Radiomics converts medical images into high-dimensional quantitative features that may capture subtle disease-related patterns not detectable through conventional analyses. The investigators pooled diagnostic performance with a random-effects meta-analysis and estimated summary sensitivity and specificity using a hierarchical summary receiver operating characteristic (HSROC) model. The researchers searched PubMed, EMBASE, IEEE Xplore and Scopus through September 17, 2026.

They followed PRISMA-DTA guidelines and used a preregistered protocol, identified as PROSPERO CRD420261286782. Risk of bias and reporting quality were assessed with QUADAS-2 and METRICS. Fourteen studies met the inclusion criteria, and eight contributed data to the main diagnostic meta-analysis.

For diagnostic discrimination, the pooled AUC was 0.81, with a 95% confidence interval of 0.73 to 0.88. The HSROC model estimated mean sensitivity at 0.736, with a 95% confidence interval of 0.666 to 0.795, and mean specificity at 0.735, with a 95% confidence interval of 0.668 to 0.793. Despite the strong overall AUC, statistical heterogeneity was substantial: the I-squared statistic was 84.2%.

This indicates that results varied markedly among the diagnostic studies contributing to the pooled estimate. The review characterized the studies’ overall quality as moderate to good, but its analysis also found that less rigorous validation procedures were associated with more optimistic performance. External validation was limited, and the reported aggregate results therefore do not establish that every radiomics model will perform equally well in routine clinical practice.

For treatment response, radiomics-based approaches also showed good discrimination in the meta-analysis. The reported AUC was 0.80 for longitudinal prediction and 0.88 for cross-sectional discrimination between treatment-response groups. These estimates concern the accuracy of statistical models in the included studies, not proof that radiomics can presently determine which treatment an individual patient should receive.

The supplied review abstract does not report confidence intervals or the number of contributing studies for the treatment-response estimates, limiting interpretation of their precision and evidence base. The authors called for standardized radiomics workflows, prospective study designs, integration with clinical markers and rigorous validation. Their findings position radiomics as a promising component of schizophrenia research for both diagnosis and treatment-response prediction, while showing that dependable individual-level clinical translation still requires stronger validation.

Subject of Research: Meta-analysis of radiomics-based MRI and PET models for schizophrenia diagnosis and treatment-response prediction

Article Title: Precision psychiatry in schizophrenia: a meta-analysis of radiomics-based advanced neuroimaging for diagnostic and treatment-response prediction

Article References: Caiazza, C., Fusco, G., Ugga, L., Toni, C., Cirillo, M., Sampogna, G., & Fiorillo, A. (2026). Precision psychiatry in schizophrenia: a meta-analysis of radiomics-based advanced neuroimaging for diagnostic and treatment-response prediction. Schizophrenia. https://doi.org/10.1038/s41537-026-00815-1

Image Credits: AI Generated

DOI: 10.1038/s41537-026-00815-1

Keywords: schizophrenia, radiomics, precision psychiatry, neuroimaging, MRI, PET, machine learning, treatment response, diagnostic accuracy, meta-analysis, Precision, psychiatry

The imaging collection combines structural and functional magnetic resonance imaging with positron emission tomography using tracers for amyloid and tau pathology. Structural and functional MRI provide complementary views of the brain, while the PET scans target two pathological features associated with Alzheimer disease. The cohort also includes clinical, cognitive, genetic and biomarker data, allowing investigators to examine imaging measures alongside other participant-level information. The source does not present the dataset as a single test or diagnostic procedure; it is a collection designed to address questions that require multiple kinds of data.

A major feature of KARI is its longitudinal scope. Participants undergo extensive imaging over time, providing repeated acquisitions rather than a one-time snapshot of the cohort. That structure is intended to help researchers investigate how different measurements relate within individuals and how imaging and biomarker findings may evolve across the disease spectrum. The 6,217 acquisitions encompass 1,645 participants spanning a 55-year age range, from ages 42 to 97, and include people representing healthy aging, preclinical Alzheimer disease and symptomatic disease.

Alongside raw imaging data, KARI provides quality-controlled processed outputs, including anatomical segmentations and biomarker quantification. These derivatives give researchers standardized outputs for the kind of measurements described in the resource, while the raw data preserve the opportunity to apply other analytical approaches. The dataset is also described as open access, with the stated aim of making the material available to the wider scientific community for work on pathophysiology, biomarker identification and therapeutic development. Access to a combined collection, however, does not itself demonstrate that any particular biomarker, diagnostic method or intervention will improve clinical outcomes.

KARI should therefore be understood as a data resource and an invitation to conduct further analyses, not as the report of a completed trial or a demonstrated clinical advance. Its value lies in the scale, multimodal design, longitudinal coverage and processed outputs, while interpretation of any resulting association will depend on the specific question, sample, analysis and limitations of the data used. By linking MRI, PET, clinical, cognitive, genetic and biomarker information across the Alzheimer disease spectrum, the Knight ADRC aims to give researchers a foundation for studies of early detection and disease progression. The resource does not by itself resolve those questions, but it makes them testable with a common and broadly accessible dataset.

Subject of Research: A multimodal neuroimaging and biomarker dataset spanning the Alzheimer disease spectrum

Article Title: Exploring healthy aging through preclinical Alzheimer disease in the Knight Alzheimer Research Imaging dataset

Article References: Hoagey, D. A., McKay, N. S., Joseph-Mathurin, N., Flores, S., Doering, S., Keefe, S. J., Tiemann Powles, S., Hornbeck, R. C., Smith, T. H., Scott, J., Chen, G., Massoumzadeh, P., LaMontagne, P. J., Wang, Q., Hassenstab, J., Ponisio, M. R., Denny, A., Balls-Berry, J. E., Snider, B. J., … Benzinger, T. L. S. (2026). Exploring healthy aging through preclinical Alzheimer disease in the Knight Alzheimer Research Imaging dataset. Nature Communications. https://doi.org/10.1038/s41467-026-78439-3

Image Credits: AI Generated

DOI: 10.1038/s41467-026-78439-3

Keywords: Alzheimer disease, brain imaging, MRI, PET, amyloid pathology, tau pathology, preclinical dementia, healthy aging, longitudinal imaging, biomarkers, Exploring, healthy

The systematic review by Ai and colleagues provides a timely synthesis of this rapidly evolving literature. By searching PubMed, Web of Science, and Scopus from database inception through 30 January 2026, the authors captured twenty-five original studies published between 2019 and 2026, a period that encompasses essentially the entire history of habitat analysis applied to head and neck tumours. This short time frame is itself informative: the field is young, and the pace of publication suggests accelerating interest. Yet the review also reveals important structural weaknesses in the evidence base. Only ten of the twenty-five studies included external validation datasets, meaning that the majority of reported models were evaluated solely on data drawn from the same institutions or patient cohorts used to develop them. This is a well-recognised threat to the generalisability of imaging prediction models, because internally validated performance frequently overstates what can be expected when a model is applied to patients from different scanners, protocols, or populations.

The dominance of nasopharyngeal carcinoma among the reviewed studies, with nine of the twenty-five articles devoted to it, reflects both the epidemiology of head and neck cancer globally and the research priorities of the regions where habitat imaging has been most actively pursued. Nasopharyngeal carcinoma is endemic in parts of southern China and Southeast Asia, and its deep anatomical location makes biopsy-based assessment of intratumoural heterogeneity challenging, which in turn creates a strong incentive for non-invasive imaging surrogates. Computed tomography was the most frequently used modality, appearing in nine studies, which is notable given that magnetic resonance imaging is often considered the mainstay of head and neck tumour assessment. The popularity of computed tomography in habitat work likely relates to its quantitative attenuation values, which lend themselves naturally to voxel-wise clustering, and to its widespread availability in routine staging workflows, making retrospective data collection straightforward.

Methodologically, K-means clustering was the predominant habitat technique, employed in twenty of the twenty-five studies. K-means is computationally simple, easy to implement, and produces readily interpretable partitions, which explains its popularity. However, its ubiquity also raises questions about whether the field has explored the full methodological space. Alternative clustering approaches, including Gaussian mixture models, fuzzy c-means, and hierarchical methods, may capture different aspects of tumour heterogeneity, and the optimal number of habitats to extract remains largely a matter of empirical choice rather than biological principle. The review’s observation that most studies selected a small number of subregions, typically two to four, is consistent with the practical need to balance model complexity against the risk of overfitting, but it also means that finer-grained biological variation may be lost in the coarser partitions that most analyses employ.

The finding that habitat imaging analysis performed consistently well, with areas under the curve exceeding 0.8, in predicting pathological and immunohistochemical biomarkers, tumour characterisation, and nodal status is encouraging. These are clinically meaningful tasks. Pathological and immunohistochemical markers, such as measures of proliferation, hypoxia, or immune infiltration, currently require tissue sampling, and non-invasive surrogates could in principle reduce the need for biopsy or guide more targeted sampling. Nodal status assessment is central to staging and treatment planning in head and neck cancer, and any imaging tool that improves on conventional size-based criteria could have real clinical impact. The fact that habitat-derived features outperformed conventional radiomics or clinical models in these domains suggests that explicitly modelling intratumoural spatial heterogeneity captures information that whole-tumour averaging discards.

In contrast, the variable results for treatment response prediction warrant careful interpretation. Predicting how an individual tumour will respond to radiotherapy, chemotherapy, or combined modalities is arguably the most valuable application imaginable for imaging-based heterogeneity analysis, because treatment de-escalation and escalation decisions in head and neck cancer carry significant trade-offs between tumour control and toxicity. The inconsistency of habitat models in this domain may reflect genuine biological complexity: treatment response is influenced by tumour factors, host factors, and treatment factors in ways that a static pretreatment imaging snapshot may not fully capture. It may also reflect methodological issues, including small sample sizes, heterogeneous endpoints such as pathological complete response versus progression-free survival, and differing time points at which imaging was performed relative to therapy.

The observation that combining multiple models only slightly improved performance in most studies is a sobering counterpoint to the prevailing enthusiasm for multimodal and multi-model integration. In many areas of imaging research, combining complementary data sources is assumed to yield additive or even synergistic gains. The review’s finding raises a practical question: if the incremental benefit of integrating habitat models with conventional radiomics or clinical predictors is marginal, does the added complexity, computational cost, and validation burden justify clinical adoption? This question is particularly pertinent in head and neck cancer, where clinical decision-making already integrates a rich array of clinical, pathological, and imaging information, and where any new tool must demonstrate clear added value over existing standards.

The heterogeneity of the current evidence base, which the authors explicitly highlight, deserves emphasis. The twenty-five included studies varied in tumour origin, imaging modality, clustering features, habitat methods, number of subregions, and clinical applications. Such heterogeneity makes it difficult to draw firm conclusions about which habitat approaches work best, for which tumour types, and for which clinical questions. It also complicates meta-analysis, since pooling results across studies with fundamentally different designs risks producing misleading summary estimates. The Radiomics Quality Score, referenced in the review’s abbreviations, is one tool that has been developed to assess the methodological rigour of imaging studies, and its application in this context underscores the field’s growing awareness of the need for standardisation.

Looking forward, the review’s call for prospectively conducted and externally validated studies is both appropriate and necessary. Prospective designs would allow standardised imaging protocols, predefined endpoints, and sample sizes powered to detect clinically meaningful effects, all of which are difficult to achieve retrospectively. External validation across multiple institutions, ideally with diverse patient populations and scanner types, would establish whether the promising internal performance of habitat models translates to real-world settings. Standardisation of habitat methodology, including consensus on clustering algorithms, feature selection, and subregion numbering, would further strengthen the evidence base and facilitate meaningful comparison across studies.

The open-access publication of this review under a Creative Commons Attribution licence ensures that its findings are freely available to clinicians, researchers, and patients worldwide, which is particularly valuable for a field in its infancy where broad access to methodological critiques can shape research priorities. The involvement of authors from both the Chinese University of Hong Kong and the University of Hong Kong reflects the regional concentration of head and neck cancer expertise and imaging research capacity, and their collaboration across radiology and oral and maxillofacial radiology disciplines illustrates the multidisciplinary nature of this work. As habitat imaging analysis matures, the transition from retrospective proof-of-concept studies to prospective, multicentre validation trials will determine whether the consistent performance observed in biomarker prediction and nodal assessment can be harnessed to improve outcomes for patients with head and neck cancer, and whether the more uncertain domain of treatment response prediction can be refined to the point where it informs individualised therapeutic decisions.

Subject of Research: Habitat imaging analysis for head and neck cancer

Article Title: Current evidence in application of habitat imaging analysis in head and neck cancer: a systematic review

Article References: Ai, Q. Y. H., Cai, Z., Leung, H. S., Kwok, H. M., So, T. Y., & Hung, K. F. (2026). Current evidence in application of habitat imaging analysis in head and neck cancer: a systematic review. BMC Medical Imaging. https://doi.org/10.1186/s12880-026-02777-3

Image Credits: AI Generated

DOI: 10.1186/s12880-026-02777-3

Keywords: habitat imaging, head and neck cancer, radiomics, nasopharyngeal carcinoma, machine learning, CT imaging, MRI, PET, K-means clustering, systematic review, tumour heterogeneity, biomarker prediction

The paper, led by Joaquim Radua of IDIBAPS in Barcelona together with Lydia Fortea and an international team spanning Spain, Germany, Italy, the United Kingdom, Sweden and South Korea, introduces the Qual-CBMA checklist: fifteen concrete criteria designed to make the methodological rigor of a coordinate-based meta-analysis visible, reviewable and discussable. The checklist grew out of a Delphi-style iterative process in which the developers of the field’s most widely used CBMA methods drafted an initial proposal and then refined it through two structured rounds of feedback among all authors. Its launch matters because the field’s previous benchmark, the influential Ten Simple Rules for neuroimaging meta-analysis published in 2018, has been cited far more often than it has been followed.

The evidence for that claim is sobering. A 2023 citation analysis found that nearly 8 percent of quotations of the Ten Simple Rules contained errors, including misinterpretation, oversimplification or outright unsupported claims. That pattern is not unique to one guideline. A separate analysis showed that a widely cited report promoting uncorrected statistical thresholds in neuroimaging was routinely invoked to justify their use, even though most of the citing studies deviated substantially from the recommended parameters. And in a now-classic irony of evidence synthesis, a commentary that explicitly criticized the Newcastle–Ottawa scale for assessing observational studies was frequently cited as if it endorsed the scale.

The problem is compounded by the fact that general reporting standards fit CBMAs poorly. PRISMA, the default checklist for systematic reviews, requires items such as a per-study effect size estimate that coordinate-based methods typically cannot supply, while ignoring procedures central to CBMAs, such as transforming all data into a common stereotactic space, ensuring consistent brain coverage across datasets and keeping statistical thresholds uniform across brain regions. When the original guidelines did include a checklist, it was optional, several items demanded open-ended answers in supplementary files without corresponding justifications, and the criteria for compliance were often unclear. The result, an exploratory search by the team suggested, was highly heterogeneous reporting of how specific recommendations had been addressed.

Crucially, the authors emphasize that inconsistent reporting does not mean the underlying work was done badly. A first-ever meta-analysis of a rare condition may legitimately include fewer datasets than the guidelines recommend and still be scientifically valuable. Qual-CBMA is therefore designed as an aid to transparent reporting, not a tribunal. Authors mark each of the fifteen criteria as met, not met or not applicable, and must comment on every response; when a criterion is not met, the explanation must also appear in the manuscript. A high-quality CBMA, in this framework, is not one that ticks every box, but one that clearly explains which boxes it missed and why those omissions are acceptable given the study’s objectives. A criterion marked No is not a negative judgment but a prompt for context.

The fifteen criteria operationalize the earlier ten rules by splitting them into sharper, more testable items. The rule on specifying the research question becomes three criteria covering imaging modality and measure, functional paradigm, and condition or disorder. The rule on data collection becomes two criteria on keyword-based searches across multiple databases and explicit inclusion and exclusion criteria. The rules on spatial standardization become three criteria: statistical thresholds independent of brain region, consistent brain or tissue coverage, and conversion of all coordinates into a common reference space such as MNI. Reporting transparency is split into detailed description of datasets, contrasts and samples, and public availability of the data and code needed for replication, with a manuscript stating only that data are available upon reasonable request falling short of the mark.

Several technical criteria deserve special attention. The checklist asks whether the protocol was preregistered in a public database such as PROSPERO or the Open Science Framework, while explicitly acknowledging that CBMAs require iterative refinement and that deviations from a preregistered plan are expected rather than problematic. It requires that each included study applied a uniform statistical threshold across the whole brain, so that peaks obtained with liberal region-of-interest or small-volume corrections are only usable if they also meet the whole-brain threshold. This prevents results biased toward overrepresented regions and preserves the assumption that every voxel has an a priori equal chance of showing an effect. Double extraction of data by at least two independent researchers, with agreement ideally reported, is also assessed, along with adjustment for multiple general linear model contrasts derived from overlapping samples.

Perhaps the most consequential criterion concerns statistical power. Drawing on large-scale simulations of the activation likelihood estimation method, the authors set a floor of 17 independent peak-coordinate datasets, rising to 23 for structural MRI studies. They also offer an optional weighting scheme: each study counts in proportion to its statistical power, so a one-sample study with 20 participants counts as one dataset, a study with only 12 to 13 participants counts as roughly half, and no study counts for more than 1.25 regardless of size. Studies providing full statistical maps count double, a deliberately conservative correction derived from simulations showing that maps carry far more information than peak coordinates alone, roughly fourfold on average though strongly dependent on effect size. These thresholds echo the optimal information size concept from the GRADE framework for intervention meta-analyses.

Notably, the checklist refuses to grade statistical significance. The authors deliberately excluded it to avoid incentivizing p-hacking, the practice of trying multiple meta-analytic methods or parameter settings until significant results emerge. Instead, they argue, the community should learn to accept large-scale CBMAs that yield no findings after correction for multiple comparisons, and they recommend presenting results across a range of thresholds rather than a single cut-off, noting that differences between analytical approaches in the Nature reproducibility study were less pronounced when compared across multiple thresholds. They also concede the apparent unfairness of requiring even a CBMA with 14 highly significant studies to acknowledge having fewer than 20 datasets, countering that small samples allow random imbalances in confounding variables to drive neuroimaging results, whereas large samples wash them out.

The authors are candid about the checklist’s limitations. Binary criteria and fixed thresholds inevitably create artificial distinctions, such as the meaningless boundary between 16 and 17 datasets, and in areas with very large effect sizes even small studies may detect effects consistently, making the thresholds less informative. The tool is also algorithm-agnostic and cannot capture method-specific optimal practices, and future iterations may add graded criteria or algorithm-specific extensions. The payoff, however, could extend well beyond individual papers: clearly defined criteria with explicit comments enable structured extraction of methodological features across hundreds of CBMAs, opening the door to large-scale meta-research and even automated reviews. The team hopes that authors, reviewers and editors will adopt Qual-CBMA as a shared standard, submitted alongside new meta-analyses, so that transparency in neuroimaging evidence synthesis becomes the norm rather than the exception. The checklist is publicly available, with the R code for the power simulations provided in the article’s supplementary material.

Subject of Research: Development of the Qual-CBMA checklist for evaluating the methodological quality of coordinate-based meta-analyses in neuroimaging

Article Title: Evaluating the methodological quality of coordinate-based meta-analyses: the qual-CBMA checklist

Article References: Radua, J., Fortea, L., Albajes-Eizagirre, A., Eickhoff, C., Fullana, M. A., Fusar-Poli, P., Mataix-Cols, D., Shin, J. I., Solanes, A., Tahmasian, M., Vieta, E., Vilajosana, E., Eickhoff, S. B., & Müller, V. I. (2026). Evaluating the methodological quality of coordinate-based meta-analyses: the qual-CBMA checklist. Frontiers in Neuroimaging, 5, Article 1753543. https://doi.org/10.3389/fnimg.2026.1753543

Image Credits: AI Generated

DOI: 10.3389/fnimg.2026.1753543

Keywords: coordinate-based meta-analysis, neuroimaging, checklist, methodological quality, CBMA, MRI, PET, preregistration, reproducibility, statistical power, publication bias, Ten Simple Rules

Researchers from the University of Hail, Imam Mohammad Ibn Saud Islamic University, Dar Al-Hekma University, and Aswan University tackled a problem that has long frustrated urban designers and landscape architects in hot-arid regions: while urban vegetation is widely recognized as a passive cooling strategy, its actual performance depends on a complex interplay between the tree’s physical characteristics and the surrounding built environment. A tree that provides exceptional shade in one courtyard orientation may perform far less effectively in another. A species that delivers dramatic peak-hour cooling may not maintain that benefit throughout the school day. Conventional assessments that rely on a single thermal variable, such as air temperature reduction, simply cannot capture this multidimensional nature of vegetation performance.

To address this gap, the research team developed what they call a Multi-Criteria Thermal Performance Framework, or MCTPF. The framework integrates high-resolution microclimatic simulations using ENVI-met software with physiological equivalent temperature, or PET, calculations to evaluate vegetation performance across three complementary dimensions. The first dimension, quantified through the Tree Cooling Efficiency Index, measures the magnitude of integrated thermal improvement relative to a non-vegetated baseline. The second, captured by the Thermal Resilience Index, assesses whether a vegetation scenario reduces temporal variability in thermal conditions throughout the assessment period. The third, expressed through the Orientation Sensitivity Index, quantifies how much cooling effectiveness changes when the same vegetation configuration is evaluated under contrasting courtyard orientations. These three indices are subsequently integrated through a Composite Thermal Performance Index after normalization to a common performance scale.

The study focused on two semi-enclosed courtyards at the Egyptian Japanese School, each with an identical footprint of 200 square meters but with critically different orientations. The primary courtyard opens toward the east, while the secondary faces west. This asymmetrical configuration creates fundamentally different patterns of solar radiation absorption and wind interaction throughout the school day, establishing diverse thermal microenvironments within the same building complex. The researchers established a non-vegetated baseline scenario for each courtyard and then modeled three alternative vegetation scenarios by introducing individual tree species: Cassia leptophylla, commonly known as the gold medallion tree; Ficus nitida, or the Indian laurel; and Cassia nodosa, the pink shower tree. Each species was represented in the ENVI-met model using its actual morphological characteristics, including tree height, canopy diameter, leaf area index, and canopy porosity, rather than imposing an artificial standardized geometry.

The simulation results revealed substantial differences among the modeled vegetation configurations and between the two courtyard orientations. Ficus nitida consistently produced the greatest cooling response across all three thermal variables. In the east-facing courtyard, this species reduced air temperature by approximately 5.02 percent, mean radiant temperature by 37.88 percent, and PET by 25.39 percent relative to the baseline. In the west-facing courtyard, the corresponding improvements were 4.57 percent, 25.48 percent, and 15.35 percent respectively. The substantially larger reductions in mean radiant temperature compared to air temperature demonstrate that the principal cooling contribution of these tree scenarios was associated with modification of the radiative environment through shading and interception of solar radiation rather than with large reductions in ambient air temperature.

However, the story became considerably more nuanced when the researchers evaluated performance across all three dimensions simultaneously. While Ficus nitida achieved the highest Tree Cooling Efficiency Index values in both courtyards, with scores of 0.251 in the east-facing courtyard and 0.162 in the west-facing courtyard, it also exhibited the greatest sensitivity to courtyard orientation, with an Orientation Sensitivity Index of 0.428. In contrast, Cassia leptophylla demonstrated the greatest orientation robustness, with the lowest OSI value of 0.272, despite producing the lowest cooling efficiency among the three species. Cassia nodosa occupied an intermediate position, combining moderate cooling performance with the highest Thermal Resilience Index in the east-facing courtyard, indicating superior temporal stabilization of thermal conditions under that orientation.

The Composite Thermal Performance Index revealed clear trade-offs among cooling magnitude, temporal stability, and orientation robustness. Ficus nitida in the east-facing courtyard achieved the highest overall CTPI value of 74.75, reflecting its dominant cooling efficiency and near-maximum temporal resilience. However, its high orientation sensitivity resulted in the lowest orientation-robustness score, which limited its composite performance under the less favorable west-facing courtyard condition, where its CTPI dropped to just 21.06. Conversely, Cassia leptophylla, despite its lower cooling efficiency, achieved the highest CTPI in the west-facing courtyard at 25.00, benefiting substantially from its superior orientation robustness. The ranking of vegetation configurations changed between the two courtyard orientations, confirming that the same tree species can exhibit fundamentally different integrated thermal performance depending on its spatial context.

The researchers conducted a sensitivity analysis to examine whether alternative weighting structures would alter the comparative ranking. Under the baseline structure, which assigned the greatest weight to cooling efficiency, Ficus nitida in the east-facing courtyard remained the preferred alternative. However, when orientation robustness received substantial importance, Cassia leptophylla became the preferred configuration. This finding demonstrates that the framework does not impose a universally optimal vegetation choice but rather provides a transparent mechanism for adapting tree-selection decisions to different thermal-design priorities while preserving the spatial influence of courtyard orientation.

Model validation showed strong agreement between simulated and measured air temperature patterns, with a coefficient of determination of 0.996 and a mean absolute error of 1.505 degrees Celsius. Relative humidity demonstrated even stronger agreement, with a coefficient of determination of 0.9998. However, the researchers acknowledge that the validation was based on baseline measurements at a single location over one 24-hour period and did not directly validate vegetation-induced thermal responses. Mean radiant temperature, wind conditions, PET, and the magnitude of vegetation cooling were not independently validated against field measurements, meaning the findings should be interpreted primarily as comparative simulation-based evidence rather than fully validated predictions of absolute thermal conditions.

The implications of this research extend well beyond a single school in Aswan. As cities across the Middle East, North Africa, and other hot-arid regions grapple with intensifying urban heat, the framework offers a transparent and reproducible pathway from microclimatic simulation to climate-responsive design selection. Rather than asking simply which tree provides the greatest temperature reduction, designers can now evaluate vegetation strategies according to clearly defined priorities: whether maximum cooling efficiency matters most, whether temporal stability throughout the occupancy period is the primary concern, or whether robustness across different spatial orientations should dominate the decision. The framework makes these trade-offs explicit, allowing vegetation selection to respond to specific courtyard conditions and design objectives rather than relying on single-variable comparisons that may mask critical performance differences. The researchers emphasize that future studies should validate vegetation-induced thermal effects through field measurements across broader environmental contexts, examine multiple seasons and climatic regions, and incorporate age-specific physiological parameters to better represent the children and adolescents who constitute the primary users of educational courtyards.

Subject of Research: Tree selection, courtyard form, and outdoor microclimatic comfort in hot-arid educational spaces: a multi-criteria decision framework

Article Title: Tree selection, courtyard form, and outdoor microclimatic comfort in hot-arid educational spaces: a multi-criteria decision framework

Article References: Aldersoni, A., Abdallah, A. S. H., Abdelhafez, M. H. H., Elkhayat, K., Gomaa, M. M., & Ragab, A. (2026). Tree selection, courtyard form, and outdoor microclimatic comfort in hot-arid educational spaces: a multi-criteria decision framework. Frontiers in Built Environment, 12, Article 1956270. https://doi.org/10.3389/fbuil.2026.1956270

Image Credits: AI Generated

DOI: 10.3389/fbuil.2026.1956270

Keywords: ENVI-met, PET, TCEI, TRI, OSI, CTPI, courtyard orientation, hot-arid climate, school courtyards, tree selection, thermal comfort, urban vegetation

The review, led by Emily Gallmetzer and Kejal Kantarci, followed PRISMA guidelines and screened 461 records from PubMed, ultimately including 18 peer-reviewed studies assessed with the Newcastle-Ottawa Scale. Fourteen were rated good quality, two fair, and two poor. The imaging modalities spanned the full modern neuroimaging toolkit: 16 studies used in vivo structural magnetic resonance imaging, four used diffusion tensor imaging, two used functional MRI, and three used positron emission tomography. Study designs ranged from massive population cohorts such as the UK Biobank, with more than 13,000 imaged participants, to small case-control studies of women recruited from familial cancer clinics.

Definitions matter enormously in this literature, and the review highlights how inconsistent they have been. Premature menopause is defined as onset before age 40, and early menopause as onset between ages 40 and 44. Both can arise spontaneously or be induced by chemotherapy, pelvic radiation, or premenopausal bilateral oophorectomy, the surgical removal of both ovaries before natural menopause. By contrast, hysterectomy with ovarian conservation stops menstruation but preserves endocrine function and is not a cause of ovarian hormone loss. The hallmark consequence of spontaneous menopause is the gradual decline of estradiol, the predominant and most biologically active ovarian estrogen during reproductive life, whereas oophorectomy causes abrupt and complete loss of estradiol and all other ovarian hormones.

Estrogen is far more than a reproductive hormone. In the brain it has important antioxidative effects, contributes to the suppression of immune activation led by astrocytes and microglia, and supports cerebral energy metabolism by upregulating glucose transporters. Early estrogen loss has been associated in experimental studies with increased oxidative stress, accelerated neuroinflammatory processes, and dysregulated iron homeostasis within the central nervous system. Loss of estrogenic support may increase vulnerability to neuronal injury, gliosis, and accumulation of pathological proteins. Because ovarian hormones act on multiple organ systems, their early loss has systemic consequences that extend well beyond reproductive health, and this accelerated somatic aging may itself exacerbate brain aging.

Across the structural MRI studies, the review found regionally specific rather than global patterns of brain differences. Women with premature or early menopause showed lower gray-matter volume or cortical thickness in medial temporal, frontal, and basal forebrain regions. In population-based cohorts, women who underwent oophorectomy before age 50 showed smaller amygdala volume and thinner parahippocampal-entorhinal cortex compared with age-matched controls. A more recent Mayo Clinic study found that oophorectomy before age 46 was associated with thinner entorhinal cortex independent of age at imaging, whereas lower temporal-lobe cortical thickness emerged only at older ages. In women with idiopathic premature ovarian insufficiency, voxel-based analysis revealed regional reductions in the bilateral olfactory cortex and right parahippocampal gyrus, together with reduced cortical thickness in parietal and insular regions and smaller hippocampal subfield and thalamic volumes. Notably, no significant global gray-matter difference was observed in the full sample, and younger participants showed exploratory global differences that were uncorrected for multiple comparisons.

Diffusion tensor imaging added a complementary microstructural picture. This technique quantifies axonal architecture through metrics such as fractional anisotropy, which measures the directionality of water diffusion, and mean diffusivity, which serves as an inverse index of membrane density and tissue integrity. Reduced fractional anisotropy, often accompanied by increased mean diffusivity, indicates loss of axonal integrity and myelination. Women who underwent oophorectomy before age 40 showed lower fractional anisotropy and higher mean diffusivity in the anterior corona radiata, the genu of the corpus callosum, and the inferior fronto-occipital fasciculus. Strikingly, women whose oophorectomy occurred between ages 40 and 44 showed no such differences, a pattern consistent with a threshold or dose-response effect of earlier hormone deprivation. In spontaneous premature menopause, altered structural connectivity involved the olfactory cortex, amygdala, hippocampus, and thalamus, while global network topology remained relatively preserved.

Functional MRI, though available in only two studies from the same Canadian research group, provided convergent functional correlates. Women who underwent bilateral salpingo-oophorectomy without hormone therapy showed lower right posterior lateral hippocampal activation during novel face-name encoding compared with those who received hormone therapy. A later study from the same group found lower posterior hippocampal activation in both treated and untreated groups compared with age-matched controls, though it did not fully replicate the earlier group-specific pattern, and anterior hippocampal activation and raw task accuracy did not differ across groups. The convergence with structural findings is only partial, and independent replication in separate cohorts is needed before firm conclusions can be drawn.

The PET evidence, though limited to three studies, offers the most provocative molecular findings. In cognitively unimpaired women with elevated amyloid-beta, earlier age at menopause was associated with higher regional tau burden, with effects clustering in medial temporal, lateral temporal, parietal, and occipito-temporal regions, a distribution consistent with Braak staging of early-to-intermediate Alzheimer’s tau pathology. In the Mayo Clinic cohort, women who underwent oophorectomy before age 46 showed higher amyloid-beta load when imaged at older ages, and tau was elevated only at higher amyloid-beta levels. Women whose oophorectomy occurred between ages 46 and 49 showed no comparable biomarker differences. These findings suggest that a woman’s age at the time of menopause, rather than menopause type alone, may be the most sensitive determinant of later molecular vulnerability.

Mechanistically, the review proposes that early loss of ovarian hormones may weaken the brain’s compensatory reserves and converge on interconnected deleterious molecular pathways. Reduced estrogen levels may diminish neurotrophic support, leading to loss of synaptic density within the medial temporal lobe and reduced neuronal resilience to age-related stressors. Estrogen normally supports cerebral energy metabolism, and its loss may contribute to mitochondrial dysfunction and impaired cellular bioenergetics, resulting in a hypometabolic state where the brain metabolizes its own myelin-derived lipids for energy. This metabolic shift may increase production of reactive oxygen species, including superoxide and hydrogen peroxide, which can damage neuronal membranes and further destabilize mitochondrial function. Estrogen also regulates enzymes involved in amyloid-beta degradation, such as the insulin-degrading enzyme, supports autophagic flux, and helps maintain microglia in a homeostatic state, so its depletion may shift the brain toward chronic neuroinflammation and failed protein clearance.

The authors are careful to emphasize that the evidence remains heterogeneous and largely cross-sectional, limiting causal inference about the temporal sequence between ovarian hormone loss and neurobiological change. Sixteen of the 18 included studies used age cutoffs that reached into the usual-age menopause range of 45 to 54 years, meaning many reported associations cannot be tied specifically to premature or early menopause. Large volunteer cohorts such as the UK Biobank tend to recruit healthier participants than the general population, and almost all studies imaged women only once. The review’s conclusions are therefore best seen as a rationale for prevention-oriented research rather than established clinical guidance. Still, the convergent pattern across imaging modalities, spanning gray-matter atrophy, white-matter microstructural deficits, reduced hippocampal activation, and preliminary evidence of altered amyloid and tau burden, suggests that women’s reproductive history deserves consideration as an integral part of dementia risk profiles. Large-scale longitudinal studies tracking the same women with serial neuroimaging over years, and evaluating interactions with genetic predisposition such as the apolipoprotein E epsilon 4 genotype, are now needed to determine when brain alterations begin in relation to the drop in estrogen levels and whether specific critical windows exist during which hormonal interventions could decelerate degeneration in networks vulnerable to Alzheimer’s disease.

Subject of Research: Neuroscience

Article Title: Premature and early menopause and neuroimaging biomarkers of neurodegenerative diseases: a systematic review

Article References: Gallmetzer, E., Mak, E., Kara, F., Zeydan, B., Kapoor, E., Shufelt, C. L., Rocca, W. A., & Kantarci, K. (2026). Premature and early menopause and neuroimaging biomarkers of neurodegenerative diseases: a systematic review. Frontiers in Dementia, 5, Article 1912295. https://doi.org/10.3389/frdem.2026.1912295

Image Credits: AI Generated

DOI: 10.3389/frdem.2026.1912295

Keywords: menopause, Alzheimer's disease, neuroimaging, estrogen, dementia, women's health, MRI, PET, Premature, early, biomarkers, neurodegenerative

The study, published in Nature Machine Intelligence by Luyang Cao, Zhijian Wu and colleagues at Nanjing University, Westlake University, Southeast University and the Shenzhen Institute of Advanced Technology, was trained and evaluated on an extraordinary corpus: 3.76 million paired images, in which each degraded clinical image is matched with a high-quality reference. The data were assembled from publicly available repositories including The Cancer Imaging Archive, OpenNeuro, NITRC, Grand Challenge, Synapse, Kaggle and GitHub, and span 34 MRI datasets covering brain, cardiac, abdominal and prostate imaging, 37 CT datasets ranging from thoracic to whole-body acquisitions, six X-ray datasets, five endoscopy datasets, four fundus photography datasets, two ultrasound datasets and one whole-body PET dataset for oncological imaging.

Architecturally, AMIFound follows a philosophy that has transformed natural-language processing but has been slower to arrive in medical image restoration: a shared backbone with adaptive, task-specific routing. A single encoder first maps any low-quality input, regardless of modality, into a common latent feature space that captures both modality cues and the nature of the degradation. A dynamic routing mechanism then selects from a pool of specialised expert modules to construct a restoration pathway tailored to each individual image. The expert pool contains fast Fourier transform attention experts that operate in the frequency domain, spatial attention experts, convolutional local-feature experts and hybrid FFT-CNN fusion experts, allowing the model to combine complementary frequency, spatial and structural information on the fly.

Training such a system required a way to generate realistic degradation across modalities. The researchers built a modality-aware degradation engine that transforms high-quality images into degraded observations using physics-informed simulations. For CT, metal artefacts are synthesised in the projection domain through random metal masks, Radon transformation, nonlinear artefact mapping and inverse reconstruction. MRI aliasing and low-resolution effects are produced by k-space truncation followed by inverse Fourier reconstruction, while PET low-dose degradation is simulated with Poisson noise, Gaussian noise and blur. Endoscopic low-light conditions are modelled by gamma correction and brightness reduction in HSV colour space, X-ray blur by cascaded filtering with anisotropic diffusion, ultrasound artefacts by reverberation echoes, geometric shadowing and depth attenuation, and fundus artefacts by random circular highlights and Gaussian softening. In total, the model handles seven representative degradation patterns.

The benchmark results are striking. Across all seven modalities, AMIFound outperformed 38 competing baselines, including both task-specific restoration networks and generalist models such as AMIR, MoCE-IR and Found-IR, on full-reference image quality metrics such as structural similarity and learned perceptual measures. Radar plots of structural similarity improvements show consistent gains over each competing framework, and detailed per-task comparisons show the model achieving the best or near-best scores against modality-specific methods for every restoration task, from CT metal artefact reduction to PET denoising and cataract-affected fundus enhancement.

Quantitative metrics alone do not convince radiologists, so the team also ran a subjective evaluation with clinicians across all seven modalities. Physicians rated the restored images on sharpness, artefact suppression, naturalness and diagnostic confidence. AMIFound achieved the highest or near-highest average scores in every modality and provided greater diagnostic confidence than competing methods. The model was also stress-tested under combined degradations, where multiple corruptions strike the same image simultaneously, and on naturally degraded clinical images rather than synthetic ones, maintaining its advantage in both settings and preserving anatomical structures and realistic intensity distributions where competitors tended to oversmooth or distort contrast.

Perhaps the most clinically consequential finding is what happens downstream. The researchers fed AMIFound-enhanced images into 44 disease detection tasks and found consistent improvements over detection on the original degraded images. The benefits extended to tumour stage prediction in non-small-cell lung cancer and to survival risk stratification, including improved individual-level prognostic ranking. Attention heatmaps from the CT-based tumour-stage classifier reveal a plausible mechanism: before enhancement, model attention is often diffuse or mislocalized and predictions are less certain, whereas after enhancement, attention concentrates on tumours, lung-parenchyma abnormalities and lesion boundaries, and the highest prediction probability more often agrees with the ground-truth stage.

The significance of this work lies partly in its economics. Hospitals generate enormous volumes of routine imaging that falls short of research-grade quality, and low-dose protocols, deliberately chosen to reduce radiation exposure, or motion artefacts from patients who cannot hold still, are unavoidable facts of clinical life. Previous approaches required a separate, carefully tuned network for each modality-degradation combination, an unsustainable proliferation as imaging technology diversifies. A single foundation model that generalises across modalities and degradations, trained once and deployed everywhere, could lower the barrier for smaller institutions and make image enhancement a routine preprocessing step for any AI diagnostic pipeline.

The authors have released their training and evaluation code publicly on GitHub, with a versioned archive on Zenodo, built on PyTorch and drawing on data-loader pipelines from prior all-in-one restoration frameworks such as PromptIR, AirNet and MoCE-IR. All datasets used were publicly available and accessed under their respective licences, and the team reports that the model preserves clinically relevant information rather than hallucinating plausible-looking detail, a critical concern for any generative approach in medicine. Funded by the National Natural Science Foundation of China and several regional programmes, the work positions AMIFound as a domain-specific foundation model for the unglamorous but essential task of making everyday clinical images good enough, not just for human eyes, but for the growing fleet of AI systems that depend on them.

Subject of Research: A multimodal foundation model for restoring and enhancing degraded medical images across seven imaging modalities

Article Title: An all-in-one foundation model for multimodal medical image restoration and enhancement

Article References: Cao, L., Wu, Z., Zhang, J., Qi, L., Gao, Y., Zheng, H., Zheng, Y., & Shi, Y. (2026). An all-in-one foundation model for multimodal medical image restoration and enhancement. Nature Machine Intelligence. https://doi.org/10.1038/s42256-026-01310-z

Image Credits: AI Generated

DOI: 10.1038/s42256-026-01310-z

Keywords: medical imaging, foundation model, image restoration, deep learning, AMIFound, CT, MRI, PET, ultrasound, endoscopy, lung cancer, diagnostic AI

The scale of the problem the team set out to tackle is enormous. Microplastics, defined as plastic particles smaller than 5 millimeters, were first named by researchers in 2004, though most particles in the environment are actually under 500 micrometers. They come from at least 14 different polymers and appear as fibers, fragments, and films. Laundry effluent is one of the richest sources, since synthetic fabrics shed fibers with every wash; other major contributors include plastic recycling wastewater, landfill leachate, and sewage sludge. Because microplastics carry chemicals from manufacturing and can also adsorb pollutants from surrounding water, their ingestion threatens wildlife and human health through combined physical and chemical toxicity. Yet the field remains exploratory, with few standards, methods, or policies governing detection, removal, and discharge.

Among available treatment options, adsorption has drawn growing attention because it is operationally simple, energy-efficient, and effective against a wide range of contaminants. MOFs, a class of crystalline materials built from metal ions connected by organic linkers, are natural candidates for the job. Their exceptional porosity and vast internal surface areas give them strong selectivity toward pollutants. In water, most microplastics acquire negative surface charges, so MOFs containing positively charged ions can capture them through strong electrostatic attraction. But earlier MOF-based removal studies have been disappointing, achieving removal efficiencies of only around 32.6 percent in one case and 76.5 percent in another. A 2025 review of more than 65 studies concluded that the field needs better synthesis efficiency, scalable applications, and environmental assessment.

The core obstacle is physical form. Pristine MOF powders exist as fine, discrete particles that lack structural stability and are difficult to deploy in continuous water treatment systems. Loose powder would clog columns, wash away, or resist handling entirely. The Sydney-led team’s solution was to embed the MOF inside a hydrogel, a three-dimensional polymer network that is porous, water-permeable, and mechanically robust. By combining the high surface activity of MOFs with the structural integrity and processability of a polymeric scaffold, the composite turns a laboratory curiosity into something that can sit at the bottom of a filtration column and treat wastewater continuously.

The fabrication process itself is deliberately simple. The researchers first synthesized two batches of a zeolitic imidazolate framework called ZIF-8 by mixing zinc nitrate with the organic linker 2-methylimidazole at different concentrations, using a conventional solvothermal method. X-ray diffraction analysis showed that the first batch, ZIF-8.1, had sharper diffraction peaks and a narrower full width at half maximum than its counterpart, indicating roughly 44 percent higher peak intensity and superior crystallinity. Nitrogen adsorption measurements confirmed classic microporous behavior, with a steep uptake at very low relative pressures and no hysteresis loop, and a surface area of about 496 square meters per gram. ZIF-8.1 was therefore selected for incorporation into the hydrogel.

To build the composite, the team dispersed 6 milligrams of ZIF-8 in water by ultrasonication, then added poly(ethylene glycol) diacrylate, a photoinitiator, and a solution of polyvinyl alcohol. The mixture was cast into a mold and exposed to ultraviolet light for 10 minutes to trigger photopolymerization, then frozen overnight and chemically cross-linked with glutaraldehyde. After rinsing and freeze-drying, the result was a disc-shaped hydrogel filter studded with ZIF-8 nanocrystals. Infrared spectroscopy confirmed the successful integration, showing new peaks corresponding to ZIF-8’s imidazole ring vibrations alongside the hydrogel’s own carbonyl and ether signals. Scanning electron microscopy revealed small crystalline particles distributed across the hydrogel skeleton without disrupting its intrinsic pore structure.

Performance testing used a low-pressure filtration column with a 30-millimeter-diameter hydrogel disc, 2.5 millimeters thick, operated at a flow rate of 1.2 liters per hour. The test water was genuine laundry wastewater collected from a residential washing machine running synthetic-fiber loads, spiked with PE and PET particles to reach 100 milligrams per liter of each polymer. The raw effluent itself contained 9,000 to 11,000 microplastic particles per liter, with particles ranging from the nanometer scale up to 200 micrometers. The bare hydrogel managed only 65 percent PET and 52 percent PE removal, hampered by its negatively charged surface, which repels similarly charged plastics. Adding 3 milligrams of ZIF-8 lifted those figures to 90 and 81 percent, and at 6 milligrams the composite reached its peak of 97 percent for PET and 98 percent for PE. At 9 milligrams, however, particle aggregation clogged the pores and wastewater could no longer pass through at all, revealing a delicate balance in MOF loading.

The mechanism behind the capture is a layered one. Zeta potential measurements showed the ZIF-8 particles carry strongly positive charges of about 26.8 millivolts, while PET, PE, and PVC particles all carry negative charges between roughly minus 22 and minus 25 millivolts, setting up powerful electrostatic attraction. Infrared analysis after filtration suggested that polar PET adsorbs through hydrogen bonding and dipole–dipole interactions between its ester groups and ZIF-8’s nitrogen-rich moieties, combined with confinement in pores around 1.16 nanometers wide. Nonpolar PE, lacking such chemical handles, is retained mainly through van der Waals forces, hydrophobic interactions, and mechanical entrapment. The interpenetrating hydrogel network adds physical sieving on top of these surface interactions, meaning removal involves filtration and particle–surface attraction working together rather than adsorption alone.

Durability and cost are where the material becomes genuinely compelling. After six filtration–regeneration cycles, PET removal had fallen only slightly, from 97 percent to 96 percent, and electron microscopy showed the composite’s morphology largely preserved after washing, with the ZIF-8 nanocrystals still anchored in place. The filter operated stably for roughly four hours per cycle, and pretreatment combined with backwashing dramatically improved long-term performance: without pretreatment, water flux dropped 21.7 percent over a 300-minute run, but with pretreatment the decline was just 5.8 percent, a stark contrast to the 9 to 54 percent flux losses reported for conventional membrane systems treating microplastic-laden municipal wastewater. The hydrogel can be freeze-dried, stored at ambient conditions without special preservation, and cast in different dimensions to fit various column configurations.

The economics look promising too. Assuming a year of continuous operation at a filtration rate of 1.7 meters per hour, with each filter replaced after six four-hour cycles, the team estimated a total annual cost of approximately 3.1 Australian dollars, or about 0.3 Australian dollars per cubic meter of laundry wastewater treated, covering only chemical and energy inputs for fabrication. That figure positions the MOF-derived hydrogel as a potentially sustainable and affordable alternative to existing technologies. The authors caution that more work is needed on long-term durability, scalability, and techno-economic feasibility at industrial scale, and they propose using X-ray photoelectron spectroscopy and Raman spectroscopy to further verify the removal mechanisms, along with testing how pH, temperature, and influent concentration affect performance. But as microplastic pollution accelerates worldwide, a cheap, reusable, sunlight-cured sponge that catches nearly every plastic particle passing through it offers a rare piece of genuinely good environmental news.

Subject of Research: Development of a ZIF-8 metal–organic framework hydrogel for removing microplastics from laundry wastewater

Article Title: Synthesis of metal organic framework-based hydrogel for microplastics removal from aquatic environments

Article References: Luu, T. T., Mao, S., Truong, D. Q., Shi, Y., Duong, T. N. O., Fu, Q., Do, V. M., Vigneswaran, S., & Nguyen, T. V. (2026). Synthesis of metal organic framework-based hydrogel for microplastics removal from aquatic environments. Environmental Science and Pollution Research, 33(28), 14165-14177. https://doi.org/10.1007/s11356-026-38156-2

Image Credits: AI Generated

DOI: 10.1007/s11356-026-38156-2

Keywords: microplastics, metal-organic frameworks, ZIF-8, hydrogel, laundry wastewater, water treatment, adsorption, filtration, PET, polyethylene, electrostatic attraction, pollution remediation

The clinical motivation is straightforward. Dysregulated brain metabolism is increasingly recognized as a central player in the development and progression of neurodegenerative diseases, not merely a byproduct of neuronal death. The workhorse technique for probing this metabolism in patients has been positron emission tomography, or PET, most famously with the glucose analog 18F-fluorodeoxyglucose. PET can show where the brain takes up a radioactive tracer, and reduced uptake in certain regions is a well-established signature of neurodegeneration. But PET has a fundamental blind spot: it reveals only the initial uptake of the tracer and offers no insight into its downstream metabolic fate. It cannot tell clinicians whether glucose that enters a neuron is being funneled into energy production, diverted into biosynthetic pathways, or left to accumulate. It also provides no window into dynamic metabolic fluxes and turnover, the actual rates of biochemical reactions that define how a cell is functioning.

MRS and MRSI approach the problem from an entirely different physical angle. Rather than detecting radioactive decay, these techniques exploit nuclear magnetic resonance, the same phenomenon underlying MRI, to identify specific chemical compounds based on the characteristic resonance frequencies of their atomic nuclei. The result is a non-invasive, spatially resolved, in vivo measurement of metabolite levels, achieved without any radioactive tracer. Proton MRS and MRSI, which detect hydrogen nuclei, are the most widely utilized variants because they can be implemented on standard clinical MRI scanners and require no special hardware beyond what most hospitals already possess. In neurodegenerative research, proton MRS has long been used to track markers such as N-acetyl-L-aspartate, a proxy for neuronal health, and the combined signal of glutamate and glutamine, which reflects excitatory neurotransmitter metabolism. These measurements, however, remain essentially equilibrium measurements: they report how much of each metabolite is present, not how quickly it is being made, consumed, or exchanged.

The review’s central conceptual contribution is a framework for moving beyond this equilibrium limitation toward dynamic metabolic imaging. In a dynamic experiment, a labeled substrate is introduced into the body and the scanner watches as the label is incorporated, step by step, into downstream metabolites. The rate at which labeled products appear encodes the flux through specific enzymatic pathways, such as the tricarboxylic acid cycle that powers oxidative phosphorylation in mitochondria. This is the kind of information that reveals metabolic reprogramming and metabolic bottlenecks, the mechanistic signatures that static concentration measurements cannot capture. The authors lay out the key considerations for designing such experiments, including the choice of tracer, the method of signal enhancement, the acquisition hardware, the image reconstruction and preprocessing pipeline, and the metabolic modeling needed to convert time-course data into meaningful biochemical rate constants.

Tracer selection is where the physics gets interesting. Proton MRSI can be extended to heteronuclear approaches that detect less abundant nuclei, including deuterium, carbon-13, nitrogen-15, and oxygen-17. Each of these broadens the scope of accessible metabolites and pathways. Deuterium, the heavy isotope of hydrogen, has emerged as a particularly attractive option. Deuterium-labeled glucose can be administered safely, and the resulting deuterium metabolic imaging, or DMI, tracks its conversion into products such as glutamate and lactate. The revival of deuterium MRSI, the review notes, has opened new avenues for dynamic metabolic investigation in the human brain without requiring hyperpolarization, making it one of the most practical routes to clinical translation. Because deuterium resonates at a frequency far from that of protons, the technique can be implemented with relatively modest hardware additions to existing scanners.

For metabolites present at very low concentrations, however, even heteronuclear detection struggles against the intrinsically poor sensitivity of magnetic resonance. This is where nuclear spin hyperpolarization enters the picture. Hyperpolarization techniques, including dissolution dynamic nuclear polarization and parahydrogen-based methods such as parahydrogen-induced polarization and signal amplification by reversible exchange, can transiently boost the nuclear polarization of a tracer by four to five orders of magnitude before it is injected. The result is a signal strong enough to detect low-abundance metabolites in vivo in real time, enabling hyperpolarized metabolic imaging of reactions that would otherwise be invisible. Dissolution DNP, for example, can hyperpolarize carbon-13-labeled pyruvate, allowing researchers to watch the enzyme lactate dehydrogenase convert it to lactate within seconds, a direct readout of a key step in cellular energy metabolism. The review discusses these hyperpolarization methods in detail, along with their technical demands, including the specialized dissolution equipment and rapid injection systems they require.

None of this is trivial to implement. The authors are candid about the major methodological and translational challenges standing between dynamic MRSI and routine clinical use. Signal-to-noise ratio remains the persistent adversary, since metabolic signals are orders of magnitude weaker than the water signal used in conventional MRI. Higher magnetic field strengths improve sensitivity and spectral resolution but bring their own problems, including increased specific absorption rate and greater hardware cost. Vendor hardware variability compounds the difficulty, as scanners from different manufacturers differ in gradient performance, coil configurations, and pulse sequence availability. Perhaps most consequentially, there is currently a lack of standardization across acquisition, reconstruction, preprocessing, and modeling strategies, which makes it difficult to compare results between sites or to pool data across studies. The review also points to the role of modern computational tools, including machine learning and deep learning approaches, in accelerating image reconstruction and improving the extraction of quantitative information from noisy spectroscopic data.

What could dynamic MRSI actually deliver for patients with neurodegenerative diseases? The authors argue that the technique provides a promising foundation for investigating brain metabolism in vivo, offering mechanistic insights into metabolic fluxes, reprogramming, and bottlenecks that could reshape biomarker development and diagnostic practice. Mitochondrial dysfunction is a recurring theme across neurodegeneration, and dynamic measurements of TCA cycle flux or oxidative phosphorylation could reveal disease mechanisms years before structural changes appear on conventional MRI. In disorders such as multiple system atrophy, progressive supranuclear palsy, and corticobasal degeneration, where distinguishing between overlapping clinical syndromes is notoriously difficult, metabolic fingerprints might aid patient stratification for clinical trials. Longitudinal disease monitoring is another compelling application: because MRSI involves no ionizing radiation, it can be repeated safely, making it well suited to tracking disease progression and treatment response over time in a way that repeated PET scans cannot easily match.

The translational path, however, requires honest acknowledgment of where the field stands. Dynamic MRSI is currently limited to a handful of research institutions with the specialized hardware, chemistry infrastructure, and analytical expertise to pull it off. Hyperpolarized tracers must be produced under good manufacturing practice conditions, and regulatory pathways under medical device regulation frameworks add further complexity. The authors emphasize that ongoing technical and methodological advances are steadily improving feasibility, biochemical specificity, and the prospects for broader implementation, but they are equally clear that standardized and harmonized acquisition, reconstruction, and preprocessing protocols are necessary before these approaches can achieve broad clinical implementation. Without such harmonization, the field risks fragmenting into incompatible local practices that cannot generate the large, comparable datasets needed for clinical validation.

The bigger picture is one of a field at an inflection point. The shift from static to dynamic MRSI, the review concludes, has great potential to advance biomarker and diagnostic tool development, patient stratification, and longitudinal disease monitoring in neurodegenerative disorders, provided that current challenges and limitations are sufficiently addressed. In an era when most experimental therapies for Alzheimer’s and Parkinson’s diseases have stumbled, partly because patients are enrolled too late in a disease process that begins decades before symptoms, a technology that can measure the metabolic fluxes of the living brain, non-invasively and repeatedly, offers exactly the kind of window that drug developers and neurologists have been seeking. The photograph is becoming a movie, and what it shows may change how neurodegeneration is understood.

Subject of Research: Dynamic magnetic resonance spectroscopic imaging of brain metabolism in neurodegenerative disorders

Article Title: From equilibrium to dynamic metabolic imaging: the evolving role of magnetic resonance spectroscopic imaging in neurodegenerative disorders

Article References: Pokotylo, M. M., Peters, J. P., Diercks, A., Mohamad, F. H., Brüggemann, N., Pravdivtsev, A. N., Hövener, J.-B., & Prasuhn, J. (2026). From equilibrium to dynamic metabolic imaging: the evolving role of magnetic resonance spectroscopic imaging in neurodegenerative disorders. Journal of Translational Medicine. https://doi.org/10.1186/s12967-026-09084-z

Image Credits: AI Generated

DOI: 10.1186/s12967-026-09084-z

Keywords: magnetic resonance spectroscopy, MRSI, hyperpolarization, deuterium metabolic imaging, neurodegenerative diseases, brain metabolism, PET, metabolic flux, Alzheimer's disease, Parkinson's disease, biomarkers, metabolic modeling

Among the most technically striking reports is a paired imaging and treatment approach for prostate cancer built on compounds that target PSMA, or prostate-specific membrane antigen, a protein abundantly expressed on prostate cancer cells. The strategy pairs a diagnostic imaging agent with a therapeutic compound designed to bind the same molecular target, an approach at the heart of theranostics, the practice of using matched agents to first visualize a disease target and then deliver radiation to it. In mouse models, the therapeutic agent accumulated in tumors at 2.2 to 3.5 times higher levels than a reference treatment, while producing similar effects on tumor growth. Just as importantly, the imaging agent showed a distribution pattern closely matching that of the therapeutic compound, suggesting it could serve as a reliable preview of where the therapy would travel in the body. That pairing is precisely the logic that makes theranostics attractive: if the diagnostic tracer lights up a tumor, the therapeutic counterpart should follow the same path.

A second study tackled one of the persistent frustrations of molecular imaging: background noise from radioactive material still circulating in the bloodstream. When unbound tracer lingers in the blood, it can wash out the contrast between tumor and surrounding tissue, making small or subtle lesions harder to see. The researchers developed a method that uses enzymes to clear unneeded radioactive material from the circulation after injection. In mice, treatment with the enzyme carboxypeptidase G2 increased the imaging signal from tumors relative to blood by more than 40 percent. The concept relies on a clever chemical design in which the tracer is engineered so that the enzyme can alter it in the blood, accelerating its clearance, while the tracer that has already bound to its cancer target remains intact. The result is a cleaner image in which the tumor stands out more clearly against a quieter background, potentially helping clinicians visualize cancer targets that would otherwise be lost in vascular noise.

Breast cancer imaging also featured prominently in the new releases. Researchers evaluated a new PET tracer that targets nectin-4, a protein found in several cancers, in a pilot study of six patients with breast cancer. The tracer performed similarly to standard imaging for primary tumors and most lymph node metastases, but in some patients it detected more bone and liver metastases than conventional approaches. Because nectin-4 is a recognized target for antibody-drug conjugates in oncology, a PET tracer that maps its expression could eventually help clinicians identify which patients are candidates for such therapies and monitor how the target behaves across different metastatic sites. The pilot study is small, and the authors’ findings will need validation in larger cohorts, but the pattern of detecting additional bone and liver lesions hints at biological information that current standard imaging may miss.

The brain was the focus of another new tracer evaluation, this one designed to detect erythropoietin receptors, which are involved in cellular responses to low oxygen levels. In two animal models of ischemic stroke, the tracer accumulated in damaged brain regions, mapping the territory where tissue is under metabolic stress. Tests of human stroke tissue also showed tracer binding, particularly in areas surrounding the most severely damaged tissue, a distribution known as the penumbra, where cells are endangered but not yet dead. A tracer that can visualize this stressed-but-salvageable tissue in living patients could become a valuable complement to existing stroke imaging, which primarily distinguishes between tissue that is already infarcted and tissue that is perfused. By revealing the molecular signature of hypoxic stress, the tracer could help clinicians and researchers better understand the time course of injury and, potentially, refine patient selection for reperfusion therapies.

Diabetes research took a different imaging angle, using PET/CT to measure insulin-producing beta cells in 26 people with type 1 diabetes. Those diagnosed less than seven years earlier had greater beta-cell mass, larger pancreases, and less pancreatic fat than those living with longer-duration disease. The study also found that beta-cell mass was associated with insulin-producing function, linking what the scanner sees to what the pancreas can still do. The findings carry implications for the design of clinical trials aimed at preserving or regenerating beta cells, since a noninvasive measure of beta-cell mass could serve as an endpoint or a stratification tool. They also add to a growing body of evidence that the pancreas itself changes in type 1 diabetes over time, not only in its endocrine cells but in its overall size and fat content.

Not all of the new work involved new chemistry. A survey of 227 patients undergoing PET/CT examined how people weigh scan quality, duration, comfort, and access when choosing among imaging options. Nearly eight in 10 respondents preferred better-quality images even if the scans took longer, and 80 percent said they were willing to wait more than an hour after tracer injection to obtain them. Most participants reported little or no discomfort during scanning. In an era when imaging protocols are often optimized for scanner throughput, the survey is a reminder that patients themselves frequently prioritize diagnostic confidence over convenience, and that longer acquisition times or longer waits between injection and imaging may be far more acceptable to patients than imaging departments assume.

Pediatric oncology and endocrine imaging each gained a practical refinement. In a study of five children ages 3 to 7 with advanced neuroblastoma, researchers evaluated a new PET tracer targeting somatostatin receptors. The imaging identified 17 cancer lesions with strong contrast between tumors and surrounding tissues, and the scan delivered a low overall radiation dose, an especially important consideration in young children. The results support further evaluation of the approach in pediatric patients, where minimizing radiation exposure while maximizing diagnostic yield is a central design constraint. Separately, a study of 119 patients with primary hyperparathyroidism found that delayed PET/CT imaging performed 60 minutes after injection improved image contrast compared with scans taken at 10 minutes. The delayed protocol changed clinical interpretation in 19.3 percent of cases and revealed a parathyroid adenoma in eight patients whose early scans had been negative, suggesting that simply waiting longer before imaging can meaningfully improve the detection of overactive parathyroid glands.

Finally, the collection included long-view safety data on radiopharmaceutical therapy for advanced prostate cancer. Researchers compared kidney function in patients treated with radiopharmaceutical therapy or chemotherapy and found that kidney function declined similarly in both groups during the first six months. Over longer follow-up, however, the picture diverged by baseline status: declines were more pronounced in patients with preexisting chronic kidney disease, while those with initially preserved kidney function showed no significant change. That distinction matters for clinicians weighing radioligand therapy for patients whose kidneys are already compromised, and it underscores the value of extended follow-up in understanding the durability of organ safety after targeted radiation treatment. Taken together, the nine studies sketch a field that is maturing on multiple fronts at once, refining the chemistry of tracers, the timing of scans, the interpretation of images, and the measurement of long-term consequences, all in service of imaging and treating disease at the level of individual molecules and individual patients.

Subject of Research: Advances in nuclear medicine and molecular imaging reported in The Journal of Nuclear Medicine

Article Title: The Journal of Nuclear Medicine ahead-of-print tip sheet: October 9, 2026

Article References: The Journal of Nuclear Medicine ahead-of-print tip sheet: October 9, 2026. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: nuclear medicine, PET, theranostics, prostate cancer, PSMA, breast cancer, nectin-4, ischemic stroke, type 1 diabetes, neuroblastoma, hyperparathyroidism, radiopharmaceutical therapy