Assessing arterial stiffness using characteristics of Korotkoff sounds

Q: What are the key findings of this study?

Machine-learning analysis of Korotkoff sounds distinguished young and older age-defined groups with up to 93.7% accuracy and showed modest but significant association with measured baPWV.

Q: What is the clinical evidence from this study?

Population: vascular ageing and arterial stiffness assessment (n=316). Intervention: Korotkoff sound analysis with machine learning vs. age-defined groups (young vs older) and baPWV-defined groups (high vs low). Primary outcome: classification accuracy of Korotkoff sound analysis distinguishing vascular stiffness-related groups, and correlation with measured baPWV (Accuracy 93.7% for SqueezeNet in age-defined group classification; Accuracy 89.3% for GoogLeNet in same task; moderate accuracy 87.5% and 81.3% in baPWV-defined task).

Q: What does this research mean for the field?

Korotkoff sound analysis achieved a classification accuracy of 93.7% in distinguishing older from younger adults, indicating its potential in assessing arterial stiffness related to vascular ageing. Novelty: ClaimNovelty.NOVEL_FINDING. Consensus alignment: ConsensusAlignment.NEUTRAL.

Shuqi Ren; Wei Zhao; C. Yi; Xiaoyan Deng; Zengsheng CHEN; Li Wang; Ling Xu; Yuheng Yang; Yubo Fan; Anqiang SUN

doi:10.3389/fcvm.2026.1654162

February 10, 2026SHILAP Revista de lepidopterologíaOpen Access

Assessing arterial stiffness using characteristics of Korotkoff sounds

View Full Paper

Ask AI

Bookmark

Share

Key result

Machine-learning analysis of Korotkoff sounds distinguishes age-related vascular stiffness with ~94% accuracy.

Accuracy 93.7% for SqueezeNet in age-defined group classification; Accuracy 89.3% for GoogLeNet in same task; moderate accuracy 87.5% and 81.3% in baPWV-defined task
n=316

Figures

Why the study?

Routine assessment of pulse wave velocity remains limited in many clinical and home settings despite arterial stiffness being a recognized marker of vascular ageing.

Can machine learning analysis of Korotkoff sounds during routine blood pressure measurement accurately classify arterial stiffness and vascular ageing?

Key Points

To assess the feasibility of detecting arterial stiffness and vascular ageing by analyzing acoustic Korotkoff sound features captured during routine cuff blood pressure measurement.
Recorded Korotkoff sounds using a custom device in 123 young (25.9 ± 2.2 years) and 112 older (67.5 ± 6.7 years) participants for age-stratified proof-of-concept testing.
Enrolled 81 hospital participants stratified by measured brachial-ankle pulse wave velocity (baPWV) to validate clinical applicability.
Extracted 10 time- and frequency-domain acoustic features and evaluated classification performance using logistic regression and deep learning convolutional neural network models.
Center of mass, skewness, and peak frequency differed significantly between age groups (p < 0.001 for all), with logistic regression reaching 83.7% accuracy and two deep learning models achieving 89.3% and 93.7% accuracy.
In the baPWV-defined classification task, deep learning models achieved accuracies of 87.5% and 81.3% for distinguishing high from low baPWV groups.
Multivariable linear regression adjusting for age, diabetes, BMI, and systolic blood pressure showed that skewness (β = 0.166, p = 0.030), center of mass (β = −0.310, p = 0.018), and peak frequency (β = −0.331, p = 0.029) remained independently associated with baPWV.

Structured PICO

Can machine learning analysis of Korotkoff sounds during routine blood pressure measurement accurately classify arterial stiffness and vascular ageing?

P

Population

316 participants, comprising a proof-of-concept cohort of 123 young (mean age 25.9 ± 2.2 years) and 112 older (mean age 67.5 ± 6.7 years) healthy adults, and a clinical feasibility cohort of 81 hospital participants with measured brachial-ankle pulse wave velocity (baPWV). Inclusion criteria for the age-stratified groups: age 20-30 or 60-79 years, BMI 20-25 kg/m2, good general health, non-smoking, and normal blood pressure and heart rate.

I

Intervention

Extraction of time- and frequency-domain features from Korotkoff sounds recorded during cuff-based blood pressure measurement using a custom-developed device, analyzed via traditional feature-based models and deep learning approaches (GoogLeNet, SqueezeNet).

C

Comparator

Age-stratified groups (older vs. young) as a surrogate for vascular ageing, and baPWV-stratified groups (high baPWV vs. low baPWV based on median split).

O

Outcome

Classification accuracy of age-stratified and baPWV-defined groups using Korotkoff sound features, and the correlation between acoustic features and measured baPWV.surrogate

Main Result

Effect estimate: Accuracy 93.7% for SqueezeNet in age-defined group classification; Accuracy 89.3% for GoogLeNet in same task; moderate accuracy 87.5% and 81.3% in baPWV-defined task

Limitations

Primary comparison used age as surrogate for arterial stiffness rather than direct PWV measurement
cfPWV measurement (reference standard) was not performed, baPWV used instead
Limited sample size for baPWV cohort may limit generalizability and performance
Incremental prognostic value beyond age and blood pressure was not established
Primary comparison used age as a surrogate label rather than a direct stiffness diagnosis
Clinical outcomes were not assessed
Limited sample size
Potential spectrum differences between cohorts
Lack of standardized carotid-femoral PWV (cfPWV) measurements

Ask about this paper

Answers are grounded in this paper and cite their sources.

Implication

Machine learning analysis of Korotkoff sounds during routine blood pressure measurement can extract features modestly associated with arterial stiffness, offering a potential non-invasive screening tool for vascular ageing.

Abstract

Background and Objectives Arterial stiffness is a recognized marker of vascular ageing and is associated with adverse cardiovascular outcomes. However, routine assessment of pulse wave velocity (PWV) remains limited in many clinical and home settings. This study investigated the feasibility of extracting arterial stiffness-related information from Korotkoff sounds recorded during cuff-based blood pressure measurement using feature analysis and machine learning. Materials and methods Korotkoff sounds were collected from 123 young (25.9 ± 2.2 years) participants and 112 older (67.5 ± 6.7 years) participants using a custom-developed device as a proof-of-concept for age-related vascular differences. In addition, 81 hospital participants with measured brachial-ankle PWV (baPWV) were enrolled and grouped according to baPWV to further evaluate clinical feasibility. Time- and frequency-domain features were extracted, and both traditional feature-based models and deep learning approaches were applied for classification. Results Extracted features including center of mass, skewness, and peak frequency showed significant differences between the age-stratified groups. Two deep learning models achieved classification accuracies of 89.3% and 93.7%, respectively, outperforming traditional feature-based analysis. In the baPWV-defined classification task, model performance was moderate (accuracy 87.5% and 81.3%). In the baPWV-measured cohort, Korotkoff sound-derived features showed a statistically significant but modest association with measured baPWV. Conclusion Korotkoff sounds contain measurable information related to vascular ageing and arterial stiffness, and machine learning can leverage these signals for group discrimination. Given that the primary comparison used age as a surrogate label and clinical outcomes were not assessed, the present data do not establish incremental value for cardiovascular risk stratification beyond age and blood pressure. Larger studies with standardized PWV measurements, ideally carotid-femoral PWV (cfPWV), and prospective validation are required before prognostic or risk-stratification claims can be made.

Discussion

No takes yet. Share an insight, caveat, or question.

Cite This Study

Ren et al. (2026) studied vascular ageing and arterial stiffness assessment (n=316). Korotkoff sound analysis with machine learning vs. age-defined groups (young vs older) and baPWV-defined groups (high vs low) was evaluated on classification accuracy of Korotkoff sound analysis distinguishing vascular stiffness-related groups, and correlation with measured baPWV (Accuracy 93.7% for SqueezeNet in age-defined group classification; Accuracy 89.3% for GoogLeNet in same task; moderate accuracy 87.5% and 81.3% in baPWV-defined task). Machine-learning analysis of Korotkoff sounds distinguished young and older age-defined groups with up to 93.7% accuracy and showed modest but significant association with measured baPWV.

synapsesocial.com/papers/698ebedd85a1ff6a93016349 https://doi.org/10.3389/fcvm.2026.1654162

View Full Paper

Ask AI

Bookmark

Share

Assessing arterial stiffness using characteristics of Korotkoff sounds

Figures

Key Points

Structured PICO

Main Result

Limitations

Ask about this paper

Implication

Abstract

Discussion

Cite This Study

Assessing arterial stiffness using characteristics of Korotkoff sounds

Figures

Key Points

Structured PICO

Main Result

Limitations

Ask about this paper

Implication

Abstract

Discussion

Cite This Study

Keep reading

Keep reading