Page 65 - 《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

Intelligence, 2021: 7657-7666. Random synaptic feedback weights support error
[53] BARDOZZO F, TERLIZZI A, SIMONCINI C, et al. backpropagation for deep learning[J]. Nat Commun,
Elegans-AI: how the connectome of a living organism 2016, 7: 13276.
could model artificial neural networks[J]. [68] MARBLSTONE A H, WAYNE G, KORDING K P.
Neurocomputing, 2024, 584: 127598. Toward an integration of deep learning and
[54] HUANG Z, NEWMAN M, VAIDA M, et al. Planarian neuroscience[J]. Front Comput Neurosci, 2016, 10: 94.
neural networks: Evolutionary patterns from basic [69] WHITTINGTON J C R, BOGACZ R. An
Bilateria shaping modern artificial neural network approximation of the error backpropagation algorithm in
architectures[EB/OL]. arXiv preprint, 2025[2025-08- a predictive coding network with local Hebbian synaptic
03]. https://arxiv.org/abs/2501.04700. plasticity[J]. Neural Comput, 2017, 29(5): 1229-1262.
[55] RONG D, DONG H, GAO X, et al. Improving [70] KIRKPATRICK J, PASCANU R, RABINOWITZ N, et
unsupervised task-driven models of ventral visual stream al. Overcoming catastrophic forgetting in neural
via relative position predictivity [EB/OL]. arXiv networks[J]. Proc Natl Acad Sci U S A, 2017, 114(13):
preprint, 2025[2025-08-03]. https://arxiv.org/abs/2505. 3521-3526.
08316. [71] ZENKE F, POOLE B, GANGULI S. Continual learning
[56] WANG X B, LIU C X Z, ZHAO M, et al. An artificial through synaptic intelligence[C]//Proceedings of the
neural network for image classification inspired by the 34th International Conference on Machine Learning.
aversive olfactory learning neural circuit in Proceedings of Machine Learning Research, 2017, 70:
Caenorhabditis elegans[J]. Adv Sci (Weinh), 2025, 3987-3995.
12(7): e2410637. [72] MASSE N Y, GRANT G D, FREEDMAN D J.
[57] JÜRGENSEN A M, KHALILI A, CHICCA E, et al. A Alleviating catastrophic forgetting using context-
neuromorphic model of olfactory processing and sparse dependent gating and synaptic stabilization[J]. Proc Natl
coding in the Drosophila larva brain[J]. Neuromorph Acad Sci U S A, 2018, 115(44): E10467-E10475.
Comput Eng, 2021, 1(2): 024008. [73] PAN S J, YANG Q. A survey on transfer learning[J].
[58] 张笃振, 程翔, 王岩松, 等. 生物结构启发基本网络算 IEEE Trans Knowl Data Eng, 2010, 22(10): 1345-1359.
子助力类脑智能研究[J]. 人工智能, 2022(6): 54-64. [74] FINN C, ABBEEL P, LEVINE S. Model-agnostic meta-
[59] SHI J H, TRIPP B, SHEA-BROWN E, et al. MouseNet: learning for fast adaptation of deep networks
a biologically constrained convolutional neural network [C]//Proceedings of the 34th International Conference on
model for the mouse visual cortex[J]. PLoS Comput Machine Learning. PMLR, 2017: 1126-1135.
Biol, 2022, 18(9): e1010427. [75] WANG Y Q, YAO Q M, KWOK J T, et al. Generalizing
[60] CHEN X X, CAI R C, FANG Y, et al. Motif graph from a few examples: A survey on few-shot learning[J].
neural network[J]. IEEE Trans Neural Netw Learn Syst, ACM Comput Surv, 2020, 53(3): 1-34.
2024, 35(10): 14833-14847. [76] WANG X, CHEN Y D, ZHU W W. A survey on
[61] MONGILLO G. Hebbian learning[A]//SEEL N M, ed. curriculum learning[J]. IEEE Trans Pattern Anal Mach
Encyclopedia of the Sciences of Learning. Boston, MA: Intell, 2022, 44(9): 4555-4576.
Springer, 2012. [77] GUI J, CHEN T, ZHANG J, et al. A survey on self-
[62] OJA E. Simplified neuron model as a principal supervised learning: algorithms, applications, and future
component analyzer[J]. J Math Biol, 1982, 15(3): 267- trends[J]. IEEE Trans Pattern Anal Mach Intell, 2024,
273. 46(12): 9052-9071.
[63] IZHIKEVICH E M, DESAI N S. Relating STDP to [78] CHAKRABORTY T, REDDY U K S, NAIK S M, et al.
BCM[J]. Neural Comput, 2003, 15(7): 1511-1523. Ten years of generative adversarial nets (GANs): a
[64] ANDRADE-TALAVERA Y, FISAHN A, survey of the state-of-the-art[J]. Mach Learn Sci
RODRÍGUEZ-MORENO A. Timing to be precise? An Technol, 2024, 5(1): 011001.
overview of spike timing-dependent plasticity, brain [79] FURLANELLO T, LIPTON Z C, TSCHANNEN M, et
rhythmicity, and glial cells interplay within neuronal al. Born again neural networks[C]//Proceedings of the
circuits[J]. Mol Psychiatry, 2023, 28(6): 2177-2188. 35th International Conference on Machine Learning.
[65] ZHANG T L, CHENG X, JIA S C, et al. Self- PMLR, 2018: 1607-1616.
backpropagation of synaptic modifications elevates the [80] ZARE M, KEBRIA P M, KHOSRAVI A, et al. A
efficiency of spiking and artificial neural networks[J]. survey of imitation learning: Algorithms, recent
Sci Adv, 2021, 7(43): eabh0146. developments, and challenges[J]. IEEE Trans Cybern,
[66] DELLAFERRERA G, KREIMAN G. Error-driven input 2024, 54(12): 7173-7186.
modulation: solving the credit assignment problem [81] GHASEMI M, MOOSAVI A H, EBRAHIMI D. A
without a backward pass[C]//Proceedings of the 39th comprehensive survey of reinforcement learning: From
International Conference on Machine Learning. PMLR, algorithms to practical challenges[EB/OL] arXiv:
2022: 4937-4955. 2411.18892, 2025[2025-08-04]. https://arxiv.org/abs/
[67] LILLICRAP T P, COWNDEN D, TWEED D B, et al. 2411.18892.

179