[College of the French Vascular Medicine Teachers (CEMV) statement: Arterial Doppler waveforms analysis (simplified Saint-Bonnet classification)]

Simplification of ankle-brachial-index measurement using Doppler-waveform classification in symptomatic patients suspected of lower extremity artery disease

Objectives

Ankle-brachial index (ABI) is commonly used for screening lower extremity peripheral artery disease (PAD) according to the international guidelines. Arterial Doppler waveform recordings is a tool to diagnose and assess PAD severity. We hypothesized that ABI measurement could be simplified by measuring only the pressure where the best arterial flow is recorded. The aim of this study was to evaluate the concordance between ABI performed according to the American Heart Association guidelines (AHA-ABI) and ABI measured according to best arterial waveform (FLOW-ABI).

Design

This was a monocentric cross-sectional study.

Methods

We included patients with exertional limb symptoms suspected of PAD. Arterial Doppler waveforms and ABI were acquired on both lower extremities at the pedis and tibial posterior arteries. Each arterial waveform was classified using the Saint-Bonnet classification. Concordances were analyzed with the kappa coefficient (confidence interval 95%). Exercise PAD study was registered n° NCT03186391.

Results

In total, one hundred and eighty-eight patients (62+/−12 years and 26.8+/−4.5 kg/m²) with exertional limb symptoms were included from May 2016 to June 2019. On each extremity, FLOW-ABI had excellent concordance for the diagnosis of PAD with the AHA-ABI with a kappa of 0.95 (95% CI: 0.90, 0.99) in the right extremity and 0.91 (95% CI: 0.86, 0.97) in the left extremity.

Conclusion

There is almost perfect concordance between AHA-ABI and FLOW-ABI. Thus, ABI can be simplified into five pressure measurements instead of seven in patient suspected of PAD with exertional limb symptoms. The question remains in patients with chronic limb ischemia.

Use of a Pre-Trained Neural Network for Automatic Classification of Arterial Doppler Flow Waveforms: A Proof of Concept

BACKGROUND

Arterial Doppler flow waveform analysis is a tool recommended for the management of lower extremity peripheral arterial disease (PAD). To standardize the waveform analysis, classifications have been proposed. Neural networks have shown a great ability to categorize data. The aim of the present study was to use an existing neural network to evaluate the potential for categorization of arterial Doppler flow waveforms according to a commonly used classification.

METHODS

The Pareto efficient ResNet-101 (ResNet-101) neural network was chosen to categorize 424 images of arterial Doppler flow waveforms according to the Simplified Saint-Bonnet classification. As a reference, the inter-operator variability between two trained vascular medicine physicians was also assessed. Accuracy was expressed in percentage, and agreement was assessed using Cohen's Kappa coefficient.

RESULTS

After retraining, ResNet-101 was able to categorize waveforms with 83.7 ± 4.6% accuracy resulting in a kappa coefficient of 0.79 (0.75-0.83) (CI 95%), compared with a kappa coefficient of 0.83 (0.79-0.87) (CI 95%) between the two physicians.

CONCLUSION

This study suggests that the use of transfer learning on a pre-trained neural network is feasible for the automatic classification of images of arterial Doppler flow waveforms.

Arterial Doppler Waveforms Are Independently Associated With Maximal Walking Distance in Suspected Peripheral Artery Disease Patients

Objective: Arterial Doppler waveform recordings are commonly used to assess lower extremity arterial disease (LEAD) severity. However, little is known about the relationship between arterial Doppler waveform profiles and patients' walking capacity. The purpose of this study was to assess whether arterial Doppler waveforms are independently associated with maximal walking distance (MWD) in patients experiencing exertional limb symptoms. Materials and Methods: This cross-sectional study included suspected LEAD patients experiencing exertional limb symptoms. In both lower extremities, arterial Doppler waveforms and ankle-brachial index (ABI) values were obtained from the pedis and tibial posterior arteries. Each arterial flow measurement was ranked using the Saint-Bonnet classification system. Treadmill stress testing (3.2 km/h, 10% slope) coupled with exercise oximetry (Exercise-TcPO2) were used to determine MWD. Delta from rest oxygen pressure (DROP) was calculated. Following treadmill stress testing, post-exercise ABI values were recorded. Univariate and multivariate analyses were used to determine the clinical variables associated with MWD. Results: 186 patients experiencing exertional limb symptoms (62 ± 12 years and 26.8 ± 4.5 kg/m2) were included between May 2016 and June 2019. Median [25th; 75th] treadmill MWD was 235 [125;500]m. Better arterial Doppler waveforms were associated with better walking distance (p = 0.0012). Whereas, median MWD was 524 [185;525]m in the group that yielded the best Doppler waveforms, it was 182 [125,305]m in the group with the poorest Doppler waveforms (p = 0.0012). MWD was significantly better (p = 0.006) in the patients with the best ABIs. However, arterial Doppler waveforms alone were significantly associated with MWD (p = 0.0009) in the multivariate model. When exercise variables (post-exercise ABI or DROP) were incorporated into the multivariate model, these were the only variables to be associated with MWD. Conclusion: Of the various clinical parameters at rest, Doppler flow waveform profiles were associated with MWD in suspected LEAD patients. A stronger link was however found between exercise variables and MWD.