Cross-Modality Transfer Learning from PSG to FMCW Radar for Event-Level Apnea–Hypopnea Segmentation

Lu, Saihu; Wang, Peng; Li, Zhenfeng; Wu, Pang; Chen, Xianxiang; Du, Lidong; Jiang, Libin; Fang, Zhen

doi:10.3390/bioengineering13030283

This is an early access version, the complete PDF, HTML, and XML versions will be available soon.

Open AccessArticle

Cross-Modality Transfer Learning from PSG to FMCW Radar for Event-Level Apnea–Hypopnea Segmentation

by

Saihu Lu

^1,2,

Peng Wang

¹

,

Zhenfeng Li

¹,

Pang Wu

¹,

Xianxiang Chen

^1,2

,

Lidong Du

^1,2

,

Libin Jiang

^3,* and

Zhen Fang

^1,2,*

¹

Aerospace Information Research Institute, Chinese Academy of Sciences (AIRCAS), Beijing 100094, China

²

School of Electronic, Electrical and Communication Engineering, University of Chinese Academy of Sciences, Beijing 100049, China

³

Beijing Tongren Eye Center, Beijing Tongren Hospital, Capital Medical University, Beijing 100730, China

^*

Authors to whom correspondence should be addressed.

Bioengineering 2026, 13(3), 283; https://doi.org/10.3390/bioengineering13030283 (registering DOI)

Submission received: 19 January 2026 / Revised: 21 February 2026 / Accepted: 25 February 2026 / Published: 27 February 2026

(This article belongs to the Special Issue AI-Driven Approaches to Diseases Detection and Diagnosis)

Download Versions Notes

Abstract

Sleep apnea–hypopnea syndrome (SAHS) is a common sleep-related breathing disorder associated with substantial cardiovascular and neurocognitive risks. Although polysomnography (PSG) remains the clinical gold standard for diagnosis, its cost, operational burden, and limited accessibility hinder scalable and longitudinal home monitoring. Frequency-modulated continuous-wave (FMCW) radar provides unobtrusive, non-contact respiration sensing, yet radar-based event detection is often constrained by scarce annotations and pronounced domain shifts relative to PSG signals. In this work, we propose a deep learning framework for apnea–hypopnea event detection from FMCW radar that combines a 1D U-Net segmentation backbone with multi-head self-attention (MHSA) and cross-modality transfer learning. The model was first pre-trained on a large public PSG dataset to learn transferable respiratory-event representations and then fine-tuned on a smaller clinically annotated radar respiration dataset using synchronized PSG labels. It produced per-sample event probabilities, which were further refined via temporal post-processing to generate event-level detections and apnea–hypopnea index (AHI) estimates. Experimental results demonstrate strong performance in the radar domain, achieving precision of

0.8137 \pm 0.0332

, recall of

0.8369 \pm 0.0470

, and an F1-score of

0.8167 \pm 0.0052

. Overall, these results indicate that PSG-to-radar transfer learning enables accurate, low-cost, and non-contact sleep apnea screening, supporting scalable longitudinal monitoring in home-like settings.

Keywords: transfer learning; apnea–hypopnea; FMCW radar; AHI estimation; health care

Share and Cite

MDPI and ACS Style

Lu, S.; Wang, P.; Li, Z.; Wu, P.; Chen, X.; Du, L.; Jiang, L.; Fang, Z. Cross-Modality Transfer Learning from PSG to FMCW Radar for Event-Level Apnea–Hypopnea Segmentation. Bioengineering 2026, 13, 283. https://doi.org/10.3390/bioengineering13030283

AMA Style

Lu S, Wang P, Li Z, Wu P, Chen X, Du L, Jiang L, Fang Z. Cross-Modality Transfer Learning from PSG to FMCW Radar for Event-Level Apnea–Hypopnea Segmentation. Bioengineering. 2026; 13(3):283. https://doi.org/10.3390/bioengineering13030283

Chicago/Turabian Style

Lu, Saihu, Peng Wang, Zhenfeng Li, Pang Wu, Xianxiang Chen, Lidong Du, Libin Jiang, and Zhen Fang. 2026. "Cross-Modality Transfer Learning from PSG to FMCW Radar for Event-Level Apnea–Hypopnea Segmentation" Bioengineering 13, no. 3: 283. https://doi.org/10.3390/bioengineering13030283

APA Style

Lu, S., Wang, P., Li, Z., Wu, P., Chen, X., Du, L., Jiang, L., & Fang, Z. (2026). Cross-Modality Transfer Learning from PSG to FMCW Radar for Event-Level Apnea–Hypopnea Segmentation. Bioengineering, 13(3), 283. https://doi.org/10.3390/bioengineering13030283

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Article metric data becomes available approximately 24 hours after publication online.

Article Menu

Cross-Modality Transfer Learning from PSG to FMCW Radar for Event-Level Apnea–Hypopnea Segmentation

Abstract

Share and Cite

Article Metrics

Further Information

Guidelines

MDPI Initiatives

Follow MDPI