Page 64 - 《中国医疗器械杂志》2026年第2期

Chinese Journal of Medical Instrumentation 2026年 第50卷 第2期

综 合 评 述

Conference on Neural Information Processing Systems. [41] YU X W, ZHANG L, DAI H X, et al. Core-periphery
Long Beach, CA, USA: Curran Associates, Inc. , 2017: principle guided redesign of self-attention in trans-
5998-6008. formers[EB/OL]. arXiv preprint, 2023[2025-11-04].
[30] DOSOVITSKIY A, BEYER L, KOLESNIKOV A, et al. https://arxiv.org/abs/2303.15569.
An image is worth 16×16 words: Transformers for [42] RAVICHANDRAN N B, LANSNER A, HERMAN P.
imagerecognition at scale[C]//International Conference Brain-like combination of feedforward and recurrent
on Learning Representations. Addis Ababa, Ethiopia: network components achieves prototype extraction and
ICLR Press, 2020. robust pattern recognition[C]//Proceedings of the Japan
[31] FAN W Q, DING Y J, NING L B, et al. A survey on Neuroscience Society Annual Meeting. Tokyo, Japan:
RAG meeting LLMs: Towards retrieval-augmented The Japan Neuroscience Society, 2022: 488-501.
large language models[C]//Proceedings of the 30th [43] SOO W W M, BATTISTA A, RADMARD P, et al.
ACM SIGKDD Conference on Knowledge Discovery Recurrent neural network dynamical systems for
and Data Mining. New York, NY, USA: Association for biological vision[C]//Proceedings of the 38th
Computing Machinery, 2024: 6491-6501. Conference on Neural Information Processing Systems.
[32] WOO S, PARK J, LEE J Y, et al. CBAM: Convolutional Vancouver, BC, Canada: Neural Information Processing
block attention module[C]//FERRARI V, HEBERT M, Systems Foundation, 2024: 135966-135982.
SMINCHISESCU C, et al. Computer Vision – ECCV [44] PAN W X, ZHAO F F, ZHAO Z Y, et al. Brain-inspired
2018. Cham: Springer International Publishing, 2018: 3- evolutionary architectures for spiking neural
19. networks[J]. IEEE Trans Artif Intell, 2024, 5(11): 5760-
[33] SUKHBAATAR S, SZLAM A, WESTON J, et al. End- 5770.
to-end memory networks[C]//Proceedings of the 29th [45] LI Y, MATEOS G, ZHANG Z W. Learning to model
International Conference on Neural Information the relationship between brain structural and functional
Processing Systems. Montréal, Canada: Neural connectomes[J]. IEEE Trans Signal Inf Process Netw,
Information Processing Systems Foundation, 2015: 2022, 8: 830-843.
2440-2448. [46] NIE W Z, REN M J, LIU A A, et al. M-GCN: Multi-
[34] WU C Y, LI Y H, MANGALAM K, et al. MeMViT: branch graph convolution network for 2D image-based
Memory-augmented multiscale vision transformer for on 3D model retrieval[J]. IEEE Trans Multimed, 2021,
efficient long-term video recognition[C]//Proceedings of 23: 1962-1976.
the 2022 IEEE/CVF Conference on Computer Vision [47] SABOUR S, FROSST N, HINTON G E. Dynamic
and Pattern Recognition (CVPR). New Orleans, LA, routing between capsules[C]//Proceedings of the 30th
USA: IEEE, 2022: 13577-13587. Advances in Neural Information Processing
[35] VAN DE VEN G M, SIEGELMANN H T, TOLIAS A Systems(NeurIPS 2017). Red Hook, NY: Curran
S. Brain-inspired replay for continual learning with Associates, Inc., 2017: 3856-3866.
artificial neural networks[J]. Nat Commun, 2020, 11(1): [48] LOTTER W, KREIMAN G, COX D. Deep predictive
4069. coding networks for video prediction and unsupervised
[36] HE X, ZHAO D C, LI Y, et al. Incorporating brain- learning[EB/OL]. arXiv preprint, 2017[2025-08-03].
inspired mechanisms for multimodal learning in https://arxiv.org/abs/1605.08104.
artificial intelligence[EB/OL]. arXiv, 2025[2025-08-03]. [49] HAN K, WEN H, ZHANG Y, et al. Deep predictive
https://arxiv.org/abs/2505.10176. coding network with local recurrent processing for
[37] PARSAPOOR M. An introduction to brain emotional object recognition[C]//Proceedings of the 31st Advances
learning inspired models (BELiMs) with an example of in Neural Information Processing Systems. Long Beach,
BELiMs’ applications[J]. Artif Intell Rev, 2019, 52(1): CA, USA: Neural Information Processing Systems
409-439. Foundation, 2018: 9221-9233.
[38] HUANG H, ZHAO L, DAI H X, et al. BI-AVAN: a [50] HAFNER D, IRPAN A, DAVIDSON J, et al. Learning
brain-inspired adversarial visual attention network for hierarchical information flow with recurrent neural
characterizing human visual attention from neural modules[C]// Proceedings of the 30th Advances in
activity[J]. IEEE Trans Multimed, 2024, 26: 11191- Neural Information Processing Systems. Long Beach,
11203. CA, USA: Neural Information Processing Systems
[39] SONG Z R, XU Y H, HE Z Z, et al. CP-ViT: Cascade Foundation, 2017: 6724-6733.
vision transformer pruning via progressive sparsity [51] WANG H E, TRIEBKORN P, BREYTON M, et al.
Prediction[EB/OL]. arXiv, 2022[2025-08-03]. https:// Virtual brain twins: from basic neuroscience to clinical
arxiv.org/abs/2203.04570. use[J]. Natl Sci Rev, 2024, 11(5): nwae079.
[40] JIANG X, ZHANG T, ZHANG S, et al. Fundamental [52] HASANI R, LECHNER M, AMINI A, et al. Liquid
functional differences between gyri and sulci: time-constant networks[C]//Proceedings of the AAAI
implications for brain function, cognition, and Conference on Artificial Intelligence. Vol 35.
behavior[J]. Psychoradiol, 2021, 1(1): 23-41. Association for the Advancement of Artificial

178