# Statistical evaluation of phase fractions in a MoSiBTiC alloy by machine-learning-assisted segmentation of scanning electron microscopy images

https://mdr.nims.go.jp/datasets/8c8556ed-8095-48f8-b5df-e33d3040a159

## File

- [1-s2.0-S2238785426020132-main.pdf](https://mdr.nims.go.jp/filesets/07d3a545-a3e3-4ca2-947e-6a1000fa62fc/download) ([Detail](https://mdr.nims.go.jp/filesets/07d3a545-a3e3-4ca2-947e-6a1000fa62fc.md))

## Id

8c8556ed-8095-48f8-b5df-e33d3040a159

## Local identifier



## Visibility

open_to_public

## State

published

## Created at

2026-09-15T23:45:56.415337Z

## Updated at

2026-09-16T01:47:16.560347Z

## Published at

2026-09-16T03:27:18.035480Z

## Doi



## First published url

https://doi.org/10.1016/j.jmrt.2026.07.282

## Date published

2026-07-30

## Recorded date published

2026-9

## Resource type

journal_article

## Manuscript type

vor

## Collection



## Title

- title: Statistical evaluation of phase fractions in a MoSiBTiC alloy by machine-learning-assisted
    segmentation of scanning electron microscopy images
  title_type: original
  lang: en

## Description

- description: This study examines how many SEM-BSE images are needed for reliable
    phase-fraction analysis of a four-phase MoSiBTiC alloy. A machine-learning-assisted
    segmentation workflow achieved 98.4% pixel accuracy and reduced the processing
    time for 100 images from about 300 days to one working day. Bootstrap analysis
    showed that using only a few images can cause errors above 15%, especially for
    TiC. About 60 images at 2000× are needed to limit sampling uncertainty in fracture-toughness
    estimates to the level of experimental error.
  description_type: abstract
  lang: und

## Creator

- name: Chihana Kudo
  role: author
- name: Masahiko Demura
  role: author
  orcid: https://orcid.org/0000-0002-7308-3041
  organization: National Institute for Materials Science
- name: Akihiro Endo
  role: author
  orcid: https://orcid.org/0009-0009-7175-6079
  organization: National Institute for Materials Science
- name: Kenji Nagata
  role: author
  orcid: https://orcid.org/0000-0001-9894-4461
  organization: National Institute for Materials Science
- name: Kyosuke Yoshimi
  role: author

## Contact agent



## Publisher

organization: Elsevier BV

## Managing organization



## Keyword

- subject: MoSiBTiC alloy
  schema: not_defined
- subject: Scanning electron microscopy
  schema: not_defined
- subject: Image segmentation
  schema: not_defined
- subject: Machine learning
  schema: not_defined
- subject: Bootstrap resampling
  schema: not_defined
- subject: Fracture toughness
  schema: not_defined

## Rights

- identifier: https://creativecommons.org/licenses/by/4.0/
  date_licensed: 2026-07-29

## Other identifier(s)



## Data origin

- data_origin_type: other

## Embargo



## Journal

- title: Journal of Materials Research and Technology
  issn: '22387854'
  volume: '44'
  start_page: 1373
  end_page: 1384

## Conference



## Related item



## Funding

- funder_name: Tohoku University
- identifier: JP21H04606
  funder_name: Japan Society for the Promotion of Science
- identifier: JPMXP1122684766
  funder_name: Government of Japan Ministry of Education Culture Sports Science and
    Technology

## Instrument



## Instrument operator



## Instrument managing organization



## Measurement method



## Specimen



## Chemical composition



## Structure for specimen



## Structural feature for specimen



## Specific property for specimen



## Process for specimen treatment



## Computational method



## Energy level/transition state



## Software



## Custom property



## Fileset

- id: 07d3a545-a3e3-4ca2-947e-6a1000fa62fc
  filename: 1-s2.0-S2238785426020132-main.pdf
  content_type: application/pdf
  size: 8914298
  md5: 3fe986f5ecb6190a0e6b83eee793a7c7

## Thumbnail

fileset_id: 07d3a545-a3e3-4ca2-947e-6a1000fa62fc
filename: 1-s2.0-S2238785426020132-main.pdf