Title: Reconstruction of central arterial pressure waveform based on CBL-iTransformer model from radial arterial pressure waveform

Authors: Xiaolu Li; Tao Peng; Libin Zhang; Hanguang Xiao

Addresses: School of Artificial Intelligence, Chongqing University of Technology, Chongqing, 401135, China ' School of Artificial Intelligence, Chongqing University of Technology, Chongqing, 401135, China ' School of Artificial Intelligence, Chongqing University of Technology, Chongqing, 401135, China ' School of Artificial Intelligence, Chongqing University of Technology, Chongqing, 401135, China

Abstract: Central arterial pressure (CAP) is a key parameter for assessing cardiovascular health and related disease risks. Accurate, non-invasive, and continuous reconstruction of CAP is crucial for cardiovascular disease evaluation, but traditional and some deep learning methods show limited precision and feature extraction ability. This study proposes the CBL-iTransformer model, which is aimed at improving the accuracy of CAP waveform reconstruction. The model is validated using radial arterial pressure and CAP data from patients, and its performance is compared with traditional and deep learning methods. The results demonstrate that the CBL-iTransformer model effectively reconstructs CAP waveforms and achieves reliable estimation for central aortic systolic pressure and diastolic pressure outperforms the compared models under the tested conditions. In addition, Bland-Altman analysis indicates a high level of agreement between the reconstructed and reference measurements.

Keywords: central aortic pressure reconstruction; central arterial pressure; deep learning; iTransformer model; BiLSTM; LSTM; waveform reconstruction.

DOI: 10.1504/IJBET.2026.153231

International Journal of Biomedical Engineering and Technology, 2026 Vol.50 No.3, pp.252 - 270

Received: 18 Aug 2025
Accepted: 24 Oct 2025
Published online: 29 Apr 2026 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article