment outcomes. These studies aim to improve diagnosis and treatment for various cancers.

SourceSociety of Nuclear Medicine and Molecular Imaging·JournalJournal of Nuclear Medicine·DateFeb 20, 2026

Magnetic resonance imaging opens the door to better treatments for underdiagnosed atypical Parkinsonisms

A new study uses advanced MRI to accurately diagnose progressive supranuclear palsy (PSP) and corticobasal degeneration (CBD), two underdiagnosed conditions. This breakthrough enables precise clinical trials and transforms treatment options for patients with balance problems, falls, stiffness, or difficulties with speech and movement.

SourceInstitut de Recerca Sant Pau (Sant Pau Research Institute)·JournalThe Journal of Prevention of Alzheimer s Disease·TypeObservational study·DateFeb 16, 2026

AI model can read and diagnose a brain MRI in seconds

A new AI-powered model can read a brain MRI and diagnose neurological conditions with up to 97.5% accuracy, predicting treatment urgency and automating alerts for immediate medical attention. The technology has the potential to transform neuroimaging at health systems across the US, reducing workload and improving patient outcomes.

SourceMichigan Medicine - University of Michigan·JournalNature Biomedical Engineering·TypeComputational simulation/modeling·DateFeb 6, 2026