Back to Evidence Hub
Case Study S046
2026 Release

Golenderov 2026: Rhythmic Spectrum Disorders in Field

Golenderov et al.
Access Paper

Quick Conclusion: Demonstrates the effectiveness of advanced neural networks in processing heterogeneous and noisy smartphone sensor data.


📊 Key Accuracy Metrics

MetricResult
Sensitivity99.80%
Positive Predictive Value (PPV)99.09%
Total Data collected7,700+ hours
ModelU-Net v3 with Multi-scale convolutions and Attention Gates


🔍 Study Analysis

Objective & Population

Technical Development / Field Study. Cohort: Volunteers using consumer smartphones (7,700+ hours of data) (N=100).

What it Supports

Validates robust AO peak detection in noisy real-world data (7,700+ hours) with 99.8% sensitivity using U-Net v3.

What it Does Not Support

The study does not provide direct validation for the diagnosis of specific rhythmic disorders (like AFib) in this paper, but rather provides the robust detection framework necessary for such diagnoses.


🛠 Technical Context

Featured Illustration

Figure 2: Labeled SCG.

Figure 2: Labeled SCG.

Study Snapshot

Metadata Summary

Target Population

Volunteers using consumer smartphones (7,700+ hours of data)

N

Sample Size

100 Subjects

Validated Metric

99.80%

Critical Appraisal
supporting

Achieved state-of-the-art accuracy for AO peak detection in noisy, field-collected smartphone SCG data.