PCA-Net: a heart segmentation model based on the meta-learning method

Abstract

In order to effectively prevent and treat heart-based diseases, the study of precise segmentation of heart parts is particularly important. The heart is divided into four parts: the left and right ventricles and the left and right atria, and the left main trunk is more important, thus the left ventricular muscle (LV-MYO), which is located in the middle part of the heart, has become the object of many researches. Deep learning medical image segmentation methods become the main means of image analysis and processing at present, but the deep learning methods based on traditional convolutional neural network (CNN) are not suitable for segmenting organs with few labels and few samples like the heart, while the meta-learning methods are able to solve the above problems and achieve better results in the direction of heart segmentation. Since the LV-MYO is wrapped in the left ventricular blood pool (LV-BP), this paper proposes a new model for heart segmentation: principle component analysis network (PCA-Net). Specifically, we redesign the coding structure of Q-Net and make improvements in threshold extraction. Experimental results confirm that PCA-Net effectively improves the accuracy of segmenting LV-MYO and LV-BP sites on the CMR dataset, and is validated on another publicly available dataset, ABD, where the results outperform other state-of-the-art (SOTA) methods.

References

[[1]] YU Z, HAN S. 3D medical image segmentation based on multi-scale MPU-Net[EB/OL]. (2023-07-11) [2023-10-12]. https://arxiv.org/abs/2307.05799.

[[2]] Liu T, Lu Y, Zhang Y, et al. . A bone segmentation method based on multi-scale features fuse U2Net and improved Dice loss in CT image process. Biomedical signal processing and control, 2022, 77: 103813. J]

[[3]] Long J, Shelhamer E, Darrell T. Fully convolutional networks for semantic segmentation. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, June 8–10, 2015, Boston, MA, USA, 2015 New York IEEE 3431-3440 [C]

[[4]] Ronneberger O, Fischer P, Brox T. U-net: convolutional networks for biomedical image segmentation. Medical Image Computing and Computer-Assisted Intervention (MICCAI 2015), October 5–9, 2015, Munich, Germany, 2015 Heidelberg Springer International Publishing 234-241. C]

[[5]] OKTAY O, SCHLEMPER J, FOLGOC L L, et al. Attention U-Net: learning where to look for the pancreas[EB/OL]. (2018-04-11) [2023-10-12]. https://arxiv.org/abs/1804.03999.

[[6]] Jha D, Smedsrud P H, Riegler M A, et al. . Resunet++: an advanced architecture for medical image segmentation. 2019 IEEE International Symposium on Multimedia (ISM), December 9–11, 2019, San Diego, CA, USA, 2019 New York IEEE 2225-2255 [C]

[[7]] Hu J, Shen L, Sun G. Squeeze-and-excitation networks. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, June 18–22, 2018, Salt Lake City, UT, USA, 2018 New York IEEE 7132-7141 [C]

[[8]] Wang Y, Liang B, Ding M, et al. . Dense semantic labeling with atrous spatial pyramid pooling and decoder for high-resolution remote sensing imagery. Remote sensing, 2018, 11(1): 20. Bibcode: , J]

[[9]] Huang H, Lin L, Tong R, et al. . Unet 3+: a full-scale connected Unet for medical image segmentation. 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), May 4–8, 2020, Virtual, 2020 New York IEEE 1055-1059 [C]

[[10]] Kattenborn T, Leitloff J, Schiefer F, et al. . Review on convolutional neural networks (CNN) in vegetation remote sensing. ISPRS journal of photogrammetry and remote sensing, 2021, 173: 24-49. Bibcode: , J]

[[11]] CHEN J, LU Y, YU Q, et al. Transunet: transformers make strong encoders for medical image segmentation[EB/OL]. (2021-02-08) [2023-10-12]. https://arxiv.org/abs/2102.04306.

[[12]] TAUD H, MAS J F. Multilayer perceptron (MLP)[J]. Geomatic approaches for modeling land change scenarios, 2018: 451–455.

[[13]] Valanarasu J M J, Patel V M. Unext: MLP-based rapid medical image segmentation network. International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI 2022), September 18–22, 2022, Singapore, 2022 Cham Springer Nature Switzerland 23-33 [C]

[[14]] Yu J, Gao H, Chen Y, et al. . Deep object detector with attentional spatiotemporal LSTM for space human-robot interaction. IEEE transactions on human-machine systems, 2022, 52(4): 784-793. J]

[[15]] Yu J, Gao H, Zhou D, et al. . Deep temporal model-based identity-aware hand detection for space human-robot interaction. IEEE transactions on cybernetics, 2021, 52(12): 13738-13751. J]

[[16]] Zhu F, Cui J, Zhu B, et al. . Semantic segmentation of urban street scene images based on improved U-Net network. Optoelectronics letters, 2023, 19(3): 179-185. Bibcode: , J]

[[17]] Çiçek Ö, Abdulkadir A, Lienkamp S S, et al. . 3D U-Net: learning dense volumetric segmentation from sparse annotation. Medical Image Computing and Computer-Assisted Intervention (MICCAI 2016), October 17–21, 2016, Athens, Greece, 2016 Heidelberg Springer International Publishing 424-432. C]

[[18]] Milletari F, Navab N, Ahmadi S A. V-net: fully convolutional neural networks for volumetric medical image segmentation. 2016 4th International Conference on 3D Vision (3DV), October 25–28, 2016, Stanford, USA, 2016 New York IEEE 565-571 [C]

[[19]] Li J, Chen Z, Zhong Y, et al. . Appearance-based gaze estimation for ASD diagnosis. IEEE transactions on cybernetics, 2022, 52(7): 6504-6517. J]

[[20]] Zhu E R, Zhao H C, Hu A F. Semi-supervised cardiac MRI image of the left ventricle segmentation algorithm based on contrastive learning. Optoelectronics letters, 2022, 18(9): 547-552. Bibcode: , J]

[[21]] HENDRYX S M, LEACH A B, HEIN P D, et al. Meta-learning initializations for image segmentation[EB/OL]. (2019-12-13) [2023-10-12]. https://arxiv.org/abs/1912.06290.

[[22]] SHEN Q, LI Y, JIN J, et al. Q-net: query-informed few-shot medical image segmentation[EB/OL]. (2022-08-24) [2023-10-12]. https://arxiv.org/abs/2208.11451.

[[23]] Liu Z, Lin W, Li X, et al. . ADNet: attention-guided deformable convolutional network for high dynamic range imaging. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, June 20–25, 2021, Nashville, TN, USA, 2021 New York IEEE 463-470 [C]

[[24]] HOWARD A G, ZHU M, CHEN B, et al. Mobilenets: efficient convolutional neural networks for mobile vision applications[EB/OL]. (2017-04-17) [2023-10-12]. https://arxiv.org/abs/1704.04861.

[[25]] Zhang X, Zhou X, Lin M, et al. . Shufflenet: an extremely efficient convolutional neural network for mobile devices. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, June 18–23, 2018, Salt Lake City, UT, USA, 2018 New York IEEE 6848-6856 [C]

[[26]] Dai Y, Gieseke F, Oehmcke S, et al. . Attentional feature fusion. Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, June 20–25, 2021, Nashville, TN, USA, 2021 New York IEEE 3560-3569 [C]

[[27]] Xu J, Zhao Y, Xu F. RDPNet: a single-path lightweight CNN with re-parameterization for CPU-type edge devices. Journal of cloud computing, 2022, 11(1): 54. J]

[[28]] ZHONG S, WEN W, QIN J. Switchable self-attention module[EB/OL]. (2022-09-13) [2023-10-12]. https://arxiv.org/abs/2209.05680.

[[29]] Zhuang X. Multivariate mixture model for cardiac segmentation from multi-sequence MRI. Medical Image Computing and Computer-Assisted Intervention (MICCAI 2016), October 17–21, 2016, Athens, Greece, 2016 Heidelberg Springer International Publishing 581-588. C]

[[30]] Kavur A E, Gezer N S, Bariş M, et al. . CHAOS challenge-combined (CT-MR) healthy abdominal organ segmentation. Medical image analysis, 2021, 69: 101950. J]