Laser & Optoelectronics Progress, Volume. 60, Issue 14, 1410015(2023)
Nucleus-Guided Cell Segmentation Method for Brightfield Micrographs
Figures & Tables(13)
Fig. 1. Flowchart of the implementation of a nucleus-guided cell segmentation method for brightfield micrographs
Fig. 2. Relationship between the σ2 value of Gaussian filter and standard deviation of the image fdg
Fig. 3. Contrast images after double Gaussian filtering and top-hat transformation. (a) Fluorescent nuclei; (b) brightfield cells; (c) images after double Gaussian filtering; (d) images after top-hat transformation
Fig. 4. Local cell segmentation image. (a) Fluorescent nucleus images; (b) brightfield cell images; (c) images after 2D OTSU threshold segmentation; (d) images after binary morphology processing
Fig. 5. Global segmentation results for cell. (a) Fluorescent nucleus image; (b) brightfield cell image; (c) local segmentation result; (d) global segmentation result; (e) fused segmentation result
Fig. 7. Watershed fusion segmentation result. (a) Fluorescent nucleus image; (b) brightfield cell image; (c) result after fusion of local segmentation and global segmentation; (d) result after watershed segmentation; (e) result of superimposing detection result on the original cell image using the Canny operator after the edge detection of the Fig.7(d)
Fig. 8. Results of the first set of cell image segmentation experiment. (a) Standard cell segmentation; (b) Jaccard algorithm; (c) dLoGH-Zhang algorithm; (d) sEGT algorithm; (e) Flight algorithm; (f) proposed algorithm
Fig. 9. Results of the second set of cell image segmentation experiment. (a) Standard cell segmentation; (b) Jaccard algorithm; (c) dLoGH-Zhang algorithm; (d) sEGT algorithm; (e) Flight algorithm; (f) proposed algorithm
Fig. 10. Results of the third set of cell image segmentation experiment. (a) Standard cell segmentation; (b) Jaccard algorithm; (c) dLoGH-Zhang algorithm; (d) sEGT algorithm; (e) Flight algorithm; (f) proposed algorithm
Fig. 11. Cell segmentation results. (a) Fluorescent cell nucleus; (b) bright field cell; (c) standard cell segmentation result; (d) segmentation result of the proposed algorithm
Table 1. Comparison of segmentation results of different algorithms
View tableTable 1. Comparison of segmentation results of different algorithms
Algorithm Precision Recall F-score Time /s Jaccard 0.817 0.681 0.743 1.55 dLoGH-Zhang 0.772 0.826 0.796 23.36 sEGT 0.785 0.824 0.802 0.40 Flight 0.804 0.861 0.830 0.12 Proposed algorithm 0.960 0.984 0.971 5.31
Table 2. Comparison of cell image counting results in three groups
View tableTable 2. Comparison of cell image counting results in three groups
Algorithm Group GT EN NTP NFP NFN Precision Recall F-score Time /s Jaccard 1 32 29 24 5 8 0.828 0.750 0.787 1.55 2 33 28 21 7 12 0.750 0.636 0.689 1.56 3 33 26 23 3 10 0.885 0.697 0.780 1.56 dLoGH-Zhang 1 32 36 26 10 6 0.722 0.813 0.765 23.22 2 33 40 27 13 6 0.675 0.818 0.740 23.14 3 33 35 29 6 4 0.829 0.879 0.853 23.16 sEGT 1 32 32 28 4 4 0.875 0.875 0.875 0.57 2 33 36 26 10 7 0.722 0.788 0.754 0.52 3 33 31 26 5 7 0.839 0.788 0.813 0.52 Flight 1 32 38 26 12 6 0.684 0.813 0.743 0.12 2 33 33 28 5 5 0.848 0.848 0.848 0.12 3 33 38 30 8 3 0.789 0.909 0.845 0.12 Proposed algorithm 1 32 32 30 2 2 0.938 0.938 0.938 4.93 2 33 33 33 0 0 1.000 1.000 1.000 5.21 3 33 34 33 1 0 0.971 1.000 0.985 5.23
Tools
Get Citation
Copy Citation Text
Yidong Wang, Yongzhao Du, Ling Li, Yuqing Fu, Yong Diao. Nucleus-Guided Cell Segmentation Method for Brightfield Micrographs[J]. Laser & Optoelectronics Progress, 2023, 60(14): 1410015
Paper Information
Category: Image Processing
Received: Sep. 1, 2022
Accepted: Sep. 23, 2022
Published Online: Jul. 14, 2023
The Author Email: Du Yongzhao (yongzhaodu@hqu.edu.cn)
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20