DiffECG: Diffusion Model-Powered Label-Efficient and Personalized Arrhythmia Diagnosis

Authors: Tianren Zhou, Zhenge Jia, Dongxiao Yu, Zhaoyan Shen

IJCAI 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results show that our proposed method outperforms the SOTA method by 37.9% and 23.9% in terms of generalization and personalization performance, respectively. The source code is available at: https://github.com/Auguuust/Diff EC.
Researcher Affiliation Academia Tianren Zhou , Zhenge Jia , Dongxiao Yu and Zhaoyan Shen School of Computer Science and Technology, Shandong University EMAIL, EMAIL
Pseudocode No The paper includes equations and figures illustrating the model architecture and processes, but no explicitly labeled pseudocode or algorithm blocks.
Open Source Code Yes The source code is available at: https://github.com/Auguuust/Diff EC.
Open Datasets Yes We evaluate all the methods based on five public ECG datasets. CPSC [Goldberger et al., 2000; Liu et al., 2018] consists of ECG records with 9 different types of arrhythmias ranging from 6 to 60 seconds.Chapman [Zheng et al., 2020] consists of ECG records with 11 different types of arrhythmias with a 10-second duration for each record. PTB [Bousseljot et al., 1995; Physio Bank, 2000] dataset contains 549 records ranging from 1 to 5 seconds. Georgia [Alday et al., 2020] consists of ECG records with 56 different types of arrhythmias, with a 10second duration for each record. LTAF [Petrutiu et al., 2007; Goldberger et al., 2000] includes 84 ECG records of subjects with paroxysmal or sustained AF.
Dataset Splits Yes Specifically, we first split each dataset (i.e., CPSC, Chapman, PTB, and Georgia) into finetuning and testing sets subject-wisely (with a splitting ratio of 2:8) to ensure the subject s data is not mixed between the finetuning and testing sets. [...] Specifically, we first split each subject s data in the LTAF dataset with a 1:9 ratio, using 10% for fine-tuning and 90% for testing.
Hardware Specification Yes The training is conducted on a server equipped with four NVIDIA RTX 4090 GPUs, an Intel Xeon Platinum 8480+ CPU, and 1 TB of memory.
Software Dependencies No We use Py Torch for all methods to build networks, train models, and report detection performance.
Experiment Setup Yes All methods are pre-trained using the CPSC and Chapman datasets, respectively, fine-tuned on the other datasets with 20 epochs. The noise-adding step t of our method is fixed to 20.