Machine Learning Performance Metrics and Diagnostic Context in Radiology

Henrik Strøm, Steven Albury, Lene Tolstrup Sørensen

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

Abstract

In this pilot study data gathered from interviewing specialists in radiology is combined with an assessment of the way machine learning metrics are used in studies of radiological work. It argues that situated context of use should be an important contributor to the design of machine learning applications in radiology. The article shows how radiologists see their professional practice as utilizing a wider range of expert knowledge than many existing studies on machine learning in radiology allow for. The article describes a case study drawn from radiology practice in a major Danish hospital and discusses a widely cited study on machine learning in radiological work. The study connects current understandings of appropriate metrics used by machine learning researchers with professional radiologists' understanding of their diagnostic work. This comparison helps identify gaps in understanding between these two communities and suggests how they might be addressed.

Original languageEnglish
Title of host publication11th CMI International Conference, 2018 : Prospects and Challenges Towards Developing a Digital Economy within the EU, PCTDDE 2018
EditorsIdongesit Williams
Number of pages6
PublisherIEEE
Publication date24 Jan 2019
Pages56-61
Article number8624718
ISBN (Print)978-1-7281-0445-4
ISBN (Electronic)978-1-7281-0444-7
DOIs
Publication statusPublished - 24 Jan 2019
Event11TH CMI INTERNATIONAL CONFERENCE - København, Denmark
Duration: 29 Nov 201830 Nov 2018
https://www.conf.cmi.aau.dk/11th+CMI+Conference+2018/

Conference

Conference11TH CMI INTERNATIONAL CONFERENCE
CountryDenmark
CityKøbenhavn
Period29/11/201830/11/2018
Internet address

Fingerprint

learning performance
diagnostic
learning
expert knowledge
community

Keywords

  • Diagnostic context
  • Machine learning
  • Metrics
  • Radiology

Cite this

Strøm, H., Albury, S., & Sørensen, L. T. (2019). Machine Learning Performance Metrics and Diagnostic Context in Radiology. In I. Williams (Ed.), 11th CMI International Conference, 2018: Prospects and Challenges Towards Developing a Digital Economy within the EU, PCTDDE 2018 (pp. 56-61). [8624718] IEEE. https://doi.org/10.1109/PCTDDE.2018.8624718
Strøm, Henrik ; Albury, Steven ; Sørensen, Lene Tolstrup. / Machine Learning Performance Metrics and Diagnostic Context in Radiology. 11th CMI International Conference, 2018: Prospects and Challenges Towards Developing a Digital Economy within the EU, PCTDDE 2018. editor / Idongesit Williams. IEEE, 2019. pp. 56-61
@inproceedings{78646ac1fdb94870a25af475f01265ac,
title = "Machine Learning Performance Metrics and Diagnostic Context in Radiology",
abstract = "In this pilot study data gathered from interviewing specialists in radiology is combined with an assessment of the way machine learning metrics are used in studies of radiological work. It argues that situated context of use should be an important contributor to the design of machine learning applications in radiology. The article shows how radiologists see their professional practice as utilizing a wider range of expert knowledge than many existing studies on machine learning in radiology allow for. The article describes a case study drawn from radiology practice in a major Danish hospital and discusses a widely cited study on machine learning in radiological work. The study connects current understandings of appropriate metrics used by machine learning researchers with professional radiologists' understanding of their diagnostic work. This comparison helps identify gaps in understanding between these two communities and suggests how they might be addressed.",
keywords = "Diagnostic context, Machine learning, Metrics, Radiology",
author = "Henrik Str{\o}m and Steven Albury and S{\o}rensen, {Lene Tolstrup}",
year = "2019",
month = "1",
day = "24",
doi = "10.1109/PCTDDE.2018.8624718",
language = "English",
isbn = "978-1-7281-0445-4",
pages = "56--61",
editor = "Idongesit Williams",
booktitle = "11th CMI International Conference, 2018",
publisher = "IEEE",
address = "United States",

}

Strøm, H, Albury, S & Sørensen, LT 2019, Machine Learning Performance Metrics and Diagnostic Context in Radiology. in I Williams (ed.), 11th CMI International Conference, 2018: Prospects and Challenges Towards Developing a Digital Economy within the EU, PCTDDE 2018., 8624718, IEEE, pp. 56-61, 11TH CMI INTERNATIONAL CONFERENCE, København, Denmark, 29/11/2018. https://doi.org/10.1109/PCTDDE.2018.8624718

Machine Learning Performance Metrics and Diagnostic Context in Radiology. / Strøm, Henrik; Albury, Steven ; Sørensen, Lene Tolstrup.

11th CMI International Conference, 2018: Prospects and Challenges Towards Developing a Digital Economy within the EU, PCTDDE 2018. ed. / Idongesit Williams. IEEE, 2019. p. 56-61 8624718.

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

TY - GEN

T1 - Machine Learning Performance Metrics and Diagnostic Context in Radiology

AU - Strøm, Henrik

AU - Albury, Steven

AU - Sørensen, Lene Tolstrup

PY - 2019/1/24

Y1 - 2019/1/24

N2 - In this pilot study data gathered from interviewing specialists in radiology is combined with an assessment of the way machine learning metrics are used in studies of radiological work. It argues that situated context of use should be an important contributor to the design of machine learning applications in radiology. The article shows how radiologists see their professional practice as utilizing a wider range of expert knowledge than many existing studies on machine learning in radiology allow for. The article describes a case study drawn from radiology practice in a major Danish hospital and discusses a widely cited study on machine learning in radiological work. The study connects current understandings of appropriate metrics used by machine learning researchers with professional radiologists' understanding of their diagnostic work. This comparison helps identify gaps in understanding between these two communities and suggests how they might be addressed.

AB - In this pilot study data gathered from interviewing specialists in radiology is combined with an assessment of the way machine learning metrics are used in studies of radiological work. It argues that situated context of use should be an important contributor to the design of machine learning applications in radiology. The article shows how radiologists see their professional practice as utilizing a wider range of expert knowledge than many existing studies on machine learning in radiology allow for. The article describes a case study drawn from radiology practice in a major Danish hospital and discusses a widely cited study on machine learning in radiological work. The study connects current understandings of appropriate metrics used by machine learning researchers with professional radiologists' understanding of their diagnostic work. This comparison helps identify gaps in understanding between these two communities and suggests how they might be addressed.

KW - Diagnostic context

KW - Machine learning

KW - Metrics

KW - Radiology

UR - http://www.scopus.com/inward/record.url?scp=85062690890&partnerID=8YFLogxK

U2 - 10.1109/PCTDDE.2018.8624718

DO - 10.1109/PCTDDE.2018.8624718

M3 - Article in proceeding

SN - 978-1-7281-0445-4

SP - 56

EP - 61

BT - 11th CMI International Conference, 2018

A2 - Williams, Idongesit

PB - IEEE

ER -

Strøm H, Albury S, Sørensen LT. Machine Learning Performance Metrics and Diagnostic Context in Radiology. In Williams I, editor, 11th CMI International Conference, 2018: Prospects and Challenges Towards Developing a Digital Economy within the EU, PCTDDE 2018. IEEE. 2019. p. 56-61. 8624718 https://doi.org/10.1109/PCTDDE.2018.8624718