What is the clinical evidence from this study?

Study design: Cohort. Population: Atherosclerotic renal artery stenosis (ARAS) (n=1405). Intervention: Stenting (in Delayed phenotype) vs. No stenting. Primary outcome: Major adverse renal events (MAREs) (HR 0.52, 95% CI 0.35 to 0.78, p=<0.01 for interaction).

What question did this study set out to answer?

To determine whether self-supervised deep learning analysis of contrast-enhanced ultrasound can decode microvascular rarefaction and predict clinical response to renal artery stenting.

June 24, 2026ResearchOpen Access

Spatiotemporal Deep Video-Phenomapping Decodes Microvascular Rarefaction in Renovascular Hypertension: A Multi-Modal Study Integrating Spatial Transcriptomics and Mitochondrial Pyroptosis

Q: What are the key findings of this study?

In patients with atherosclerotic renal artery stenosis, stenting reduced 3-year major adverse renal events in the AI-identified Delayed phenotype (HR 0.52; 95% CI 0.35-0.78; ARR 12.2%).

Q: What question did this study set out to answer?

To determine whether self-supervised deep learning analysis of contrast-enhanced ultrasound can decode microvascular rarefaction and predict clinical response to renal artery stenting.

View Full Paper

Ask AI

Bookmark

Key result

Renal stenting is linked to ~48% fewer major renal events in AI-identified delayed-perfusion phenotypes.

HR 0.52
95% CI 0.35 to 0.78
P<0.01 for interaction
NNT 8
n=1,405

Why the study?

Does AI-based video-phenomapping of contrast-enhanced ultrasound predict major adverse renal events and identify patients who benefit from stenting in atherosclerotic renal artery stenosis?

Population

1,405 middle-aged and older patients with atherosclerotic renal artery stenosis across a multi-center…

Comparison

Renal-Video-AI phenomapping and renal artery… vs Anatomical stenosis severity; non-stented…

Design

Cohort

Follow-up

3 years

Authors

LJLingjie JuBeijing Hospital RJRi JiShanghai Jiao Tong University HMHong MengFu Wai Hospital

Discussion

Member takes

Implication

AI-based video-phenomapping of renal microvascular hemodynamics identifies a specific 'Delayed' phenotype that benefits from stenting in atherosclerotic renal artery stenosis, while identifying therapeutic futility in 'Rarefied' phenotypes.

Key Points

To determine whether self-supervised deep learning analysis of contrast-enhanced ultrasound can decode microvascular rarefaction and predict clinical response to renal artery stenting.
Developed Renal-Video-AI, a self-supervised Video Swin Transformer model, and evaluated it in a multicenter Discovery Cohort (N = 1,226) and an independent External Validation Cohort (N = 122).
Characterized biological mechanisms using a prospective Multimodal Cohort with paired 10x Visium spatial transcriptomics (N = 57) and an aged two-kidney-one-clip (2K1C) murine model treated with the NLRP3 inhibitor MCC950.
Unsupervised phenomapping identified Preserved, Delayed, and Rarefied phenotypes; the Rarefied phenotype predicted major adverse renal events (HR 4.82, 95% CI 3.10 to 6.50; sHR 5.1) and increased model C-statistic from 0.72 to 0.88.
Renal stenting improved outcomes exclusively in the Delayed phenotype (HR 0.52, 95% CI 0.35 to 0.78; 3-year incidence 13.2% vs 25.4%; ARR 12.2%, NNT = 8), with no benefit in the Preserved (HR 0.98) or Rarefied (HR 1.05) phenotypes (P < 0.01 for interaction).
Spatial transcriptomics localized hypoxia and pyroptosis signatures to rarefied zones, while MCC950 inhibition of the mitochondrial ROS–NLRP3–pyroptosis axis restored microvascular perfusion in mice.

Study Design

Type

Cohort (n=1,405)

Multicenter

Yes

Structured PICO

Does AI-based video-phenomapping of contrast-enhanced ultrasound predict major adverse renal events and identify patients who benefit from stenting in atherosclerotic renal artery stenosis?

Population

1,405 middle-aged and older patients with atherosclerotic renal artery stenosis across discovery, validation, and multimodal cohorts.

Exposure

Renal-Video-AI phenomapping (extracting spatiotemporal hemodynamic features from contrast-enhanced ultrasound) and renal artery stenting.

Comparator

Anatomical stenosis severity (standard clinical model); non-stented patients.

Outcome

Major adverse renal events (MAREs)composite

Main Result

Hazard Ratio: 0.52 (95% CI 0.35–0.78)

Absolute Event Rate: 13.2% vs 25.4%

Absolute Risk Reduction: 12.2%

Number Needed to Treat: 8

p-value: p=<0.01 for interaction

Cite This Study

Ju et al. (2026) conducted a cohort in Atherosclerotic renal artery stenosis (ARAS) (n=1,405). Stenting (in Delayed phenotype) vs. No stenting was evaluated on Major adverse renal events (MAREs) (HR 0.52, 95% CI 0.35 to 0.78, p=<0.01 for interaction). In patients with atherosclerotic renal artery stenosis, stenting reduced 3-year major adverse renal events in the AI-identified Delayed phenotype (HR 0.52; 95% CI 0.35-0.78; ARR 12.2%).

synapsesocial.com/papers/6a3c2352d15afadd906f9fcf https://doi.org/10.34133/research.1339

View Full Paper

Ask AI

Bookmark

Spatiotemporal Deep Video-Phenomapping Decodes Microvascular Rarefaction in Renovascular Hypertension: A Multi-Modal Study Integrating Spatial Transcriptomics and Mitochondrial Pyroptosis

Discussion

Member takes

Implication

Key Points

Study Design

Structured PICO

Main Result

Cite This Study

Spatiotemporal Deep Video-Phenomapping Decodes Microvascular Rarefaction in Renovascular Hypertension: A Multi-Modal Study Integrating Spatial Transcriptomics and Mitochondrial Pyroptosis

Discussion

Member takes

Implication

Key Points

Study Design

Structured PICO

Main Result

Cite This Study

Also Consider

Also Consider