PMID- 37231288
OWN - NLM
STAT- MEDLINE
DCOM- 20230809
LR  - 20230810
IS  - 1618-727X (Electronic)
IS  - 0897-1889 (Print)
IS  - 0897-1889 (Linking)
VI  - 36
IP  - 4
DP  - 2023 Aug
TI  - A Lightweight and Robust Framework for Circulating Genetically Abnormal Cells 
      (CACs) Identification Using 4-Color Fluorescence In Situ Hybridization (FISH) 
      Image and Deep Refined Learning.
PG  - 1687-1700
LID - 10.1007/s10278-023-00843-8 [doi]
AB  - Circulating genetically abnormal cells (CACs) constitute an important biomarker 
      for cancer diagnosis and prognosis. This biomarker offers high safety, low cost, 
      and high repeatability, which can serve as a key reference in clinical diagnosis. 
      These cells are identified by counting fluorescence signals using 4-color 
      fluorescence in situ hybridization (FISH) technology, which has a high level of 
      stability, sensitivity, and specificity. However, there are some challenges in 
      CACs identification, due to the difference in the morphology and intensity of 
      staining signals. In this concern, we developed a deep learning network 
      (FISH-Net) based on 4-color FISH image for CACs identification. Firstly, a 
      lightweight object detection network based on the statistical information of 
      signal size was designed to improve the clinical detection rate. Secondly, the 
      rotated Gaussian heatmap with a covariance matrix was defined to standardize the 
      staining signals with different morphologies. Then, the heatmap refinement model 
      was proposed to solve the fluorescent noise interference of 4-color FISH image. 
      Finally, an online repetitive training strategy was used to improve the model's 
      feature extraction ability for hard samples (i.e., fracture signal, weak signal, 
      and adjacent signals). The results showed that the precision was superior to 96%, 
      and the sensitivity was higher than 98%, for fluorescent signal detection. 
      Additionally, validation was performed using the clinical samples of 853 patients 
      from 10 centers. The sensitivity was 97.18% (CI 96.72-97.64%) for CACs 
      identification. The number of parameters of FISH-Net was 2.24 M, compared to 
      36.9 M for the popularly used lightweight network (YOLO-V7s). The detection speed 
      was about 800 times greater than that of a pathologist. In summary, the proposed 
      network was lightweight and robust for CACs identification. It could greatly 
      increase the review accuracy, enhance the efficiency of reviewers, and reduce the 
      review turnaround time during CACs identification.
CI  - (c) 2023. The Author(s) under exclusive licence to Society for Imaging Informatics 
      in Medicine.
FAU - Xu, Xu
AU  - Xu X
AD  - China Academy of Information and Communications Technology, No.52, Huayuan bei 
      Road, 100191, Beijing, China.
FAU - Li, Congsheng
AU  - Li C
AD  - China Academy of Information and Communications Technology, No.52, Huayuan bei 
      Road, 100191, Beijing, China.
FAU - Lan, Xingjie
AU  - Lan X
AD  - Zhuhai Sanmed Biotech Ltd, Zhuhai, 519060, Guangdong, China.
FAU - Fan, Xianjun
AU  - Fan X
AD  - Zhuhai Sanmed Biotech Ltd, Zhuhai, 519060, Guangdong, China.
FAU - Lv, Xing
AU  - Lv X
AD  - Zhuhai Sanmed Biotech Ltd, Zhuhai, 519060, Guangdong, China.
FAU - Ye, Xin
AU  - Ye X
AD  - Zhuhai Sanmed Biotech Ltd, Zhuhai, 519060, Guangdong, China.
FAU - Wu, Tongning
AU  - Wu T
AUID- ORCID: 0000-0002-9894-9518
AD  - China Academy of Information and Communications Technology, No.52, Huayuan bei 
      Road, 100191, Beijing, China. wutongning@caict.ac.cn.
LA  - eng
PT  - Journal Article
PT  - Research Support, Non-U.S. Gov't
DEP - 20230525
PL  - United States
TA  - J Digit Imaging
JT  - Journal of digital imaging
JID - 9100529
SB  - IM
MH  - *In Situ Hybridization, Fluorescence/methods
MH  - *Image Interpretation, Computer-Assisted
PMC - PMC10406746
OTO - NOTNLM
OT  - Circulating genetically abnormal cell
OT  - Deep learning
OT  - Fluorescence in situ hybridization
OT  - Object detection
COIS- The authors declare no competing interests.
EDAT- 2023/05/26 01:05
MHDA- 2023/08/09 06:43
PMCR- 2024/08/01
CRDT- 2023/05/25 23:33
PHST- 2022/12/01 00:00 [received]
PHST- 2023/05/03 00:00 [accepted]
PHST- 2023/04/13 00:00 [revised]
PHST- 2024/08/01 00:00 [pmc-release]
PHST- 2023/08/09 06:43 [medline]
PHST- 2023/05/26 01:05 [pubmed]
PHST- 2023/05/25 23:33 [entrez]
AID - 10.1007/s10278-023-00843-8 [pii]
AID - 843 [pii]
AID - 10.1007/s10278-023-00843-8 [doi]
PST - ppublish
SO  - J Digit Imaging. 2023 Aug;36(4):1687-1700. doi: 10.1007/s10278-023-00843-8. Epub 
      2023 May 25.