Automated breast positioning evaluation of screening mammograms in the UK

Congress:

ECR 2021

Poster Number:

C-13781

Type:

Scientific Exhibit

Keywords:

Artificial Intelligence, Breast, Radiographers, Mammography, Screening, Technology assessment, Education and training, Image verification, Quality assurance

Authors:

H.-M. Gilroy¹, M. L. Hill¹, A. Chan¹, M. Halling-Brown², R. P. Highnam¹; ¹Wellington/NZ, ²Guildford/UK

DOI:

10.26044/ecr2021/C-13781

DOI-Link:

https://dx.doi.org/10.26044/ecr2021/C-13781

Personal information and conflict of interest

H-M. Gilroy: Employee: Volpara Health Limited

M. L. Hill: Consultant: Volpara Health Limited

A. Chan: Employee: Volpara Health Limited

M. Halling-Brown: Grant Recipient: CRUK

R. P. Highnam: CEO: Volpara Health Limited

Fig 1: Study image selection flow chart.

Fig 2: Summary of age, breast volume, volumetric breast density, and compression...

Fig 3: Percentage of images scored as perfect, good, moderate, and inadequate by...

Fig 4: Percentage of medio-lateral (MLO) and cranio-caudal (CC) images that failed...

Fig 5: (a) Original view assessed as having a convex pectoral muscle and scored...

Fig 6: (a) Original view with a body part obscuring breast tissue and scored Perfect...

Fig 7: (a) Original view scored Perfect by Volpara PGMI with a Posterior Nipple Length...

Fig 8: (a) Original view scored Inadequate by Volpara PGMI (b) repeat scored Perfect...

Fig 9: (a) Original view scored Inadequate by Volpara PGMI (b) repeat scored Perfect...

Fig 10: (a) Original view and (b) repeat view both scored Moderate by Volpara PGMI.

Fig 11: (a) Original view scored Good and (b) repeat view scored Inadequate by Volpara...

Fig 12: (a) Original view scored Good and (b) repeat scored Inadequate by Volpara PGMI.