Heterogeneous breast phantom for computed tomography and magnetic resonance imaging

PLoS One. 2023 Apr 13;18(4):e0284531. doi: 10.1371/journal.pone.0284531. eCollection 2023.

Abstract

In this article, a heterogeneous multimodal anthropomorphic breast phantom with carcinoma is introduced to meet the response of the natural breast tissue when imaged using ionizing and non-ionizing machines. The skin, adipose, fibroglandular, pectoral muscle, and carcinoma tissue were mimicked. A T1-weighted breast magnetic resonance image with BI-RADS I tissue segmentation was used for molds creation. The tissue-mimicking materials (TMMs) were tailored in terms of their elemental composition weight fractions and their response to ionization radiation parameters. These are the mass attenuation coefficient (MAC), electron density (ne) and effective atomic number (Zeff). The behaviour of the TMMs, when exposed to a wide range of ionization radiation energy, was investigated analytically and numerically using X-COM. The achieved results showed an excellent agreement with the corresponding properties of the natural breast elemental compositions as reported by the International Commission on Radiation Units and Measurements (ICRU). The MAC of the TMMs and the ICRU-based breast tissue were found to be consistent. The maximum percentage of error in ne and Zeff amounts to only 2.93% and 5.76%, respectively. For non-ionizing imaging, the TMMs were characterized in term of T1 and T2 relaxation times. Using our preclinical MRI unit, the TMMs relaxation times were measured and compared to the natural tissue. The fabricated phantom was validated experimentally using CT, MRI, and Mammographic machines. The achieved images of the TMMs were in alignment with the real tissue in terms of CT HU values and grayscale colors. T1W and T2W images on MRI revealed the expected contrast between TMMs as in natural tissue.

MeSH terms

  • Magnetic Resonance Imaging / methods
  • Mammography*
  • Phantoms, Imaging
  • Tomography, X-Ray Computed* / methods

Associated data

  • figshare/10.6084/m9.figshare.22105085
  • figshare/10.6084/m9.figshare.c.6436949

Grants and funding

The author(s) received no specific funding for this work.