A Class-Imbalance Aware and Explainable Spatio-Temporal Graph Attention Network for Neonatal Seizure Detection

Raeisi, Khadijeh; Khazaei, Mohammad Reza; Tamburro, Gabriella; Croce, Pierpaolo; Comani, Silvia; Zappasodi, Filippo

doi:10.1142/s0129065723500466

Cited by 17 publications

(2 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While such methods have shown promising results, physiological research has shown that the functional connectivity of the brain is not solely related to distance [16], and thus, defining an appropriate graph structure of EEG signals remains an open problem.Moreover, constructing the graph structure manually may not be the best option since each individual's brain functional connectivity has its specificity, making it challenging to generalize a general graph structure across different individuals [23][24][25]. Recently, Chen et al [26] proposed a soft connection graph learning model leveraging a self-attention mechanism for depression detection tasks, which results in a more precise and concise representation of the EEG signals.…”

Section: Introductionmentioning

confidence: 99%

Attention-guided graph structure learning network for EEG-enabled auditory attention detection

Zeng,

Cai,

Xie

2024

J. Neural Eng.

View full text Add to dashboard Cite

Humans possess the remarkable ability to selectively focus on one sound source in a cocktail party scenario. Decoding auditory attention from brain signals is essential for the development of neuro-steered hearing aids. However, it remains challenging to extract discriminative feature representation from electroencephalography (EEG) signals for auditory attention detection (AAD) tasks, and most methods ignore the intrinsic relationship between different EEG channels. To address these challenges, we propose a novel attention-guided graph structure learning network, AGSLnet, which leverages potential relationships between EEG channels to improve AAD performance. Specifically, AGSLnet is designed to dynamically capture latent relationships between channels and construct a graph structure of EEG signals. We evaluated AGSLnet on two publicly available AAD datasets and demonstrated its superiority and robustness over state-of-the-art models. Furthermore, visualization of the graph structure trained by AGSLnet supports previous neuroscience findings, enhancing our understanding of the underlying neural mechanisms.

show abstract

Section: Introductionmentioning

confidence: 99%

Attention-guided graph structure learning network for EEG-enabled auditory attention detection

Zeng,

Cai,

Xie

2024

J. Neural Eng.

View full text Add to dashboard Cite

show abstract

“…Just like in other fields [20][21][22][23][24][25][26][27][28], weather forecasting systems driven by spatio-temporal data and based on deep learning (DL) [29][30][31][32][33][34][35][36][37] have demonstrated superior performance compared to Numerical Weather Prediction (NWP) methods [38][39][40][41][42]. Nevertheless, these global weather forecasting models face challenges in balancing the trade-off between forecasting coverage and computational efficiency.…”

Section: Introductionmentioning

confidence: 99%

A spatio-temporal fusion deep learning network with application to lightning nowcasting

Zhou,

Fan,

Neri

2024

ICA

View full text Add to dashboard Cite

Lightning is a rapidly evolving phenomenon, exhibiting both mesoscale and microscale characteristics. Its prediction significantly relies on timely and accurate data observation. With the implementation of new generation weather radar systems and lightning detection networks, radar reflectivity image products, and lightning observation data are becoming increasingly abundant. Research focus has shifted towards lightning nowcasting (prediction of imminent events), utilizing deep learning (DL) methods to extract lightning features from very large data sets. In this paper, we propose a novel spatio-temporal fusion deep learning lightning nowcasting network (STF-LightNet) for lightning nowcasting. The network is based on a 3-dimensional U-Net architecture with encoder-decoder blocks and adopts a structure of multiple branches as well as the main path for the encoder block. To address the challenges of feature extraction and fusion of multi-source data, multiple branches are used to extract different data features independently, and the main path fuses these features. Additionally, a spatial attention (SA) module is added to each branch and the main path to automatically identify lightning areas and enhance their features. The main path fusion is conducted in two steps: the first step fuses features from the branches, and the second fuses features from the previous and current levels of the main path using two different methodsthe weighted summation fusion method and the attention gate fusion method. To overcome the sparsity of lightning observations, we employ an inverse frequency weighted cross-entropy loss function. Finally, STF-LightNet is trained using observations from the previous half hour to predict lightning in the next hour. The outcomes illustrate that the fusion of both the multi-branch and main path structures enhances the network’s ability to effectively integrate features from diverse data sources. Attention mechanisms and fusion modules allow the network to capture more detailed features in the images.

show abstract

Automated machine learning with interpretation: A systematic review of methodologies and applications in healthcare

Yuan,

Yu,

Xie

et al. 2024

Medicine Advances

View full text Add to dashboard Cite

Machine learning (ML) has achieved substantial success in performing healthcare tasks in which the configuration of every part of the ML pipeline relies heavily on technical knowledge. To help professionals with borderline expertise to better use ML techniques, Automated ML (AutoML) has emerged as a prospective solution. However, most models generated by AutoML are black boxes that are challenging to comprehend and deploy in healthcare settings. We conducted a systematic review to examine AutoML with interpretation systems for healthcare. We searched four databases (MEDLINE, EMBASE, Web of Science, and Scopus) complemented with seven prestigious ML conferences (AAAI, ACL, ICLR, ICML, IJCAI, KDD, and NeurIPS) that reported AutoML with interpretation for healthcare before September 1, 2023. We included 118 articles related to AutoML with interpretation in healthcare. First, we illustrated AutoML techniques used in the included publications, including automated data preparation, automated feature engineering, and automated model development, accompanied by a real‐world case study to demonstrate the advantages of AutoML over classic ML. Then, we summarized interpretation methods: feature interaction and importance, data dimensionality reduction, intrinsically interpretable models, and knowledge distillation and rule extraction. Finally, we detailed how AutoML with interpretation has been used for six major data types: image, free text, tabular data, signal, genomic sequences, and multi‐modality. To some extent, AutoML with interpretation provides effortless development and improves users' trust in ML in healthcare settings. In future studies, researchers should explore automated data preparation, seamless integration of automation and interpretation, compatibility with multi‐modality, and utilization of foundation models.

show abstract

A Class-Imbalance Aware and Explainable Spatio-Temporal Graph Attention Network for Neonatal Seizure Detection

Cited by 17 publications

References 49 publications

Attention-guided graph structure learning network for EEG-enabled auditory attention detection

Attention-guided graph structure learning network for EEG-enabled auditory attention detection

A spatio-temporal fusion deep learning network with application to lightning nowcasting

Automated machine learning with interpretation: A systematic review of methodologies and applications in healthcare

Contact Info

Product

Resources

About