Automatic microscopic cell counting by use of deeply-supervised density regression model

Shenghua He, Kyaw T. Minn, Lilianna Solnica-krezel, Mark Anastasio, Hua Li

Research output: Contribution to conferencePaper

Abstract

Accurately counting cells in microscopic images is important for medical diagnoses and biological studies, but manual cell counting is very tedious, time-consuming, and prone to subjective errors, and automatic counting can be less accurate than desired. To improve the accuracy of automatic cell counting, we propose here a novel method that employs deeply-supervised density regression. A fully convolutional neural network (FCNN) serves as the primary FCNN for density map regression. Innovatively, a set of auxiliary FCNNs are employed to provide additional supervision for learning the intermediate layers of the primary CNN to improve network performance. In addition, the primary CNN is designed as a concatenating framework to integrate multi-scale features through shortcut connections in the network, which improves the granularity of the features extracted from the intermediate CNN layers and further supports the final density map estimation. The experimental results on immunofluorescent images of human embryonic stem cells demonstrate the superior performance of the proposed method over other state-of-the-art methods.
Original languageEnglish (US)
Pages19
DOIs
StatePublished - Mar 18 2019
Externally publishedYes
EventDigital Pathology - San Diego, United States
Duration: Feb 16 2019Feb 21 2019

Conference

ConferenceDigital Pathology
Period2/16/192/21/19

Fingerprint

Neural networks
Network performance
Stem cells

Cite this

He, S., Minn, K. T., Solnica-krezel, L., Anastasio, M., & Li, H. (2019). Automatic microscopic cell counting by use of deeply-supervised density regression model. 19. Paper presented at Digital Pathology, . https://doi.org/10.1117/12.2513045

Automatic microscopic cell counting by use of deeply-supervised density regression model. / He, Shenghua; Minn, Kyaw T.; Solnica-krezel, Lilianna; Anastasio, Mark; Li, Hua.

2019. 19 Paper presented at Digital Pathology, .

Research output: Contribution to conferencePaper

He, S, Minn, KT, Solnica-krezel, L, Anastasio, M & Li, H 2019, 'Automatic microscopic cell counting by use of deeply-supervised density regression model' Paper presented at Digital Pathology, 2/16/19 - 2/21/19, pp. 19. https://doi.org/10.1117/12.2513045
He, Shenghua ; Minn, Kyaw T. ; Solnica-krezel, Lilianna ; Anastasio, Mark ; Li, Hua. / Automatic microscopic cell counting by use of deeply-supervised density regression model. Paper presented at Digital Pathology, .
@conference{11a65e98ad834c708d8e6bb8fde3c166,
title = "Automatic microscopic cell counting by use of deeply-supervised density regression model",
abstract = "Accurately counting cells in microscopic images is important for medical diagnoses and biological studies, but manual cell counting is very tedious, time-consuming, and prone to subjective errors, and automatic counting can be less accurate than desired. To improve the accuracy of automatic cell counting, we propose here a novel method that employs deeply-supervised density regression. A fully convolutional neural network (FCNN) serves as the primary FCNN for density map regression. Innovatively, a set of auxiliary FCNNs are employed to provide additional supervision for learning the intermediate layers of the primary CNN to improve network performance. In addition, the primary CNN is designed as a concatenating framework to integrate multi-scale features through shortcut connections in the network, which improves the granularity of the features extracted from the intermediate CNN layers and further supports the final density map estimation. The experimental results on immunofluorescent images of human embryonic stem cells demonstrate the superior performance of the proposed method over other state-of-the-art methods.",
author = "Shenghua He and Minn, {Kyaw T.} and Lilianna Solnica-krezel and Mark Anastasio and Hua Li",
year = "2019",
month = "3",
day = "18",
doi = "10.1117/12.2513045",
language = "English (US)",
pages = "19",
note = "Digital Pathology ; Conference date: 16-02-2019 Through 21-02-2019",

}

TY - CONF

T1 - Automatic microscopic cell counting by use of deeply-supervised density regression model

AU - He, Shenghua

AU - Minn, Kyaw T.

AU - Solnica-krezel, Lilianna

AU - Anastasio, Mark

AU - Li, Hua

PY - 2019/3/18

Y1 - 2019/3/18

N2 - Accurately counting cells in microscopic images is important for medical diagnoses and biological studies, but manual cell counting is very tedious, time-consuming, and prone to subjective errors, and automatic counting can be less accurate than desired. To improve the accuracy of automatic cell counting, we propose here a novel method that employs deeply-supervised density regression. A fully convolutional neural network (FCNN) serves as the primary FCNN for density map regression. Innovatively, a set of auxiliary FCNNs are employed to provide additional supervision for learning the intermediate layers of the primary CNN to improve network performance. In addition, the primary CNN is designed as a concatenating framework to integrate multi-scale features through shortcut connections in the network, which improves the granularity of the features extracted from the intermediate CNN layers and further supports the final density map estimation. The experimental results on immunofluorescent images of human embryonic stem cells demonstrate the superior performance of the proposed method over other state-of-the-art methods.

AB - Accurately counting cells in microscopic images is important for medical diagnoses and biological studies, but manual cell counting is very tedious, time-consuming, and prone to subjective errors, and automatic counting can be less accurate than desired. To improve the accuracy of automatic cell counting, we propose here a novel method that employs deeply-supervised density regression. A fully convolutional neural network (FCNN) serves as the primary FCNN for density map regression. Innovatively, a set of auxiliary FCNNs are employed to provide additional supervision for learning the intermediate layers of the primary CNN to improve network performance. In addition, the primary CNN is designed as a concatenating framework to integrate multi-scale features through shortcut connections in the network, which improves the granularity of the features extracted from the intermediate CNN layers and further supports the final density map estimation. The experimental results on immunofluorescent images of human embryonic stem cells demonstrate the superior performance of the proposed method over other state-of-the-art methods.

U2 - 10.1117/12.2513045

DO - 10.1117/12.2513045

M3 - Paper

SP - 19

ER -