Parameter-Efficient 12-Lead ECG Reconstruction from a Single Lead

Junseok Lee; Yeonho Yoo; Jinkyu Kim; Dosun Lim; Gyeongsik Yang; Chuck Yoo

doi:10.1007/978-3-032-04937-7_41

Parameter-Efficient 12-Lead ECG Reconstruction from a Single Lead

Junseok Lee
, Yeonho Yoo
, Jinkyu Kim
, Dosun Lim
, Gyeongsik Yang^*
, Chuck Yoo

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

With the rise of wearable IoT devices such as smartwatches and smart rings, ECG signals have become more accessible and made cardiovascular monitoring a reality. However, analyzing the ECG signals for complex conditions, such as bundle branch blocks and myocardial infarction, requires multi-lead ECG data. Although various deep learning models for ECG reconstruction have been proposed, they are computationally expensive and unsuitable on resource-constrained wearable IoT devices. To address this challenge, we propose mEcgNet, a parameter-efficient model for reconstructing 12-lead ECG signals from a single lead. mEcgNet introduces a modular deep learning architecture for parameter efficiency and separates the single lead-I signal into multiple frequency segments to improve accuracy. Our experiments demonstrate that mEcgNet significantly reduces the number of parameters and inference time by ∼23.1× and ∼5.4×, respectively, compared to existing state-of-the-art models. Furthermore, it reduces the reconstruction error by ∼22.1%, demonstrating its high accuracy and efficiency.

Original language	English
Title of host publication	Medical Image Computing and Computer Assisted Intervention, MICCAI 2025 - 28th International Conference, 2025, Proceedings
Editors	James C. Gee, Jaesung Hong, Carole H. Sudre, Polina Golland, Daniel C. Alexander, Juan Eugenio Iglesias, Archana Venkataraman, Jong Hyo Kim
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	431-441
Number of pages	11
ISBN (Print)	9783032049360
DOIs	https://doi.org/10.1007/978-3-032-04937-7_41
Publication status	Published - 2026
Event	28th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2025 - Daejeon, Korea, Republic of Duration: 2025 Sept 23 → 2025 Sept 27

Publication series

Name	Lecture Notes in Computer Science
Volume	15961 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	28th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2025
Country/Territory	Korea, Republic of
City	Daejeon
Period	25/9/23 → 25/9/27

Bibliographical note

Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2026.

Keywords

ECG reconstruction
Frequency-based segment partitioning
mEcgNet
Parameter-efficient model
Wearable IoT device

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-032-04937-7_41

Cite this

Lee, J., Yoo, Y., Kim, J., Lim, D., Yang, G., & Yoo, C. (2026). Parameter-Efficient 12-Lead ECG Reconstruction from a Single Lead. In J. C. Gee, J. Hong, C. H. Sudre, P. Golland, D. C. Alexander, J. E. Iglesias, A. Venkataraman, & J. H. Kim (Eds.), Medical Image Computing and Computer Assisted Intervention, MICCAI 2025 - 28th International Conference, 2025, Proceedings (pp. 431-441). (Lecture Notes in Computer Science; Vol. 15961 LNCS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-032-04937-7_41

Parameter-Efficient 12-Lead ECG Reconstruction from a Single Lead. / Lee, Junseok; Yoo, Yeonho; Kim, Jinkyu et al.
Medical Image Computing and Computer Assisted Intervention, MICCAI 2025 - 28th International Conference, 2025, Proceedings. ed. / James C. Gee; Jaesung Hong; Carole H. Sudre; Polina Golland; Daniel C. Alexander; Juan Eugenio Iglesias; Archana Venkataraman; Jong Hyo Kim. Springer Science and Business Media Deutschland GmbH, 2026. p. 431-441 (Lecture Notes in Computer Science; Vol. 15961 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Lee, J, Yoo, Y, Kim, J , Lim, D , Yang, G & Yoo, C 2026, Parameter-Efficient 12-Lead ECG Reconstruction from a Single Lead. in JC Gee, J Hong, CH Sudre, P Golland, DC Alexander, JE Iglesias, A Venkataraman & JH Kim (eds), Medical Image Computing and Computer Assisted Intervention, MICCAI 2025 - 28th International Conference, 2025, Proceedings. Lecture Notes in Computer Science, vol. 15961 LNCS, Springer Science and Business Media Deutschland GmbH, pp. 431-441, 28th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2025, Daejeon, Korea, Republic of, 25/9/23. https://doi.org/10.1007/978-3-032-04937-7_41

Lee J, Yoo Y, Kim J , Lim D , Yang G , Yoo C. Parameter-Efficient 12-Lead ECG Reconstruction from a Single Lead. In Gee JC, Hong J, Sudre CH, Golland P, Alexander DC, Iglesias JE, Venkataraman A, Kim JH, editors, Medical Image Computing and Computer Assisted Intervention, MICCAI 2025 - 28th International Conference, 2025, Proceedings. Springer Science and Business Media Deutschland GmbH. 2026. p. 431-441. (Lecture Notes in Computer Science). doi: 10.1007/978-3-032-04937-7_41

Lee, Junseok ; Yoo, Yeonho ; Kim, Jinkyu et al. / Parameter-Efficient 12-Lead ECG Reconstruction from a Single Lead. Medical Image Computing and Computer Assisted Intervention, MICCAI 2025 - 28th International Conference, 2025, Proceedings. editor / James C. Gee ; Jaesung Hong ; Carole H. Sudre ; Polina Golland ; Daniel C. Alexander ; Juan Eugenio Iglesias ; Archana Venkataraman ; Jong Hyo Kim. Springer Science and Business Media Deutschland GmbH, 2026. pp. 431-441 (Lecture Notes in Computer Science).

@inproceedings{32973428bd924b5aa36ecf6d8f8df50e,

title = "Parameter-Efficient 12-Lead ECG Reconstruction from a Single Lead",

abstract = "With the rise of wearable IoT devices such as smartwatches and smart rings, ECG signals have become more accessible and made cardiovascular monitoring a reality. However, analyzing the ECG signals for complex conditions, such as bundle branch blocks and myocardial infarction, requires multi-lead ECG data. Although various deep learning models for ECG reconstruction have been proposed, they are computationally expensive and unsuitable on resource-constrained wearable IoT devices. To address this challenge, we propose mEcgNet, a parameter-efficient model for reconstructing 12-lead ECG signals from a single lead. mEcgNet introduces a modular deep learning architecture for parameter efficiency and separates the single lead-I signal into multiple frequency segments to improve accuracy. Our experiments demonstrate that mEcgNet significantly reduces the number of parameters and inference time by ∼23.1× and ∼5.4×, respectively, compared to existing state-of-the-art models. Furthermore, it reduces the reconstruction error by ∼22.1\%, demonstrating its high accuracy and efficiency.",

keywords = "ECG reconstruction, Frequency-based segment partitioning, mEcgNet, Parameter-efficient model, Wearable IoT device",

author = "Junseok Lee and Yeonho Yoo and Jinkyu Kim and Dosun Lim and Gyeongsik Yang and Chuck Yoo",

note = "Publisher Copyright: {\textcopyright} The Author(s), under exclusive license to Springer Nature Switzerland AG 2026.; 28th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2025 ; Conference date: 23-09-2025 Through 27-09-2025",

year = "2026",

doi = "10.1007/978-3-032-04937-7\_41",

language = "English",

isbn = "9783032049360",

series = "Lecture Notes in Computer Science",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "431--441",

editor = "Gee, \{James C.\} and Jaesung Hong and Sudre, \{Carole H.\} and Polina Golland and Alexander, \{Daniel C.\} and Iglesias, \{Juan Eugenio\} and Archana Venkataraman and Kim, \{Jong Hyo\}",

booktitle = "Medical Image Computing and Computer Assisted Intervention, MICCAI 2025 - 28th International Conference, 2025, Proceedings",

address = "Germany",

}

TY - GEN

T1 - Parameter-Efficient 12-Lead ECG Reconstruction from a Single Lead

AU - Lee, Junseok

AU - Yoo, Yeonho

AU - Kim, Jinkyu

AU - Lim, Dosun

AU - Yang, Gyeongsik

AU - Yoo, Chuck

N1 - Publisher Copyright: © The Author(s), under exclusive license to Springer Nature Switzerland AG 2026.

PY - 2026

Y1 - 2026

N2 - With the rise of wearable IoT devices such as smartwatches and smart rings, ECG signals have become more accessible and made cardiovascular monitoring a reality. However, analyzing the ECG signals for complex conditions, such as bundle branch blocks and myocardial infarction, requires multi-lead ECG data. Although various deep learning models for ECG reconstruction have been proposed, they are computationally expensive and unsuitable on resource-constrained wearable IoT devices. To address this challenge, we propose mEcgNet, a parameter-efficient model for reconstructing 12-lead ECG signals from a single lead. mEcgNet introduces a modular deep learning architecture for parameter efficiency and separates the single lead-I signal into multiple frequency segments to improve accuracy. Our experiments demonstrate that mEcgNet significantly reduces the number of parameters and inference time by ∼23.1× and ∼5.4×, respectively, compared to existing state-of-the-art models. Furthermore, it reduces the reconstruction error by ∼22.1%, demonstrating its high accuracy and efficiency.

AB - With the rise of wearable IoT devices such as smartwatches and smart rings, ECG signals have become more accessible and made cardiovascular monitoring a reality. However, analyzing the ECG signals for complex conditions, such as bundle branch blocks and myocardial infarction, requires multi-lead ECG data. Although various deep learning models for ECG reconstruction have been proposed, they are computationally expensive and unsuitable on resource-constrained wearable IoT devices. To address this challenge, we propose mEcgNet, a parameter-efficient model for reconstructing 12-lead ECG signals from a single lead. mEcgNet introduces a modular deep learning architecture for parameter efficiency and separates the single lead-I signal into multiple frequency segments to improve accuracy. Our experiments demonstrate that mEcgNet significantly reduces the number of parameters and inference time by ∼23.1× and ∼5.4×, respectively, compared to existing state-of-the-art models. Furthermore, it reduces the reconstruction error by ∼22.1%, demonstrating its high accuracy and efficiency.

KW - ECG reconstruction

KW - Frequency-based segment partitioning

KW - mEcgNet

KW - Parameter-efficient model

KW - Wearable IoT device

UR - https://www.scopus.com/pages/publications/105017858230

U2 - 10.1007/978-3-032-04937-7_41

DO - 10.1007/978-3-032-04937-7_41

M3 - Conference contribution

AN - SCOPUS:105017858230

SN - 9783032049360

T3 - Lecture Notes in Computer Science

SP - 431

EP - 441

BT - Medical Image Computing and Computer Assisted Intervention, MICCAI 2025 - 28th International Conference, 2025, Proceedings

A2 - Gee, James C.

A2 - Hong, Jaesung

A2 - Sudre, Carole H.

A2 - Golland, Polina

A2 - Alexander, Daniel C.

A2 - Iglesias, Juan Eugenio

A2 - Venkataraman, Archana

A2 - Kim, Jong Hyo

PB - Springer Science and Business Media Deutschland GmbH

T2 - 28th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2025

Y2 - 23 September 2025 through 27 September 2025

ER -