Diabetic nephropathy (DN) is a major cause of chronic kidney disease. Conventional clinical tests such as albuminuria measurement or estimation of glomerular filtration rate (GFR) can be time-consuming and have limited sensitivity for early disease. The study summarized here examined whether photoacoustic imaging (PAI) can detect early pathophysiological changes in the kidney in vivo using mouse models of DN and therefore serve as a sensitive marker for early detection.
The source report is an abstract published in J Biophotonics (2026) and presents in vivo PAI measurements of renal oxygenation, hemoglobin content, and lipid signal intensity across disease stages. The study focuses on imaging-derived functional and compositional biomarkers rather than conventional biochemical tests.
The authors evaluated three primary PAI-derived metrics:
Average blood oxygen saturation (sO2 Avr): a measure of tissue oxygenation derived from photoacoustic signals at oxygenation-sensitive wavelengths.
Average total hemoglobin concentration (HbT Avr): an estimate of regional hemoglobin content inferred from PA signal amplitude.
Lipid PA signal intensity: a measure of lipid-associated photoacoustic contrast within the renal tissue.
These metrics were compared between control animals and mice with experimentally induced DN at early and advanced stages to characterize temporal changes in renal physiology and composition detectable by PAI.
The principal observations reported in the abstract are:
In early-stage DN, sO2 Avr was significantly elevated relative to the control group. This suggests increased renal oxygen saturation at an early disease stage as detected by PAI.
As disease progressed to advanced-stage DN, sO2 Avr gradually decreased compared with the early-stage measurements.
At advanced-stage DN, both HbT Avr and lipid PA signal intensity showed dramatic reductions compared with controls or earlier stages, indicating loss of hemoglobin signal and reduced lipid-associated PA contrast in later disease.
These stage-dependent patterns imply that PAI can detect dynamic changes in renal oxygenation and tissue composition during DN progression: an initial increase in measured oxygen saturation followed by decline, and late-stage decreases in hemoglobin content and lipid signal.
The abstract reports area under the receiver operating characteristic curve (AUC) values for PAI metrics distinguishing disease stages:
The AUC of sO2 Avr for diagnosing early-stage DN was 0.847 (95% CI: 0.636–1.0), indicating good discrimination for early disease in this experimental setting.
For advanced-stage DN, the AUC for HbT Avr was 0.917 (95% CI: 0.795–1.0), and the AUC for lipid PA signal intensity was 0.875 (95% CI: 0.734–1.0). These values indicate strong discrimination of advanced disease based on hemoglobin and lipid PA signals.
These reported AUCs suggest that different PAI-derived biomarkers may be useful at distinct stages of DN: oxygenation metrics for early detection and hemoglobin/lipid signals for identifying advanced pathology.
This investigation was performed in mouse models; MeSH terms in the record include Animals, Mice, and Male. The abstract concludes that PAI was potentially useful for the early evaluation of diabetic nephropathy based on observed changes in sO2 Avr, HbT Avr, and lipid PA signal intensity and the reported diagnostic AUCs.
The PubMed abstract does not report several details that would be important for clinical interpretation and translation, including sample sizes, the precise animal model and induction method for DN, full imaging protocol and wavelengths used, spatial resolution and region-of-interest definitions, statistical methods beyond AUC reporting, and any histopathologic correlation. These methodological and validation details were not provided in the abstract available on PubMed and would need to be reviewed in the full text to assess reproducibility and applicability to clinical practice.
In summary, the source reports that photoacoustic imaging detected stage-dependent changes in renal oxygenation, hemoglobin content, and lipid signal in mouse models of diabetic nephropathy, with promising diagnostic AUCs for early and advanced stages. The authors propose that photoacoustic imaging may be a promising tool for early evaluation of DN, while full methodological details and further validation are required to determine translational potential.