Automated correlative segmentation of large Transmission X-ray Microscopy (TXM) tomograms using deep learning

C. Shashank Kaira, Xiaogang Yang, Vincent De Andrade, Francesco De Carlo, William Scullin, Doga Gursoy, Nikhilesh Chawla

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

A unique correlative approach for automated segmentation of large 3D nanotomography datasets obtained using Transmission X-ray Microscopy (TXM) in an Al-Cu alloy has been introduced. Automated segmentation using a Convolutional Neural Network (CNN) architecture based on a deep learning approach was employed. This extremely versatile technique is capable of emulating the manual segmentation process effectively. Coupling this technique with post-scanning SEM imaging ensured precise estimation of 3D morphological parameters from nanotomography. The segmentation process as well as subsequent analysis was expedited by several orders of magnitude. Quantitative comparison between segmentation performed manually and using the CNN architecture established the accuracy of this automated technique. Its ability to robustly process ultra-large volumes of data in relatively small time frames can exponentially accelerate tomographic data analysis, possibly opening up novel avenues for performing 4D characterization experiments with finer time steps.

Original languageEnglish (US)
Pages (from-to)203-210
Number of pages8
JournalMaterials Characterization
Volume142
DOIs
StatePublished - Aug 1 2018

Fingerprint

Network architecture
learning
Microscopic examination
microscopy
Neural networks
X rays
x rays
Scanning
Imaging techniques
Scanning electron microscopy
scanning electron microscopy
scanning
Experiments
Deep learning

Keywords

  • Aluminum alloys
  • Deep learning
  • Precipitates
  • Segmentation
  • Transmission X-ray Microscopy (TXM)

ASJC Scopus subject areas

  • Materials Science(all)
  • Condensed Matter Physics
  • Mechanics of Materials
  • Mechanical Engineering

Cite this

Automated correlative segmentation of large Transmission X-ray Microscopy (TXM) tomograms using deep learning. / Shashank Kaira, C.; Yang, Xiaogang; De Andrade, Vincent; De Carlo, Francesco; Scullin, William; Gursoy, Doga; Chawla, Nikhilesh.

In: Materials Characterization, Vol. 142, 01.08.2018, p. 203-210.

Research output: Contribution to journalArticle

Shashank Kaira, C. ; Yang, Xiaogang ; De Andrade, Vincent ; De Carlo, Francesco ; Scullin, William ; Gursoy, Doga ; Chawla, Nikhilesh. / Automated correlative segmentation of large Transmission X-ray Microscopy (TXM) tomograms using deep learning. In: Materials Characterization. 2018 ; Vol. 142. pp. 203-210.
@article{682247158f5f4fbd9b5b73286685ac13,
title = "Automated correlative segmentation of large Transmission X-ray Microscopy (TXM) tomograms using deep learning",
abstract = "A unique correlative approach for automated segmentation of large 3D nanotomography datasets obtained using Transmission X-ray Microscopy (TXM) in an Al-Cu alloy has been introduced. Automated segmentation using a Convolutional Neural Network (CNN) architecture based on a deep learning approach was employed. This extremely versatile technique is capable of emulating the manual segmentation process effectively. Coupling this technique with post-scanning SEM imaging ensured precise estimation of 3D morphological parameters from nanotomography. The segmentation process as well as subsequent analysis was expedited by several orders of magnitude. Quantitative comparison between segmentation performed manually and using the CNN architecture established the accuracy of this automated technique. Its ability to robustly process ultra-large volumes of data in relatively small time frames can exponentially accelerate tomographic data analysis, possibly opening up novel avenues for performing 4D characterization experiments with finer time steps.",
keywords = "Aluminum alloys, Deep learning, Precipitates, Segmentation, Transmission X-ray Microscopy (TXM)",
author = "{Shashank Kaira}, C. and Xiaogang Yang and {De Andrade}, Vincent and {De Carlo}, Francesco and William Scullin and Doga Gursoy and Nikhilesh Chawla",
year = "2018",
month = "8",
day = "1",
doi = "10.1016/j.matchar.2018.05.053",
language = "English (US)",
volume = "142",
pages = "203--210",
journal = "Materials Characterization",
issn = "1044-5803",
publisher = "Elsevier Inc.",

}

TY - JOUR

T1 - Automated correlative segmentation of large Transmission X-ray Microscopy (TXM) tomograms using deep learning

AU - Shashank Kaira, C.

AU - Yang, Xiaogang

AU - De Andrade, Vincent

AU - De Carlo, Francesco

AU - Scullin, William

AU - Gursoy, Doga

AU - Chawla, Nikhilesh

PY - 2018/8/1

Y1 - 2018/8/1

N2 - A unique correlative approach for automated segmentation of large 3D nanotomography datasets obtained using Transmission X-ray Microscopy (TXM) in an Al-Cu alloy has been introduced. Automated segmentation using a Convolutional Neural Network (CNN) architecture based on a deep learning approach was employed. This extremely versatile technique is capable of emulating the manual segmentation process effectively. Coupling this technique with post-scanning SEM imaging ensured precise estimation of 3D morphological parameters from nanotomography. The segmentation process as well as subsequent analysis was expedited by several orders of magnitude. Quantitative comparison between segmentation performed manually and using the CNN architecture established the accuracy of this automated technique. Its ability to robustly process ultra-large volumes of data in relatively small time frames can exponentially accelerate tomographic data analysis, possibly opening up novel avenues for performing 4D characterization experiments with finer time steps.

AB - A unique correlative approach for automated segmentation of large 3D nanotomography datasets obtained using Transmission X-ray Microscopy (TXM) in an Al-Cu alloy has been introduced. Automated segmentation using a Convolutional Neural Network (CNN) architecture based on a deep learning approach was employed. This extremely versatile technique is capable of emulating the manual segmentation process effectively. Coupling this technique with post-scanning SEM imaging ensured precise estimation of 3D morphological parameters from nanotomography. The segmentation process as well as subsequent analysis was expedited by several orders of magnitude. Quantitative comparison between segmentation performed manually and using the CNN architecture established the accuracy of this automated technique. Its ability to robustly process ultra-large volumes of data in relatively small time frames can exponentially accelerate tomographic data analysis, possibly opening up novel avenues for performing 4D characterization experiments with finer time steps.

KW - Aluminum alloys

KW - Deep learning

KW - Precipitates

KW - Segmentation

KW - Transmission X-ray Microscopy (TXM)

UR - http://www.scopus.com/inward/record.url?scp=85048516553&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85048516553&partnerID=8YFLogxK

U2 - 10.1016/j.matchar.2018.05.053

DO - 10.1016/j.matchar.2018.05.053

M3 - Article

VL - 142

SP - 203

EP - 210

JO - Materials Characterization

JF - Materials Characterization

SN - 1044-5803

ER -