-
Automated Medical Report Generation for ECG Data: Bridging Medical Text and Signal Processing with Deep Learning
Authors:
Amnon Bleich,
Antje Linnemann,
Bjoern H. Diem,
Tim OF Conrad
Abstract:
Recent advances in deep learning and natural language generation have significantly improved image captioning, enabling automated, human-like descriptions for visual content. In this work, we apply these captioning techniques to generate clinician-like interpretations of ECG data. This study leverages existing ECG datasets accompanied by free-text reports authored by healthcare professionals (HCPs…
▽ More
Recent advances in deep learning and natural language generation have significantly improved image captioning, enabling automated, human-like descriptions for visual content. In this work, we apply these captioning techniques to generate clinician-like interpretations of ECG data. This study leverages existing ECG datasets accompanied by free-text reports authored by healthcare professionals (HCPs) as training data. These reports, while often inconsistent, provide a valuable foundation for automated learning. We introduce an encoder-decoder-based method that uses these reports to train models to generate detailed descriptions of ECG episodes. This represents a significant advancement in ECG analysis automation, with potential applications in zero-shot classification and automated clinical decision support.
The model is tested on various datasets, including both 1- and 12-lead ECGs. It significantly outperforms the state-of-the-art reference model by Qiu et al., achieving a METEOR score of 55.53% compared to 24.51% achieved by the reference model. Furthermore, several key design choices are discussed, providing a comprehensive overview of current challenges and innovations in this domain.
The source codes for this research are publicly available in our Git repository https://git.zib.de/ableich/ecg-comment-generation-public
△ Less
Submitted 5 December, 2024;
originally announced December 2024.
-
Enhancing ECG Analysis of Implantable Cardiac Monitor Data: An Efficient Pipeline for Multi-Label Classification
Authors:
Amnon Bleich,
Antje Linnemann,
Benjamin Jaidi,
Björn H Diem,
Tim OF Conrad
Abstract:
Implantable Cardiac Monitor (ICM) devices are demonstrating as of today, the fastest-growing market for implantable cardiac devices. As such, they are becoming increasingly common in patients for measuring heart electrical activity. ICMs constantly monitor and record a patient's heart rhythm and when triggered - send it to a secure server where health care professionals (denote HCPs from here on)…
▽ More
Implantable Cardiac Monitor (ICM) devices are demonstrating as of today, the fastest-growing market for implantable cardiac devices. As such, they are becoming increasingly common in patients for measuring heart electrical activity. ICMs constantly monitor and record a patient's heart rhythm and when triggered - send it to a secure server where health care professionals (denote HCPs from here on) can review it. These devices employ a relatively simplistic rule-based algorithm (due to energy consumption constraints) to alert for abnormal heart rhythms. This algorithm is usually parameterized to an over-sensitive mode in order to not miss a case (resulting in relatively high false-positive rate) and this, combined with the device's nature of constantly monitoring the heart rhythm and its growing popularity, results in HCPs having to analyze and diagnose an increasingly growing amount of data. In order to reduce the load on the latter, automated methods for ECG analysis are nowadays becoming a great tool to assist HCPs in their analysis. While state-of-the-art algorithms are data-driven rather than rule-based, training data for ICMs often consist of specific characteristics which make its analysis unique and particularly challenging. This study presents the challenges and solutions in automatically analyzing ICM data and introduces a method for its classification that outperforms existing methods on such data. As such, it could be used in numerous ways such as aiding HCPs in the analysis of ECGs originating from ICMs by e.g. suggesting a rhythm type.
△ Less
Submitted 12 July, 2023;
originally announced July 2023.
-
Fermionic-bosonic couplings in a weakly-deformed odd-mass nucleus, $^{93}_{41}$Nb
Authors:
J. N. Orce,
J. D. Holt,
A. Linnemann,
C. J. McKay,
C. Fransen,
J. Jolie,
T. T. S. Kuo,
S. R. Lesher,
M. T. McEllistrem,
N. Pietralla,
V. Werner,
S. W. Yates
Abstract:
A comprehensive decay scheme of $^{93}$Nb below 2 MeV has been constructed from information obtained with the $^{93}$Nb(n,n$^\prime$$γ$) and $^{94}$Zr(p,2n$γ$$γ$)$^{93}$Nb reactions. Branching ratios, lifetimes, transition multipolarities and spin assignments have been determined. From $M1$ and $E2$ strengths, fermionic-bosonic excitations of isoscalar and isovector character have been identified…
▽ More
A comprehensive decay scheme of $^{93}$Nb below 2 MeV has been constructed from information obtained with the $^{93}$Nb(n,n$^\prime$$γ$) and $^{94}$Zr(p,2n$γ$$γ$)$^{93}$Nb reactions. Branching ratios, lifetimes, transition multipolarities and spin assignments have been determined. From $M1$ and $E2$ strengths, fermionic-bosonic excitations of isoscalar and isovector character have been identified from the weak coupling $\pi1g_{9/2}$$\otimes$$^{92}_{40}$Zr and $\pi2p_{1/2}^{-1}$$\otimes$$^{94}_{42}$Mo configurations. A microscopic interpretation of such excitations is attained from shell-model calculations using low-momentum effective interactions.
△ Less
Submitted 21 July, 2010; v1 submitted 16 July, 2010;
originally announced July 2010.
-
Identification of mixed-symmetry states in an odd-mass nearly-spherical nucleus
Authors:
J. N. Orce,
J. D. Holt,
A. Linnemann,
C. J. McKay,
S. R. Lesher,
C. Fransen,
J. W. Holt,
A. Kumar,
N. Warr,
V. Werner,
J. Jolie,
T. T. S. Kuo,
M. T. McEllistrem,
N. Pietralla,
S. W. Yates
Abstract:
The low-spin structure of 93Nb has been studied using the (n,n' gamma) reaction at neutron energies ranging from 1.5 to 3.0 MeV and the 94Zr(p,2n gamma)93Nb reaction at bombarding energies from 11.5 to 19 MeV. States at 1779.7 and 1840.6 keV, respectively, are proposed as mixed-symmetry states associated with the coupling of a proton hole in the p_1/2 orbit to the 2+_1,ms state in 94Mo. These as…
▽ More
The low-spin structure of 93Nb has been studied using the (n,n' gamma) reaction at neutron energies ranging from 1.5 to 3.0 MeV and the 94Zr(p,2n gamma)93Nb reaction at bombarding energies from 11.5 to 19 MeV. States at 1779.7 and 1840.6 keV, respectively, are proposed as mixed-symmetry states associated with the coupling of a proton hole in the p_1/2 orbit to the 2+_1,ms state in 94Mo. These assignments are derived from the observed M1 and E2 transition strengths to the symmetric one-phonon states, energy systematics, spins and parities, and comparison with shell model calculations.
△ Less
Submitted 24 July, 2006;
originally announced July 2006.
-
Transition Rates between Mixed Symmetry States: First Measurement in 94Mo
Authors:
N. Pietralla,
C. Fransen,
D. Belic,
P. von Brentano,
C. Friessner,
U. Kneissl,
A. Linnemann,
A. Nord,
H. H. Pitz,
T. Otsuka,
I. Schneider,
V. Werner,
I. Wiedenhover
Abstract:
The nucleus 94Mo was investigated using a powerful combination of gamma-singles photon scattering experiments and gamma-gamma-coincidence studies following the beta-decay of 94mTc. The data survey short-lived J^pi=1+,2+ states and include branching ratios, E2/M1 mixing ratios, lifetimes, and transition strengths. The mixed-symmetry (MS) 1+ scissors mode and the 2+ MS state are identified from M1…
▽ More
The nucleus 94Mo was investigated using a powerful combination of gamma-singles photon scattering experiments and gamma-gamma-coincidence studies following the beta-decay of 94mTc. The data survey short-lived J^pi=1+,2+ states and include branching ratios, E2/M1 mixing ratios, lifetimes, and transition strengths. The mixed-symmetry (MS) 1+ scissors mode and the 2+ MS state are identified from M1 strengths. A gamma transition between MS states was observed and its rate was measured. Nine M1 and E2 strengths involving MS states agree with the O(6) limit of the interacting boson model-2 using the proton boson E2 charge as the only free parameter.
△ Less
Submitted 21 July, 1999;
originally announced July 1999.
