Abstract

Purpose: The purpose of this study was to investigate whether the lesion risk score provided by an AI system is influenced by the selection of exposure parameters. Methods: A breast phantom which contains a lesion, was imaged with digital mammography with different imaging conditions. The tube voltage, the dose level and the anode-filter combination were varied based on an exposure obtained with automatic exposure control. The organ dose for each image was extracted from the DICOM header. The images were analyzed with an AI system, which provided a lesion risk score (suspicion for malignancy) for each exposure condition. Correlations between the lesion risk score and the exposure conditions were investigated. Results: The results of the study showed that the organ dose had a strong impact on the lesion risk score. Reducing the organ dose to a low level resulted in that the AI system no longer detected the lesion. Conclusions: Images of suboptimal quality may result in inaccurate AI system performance. In our preliminary analysis, the breast phantom and the lesion were proven to be realistic enough for being analyzed by the AI system.

Original languageEnglish
Title of host publication17th International Workshop on Breast Imaging, IWBI 2024
EditorsMaryellen L. Giger, Heather M. Whitney, Karen Drukker, Hui Li
PublisherSPIE
ISBN (Electronic)9781510680203
DOIs
Publication statusPublished - 2024
Event17th International Workshop on Breast Imaging, IWBI 2024 - Chicago, United States
Duration: 2024 Jun 92024 Jun 12

Publication series

NameProceedings of SPIE - The International Society for Optical Engineering
Volume13174
ISSN (Print)0277-786X
ISSN (Electronic)1996-756X

Conference

Conference17th International Workshop on Breast Imaging, IWBI 2024
Country/TerritoryUnited States
CityChicago
Period2024/06/092024/06/12

Subject classification (UKÄ)

  • Radiology and Medical Imaging

Free keywords

  • Artificial intelligence
  • Breast organ dose
  • Breast phantom
  • Digital mammography
  • Image quality
  • Suspicion for malignancy

Fingerprint

Dive into the research topics of 'AI lesion risk score at different exposure settings'. Together they form a unique fingerprint.

Cite this