Multi-View Integrative Attention-Based Deep Representation Learning for Irregular Clinical Time-Series Data

Lee, Yurim; Jun, Eunji; Choi, Jaehun; Suk, Heung-Il

doi:10.1109/jbhi.2022.3172549

Cited by 11 publications

(12 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…SAM has emerged as a novel method for processing clinical data. Lee et al used SAM to process irregular multivariate time-series data in the EHR to predict in-hospital mortality, length of stay, and phenotyping [ 24 ]. Xu Y et al used SAM to selectively learn different positions in pathological slide images to improve the performance of colorectal cancer diagnosis [ 25 ].…”

Section: Discussionmentioning

confidence: 99%

Multi-head self-attention mechanism enabled individualized hemoglobin prediction and treatment recommendation systems in anemia management for hemodialysis patients

Yang¹,

Lee²,

Liao³

et al. 2023

Heliyon

View full text Add to dashboard Cite

Section: Discussionmentioning

confidence: 99%

Multi-head self-attention mechanism enabled individualized hemoglobin prediction and treatment recommendation systems in anemia management for hemodialysis patients

Yang¹,

Lee²,

Liao³

et al. 2023

Heliyon

View full text Add to dashboard Cite

“…Uneven time intervals between measurements of a patient's attributes, such as heart rate, lead to irregularity in the electronic health records (EHR), which results in missing values while preparing the data for processing with machine learning models [2]. In EHR, irregular time-series (ITS) occur due to several reasons.…”

Section: Introductionmentioning

confidence: 99%

“…Due to the importance and wide prevalence of ITS in healthcare, ITS and the resulting missing values in EHR have received increasing attention from the research community, and there has been extensive research to address the irregularity, e.g., [2], [3], [8]- [10]. A wide variety of techniques have been proposed to handle ITS, e.g., from the traditional statistical techniques for replacement of missing values (such as using mean and median values etc.…”

Section: Introductionmentioning

confidence: 99%

“…Many other modern techniques for modelling fail to capture featurecorrelations in time-series [12], separate the modelling of missingness from the downstream task, and fail to learn the missingness pattern, or are not adequately efficient to handle long sequences, noise and multi-modality of the data [11] etc. Moreover, there are scarce techniques that can handle completely missing time-steps, such as in [8] and others mostly handle partially missing values, such as in [2].…”

Section: Introductionmentioning

confidence: 99%

“…The transformer based models are one of the most successful deep learning techniques, which have shown great results across different domains [2], [18]. However, the quadratic dependence of the transformer based models on the input limits their application to long sequences.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Continuous Patient State Attention Models

Chauhan

Thakur

O'Donoghue

et al. 2022

Preprint

View full text Add to dashboard Cite

Irregular time-series (ITS) are prevalent in the electronic health records (EHR) as the data is recorded in EHR system as per the clinical guidelines/requirements but not for research and also depends on the patient health status. ITS present challenges in training of machine learning algorithms, which are mostly built on assumption of coherent fixed dimensional feature space. In this paper, we propose a computationally efficient variant of the transformer based on the idea of cross-attention, called Perceiver, for time-series in healthcare. We further develop continuous patient state attention models, using the Perceiver and the transformer to deal with ITS in EHR. The continuous patient state models utilise neural ordinary differential equations to learn the patient health dynamics, i.e., patient health trajectory from the observed irregular time-steps, which enables them to sample any number of time-steps at any time. The performance of the proposed models is evaluated on in-hospital-mortality prediction task on Physionet-2012 challenge and MIMIC-III datasets. The Perceiver model significantly outperforms the baselines and reduces the computational complexity, as compared with the transformer model, without significant loss of performance. The carefully designed experiments to study irregularity in healthcare also show that the continuous patient state models outperform the baselines. The code is publicly released and verified at https://codeocean.com/capsule/4587224

show abstract

Contrastive Learning-Based Imputation-Prediction Networks for In-hospital Mortality Risk Modeling Using EHRs

Liu,

Zhang,

Qin

et al. 2023

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Multi-View Integrative Attention-Based Deep Representation Learning for Irregular Clinical Time-Series Data

Cited by 11 publications

References 14 publications

Multi-head self-attention mechanism enabled individualized hemoglobin prediction and treatment recommendation systems in anemia management for hemodialysis patients

Multi-head self-attention mechanism enabled individualized hemoglobin prediction and treatment recommendation systems in anemia management for hemodialysis patients

Continuous Patient State Attention Models

Contrastive Learning-Based Imputation-Prediction Networks for In-hospital Mortality Risk Modeling Using EHRs

Contact Info

Product

Resources

About