Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Deep learning analysis of serial digital breast tomosynthesis images in a prospective cohort of breast cancer patients who received neoadjuvant chemotherapy

Förnvik, Daniel LU ; Borgquist, Signe LU ; Larsson, Måns ; Zackrisson, Sophia LU and Skarping, Ida LU orcid (2024) In European Journal of Radiology 178.
Abstract

Purpose: Different imaging tools, including digital breast tomosynthesis (DBT), are frequently used for evaluating tumor response during neoadjuvant chemotherapy (NACT). This study aimed to explore whether using artificial intelligence (AI) for serial DBT acquisitions during NACT for breast cancer can predict pathological complete response (pCR) after completion of NACT. Methods: A total of 149 women (mean age 53 years, pCR rate 22 %) with breast cancer treated with NACT at Skane University Hospital, Sweden, between 2014 and 2019, were prospectively included in this observational cohort study (ClinicalTrials.gov: NCT02306096). DBT images from both the cancer and contralateral healthy breasts acquired at three time points: pre-NACT,... (More)

Purpose: Different imaging tools, including digital breast tomosynthesis (DBT), are frequently used for evaluating tumor response during neoadjuvant chemotherapy (NACT). This study aimed to explore whether using artificial intelligence (AI) for serial DBT acquisitions during NACT for breast cancer can predict pathological complete response (pCR) after completion of NACT. Methods: A total of 149 women (mean age 53 years, pCR rate 22 %) with breast cancer treated with NACT at Skane University Hospital, Sweden, between 2014 and 2019, were prospectively included in this observational cohort study (ClinicalTrials.gov: NCT02306096). DBT images from both the cancer and contralateral healthy breasts acquired at three time points: pre-NACT, after two cycles of NACT, and after the completion of six cycles of NACT (pre-surgery) were analyzed. The deep learning AI system used to predict pCR consisted of a backbone 3D ResNet and an attention and prediction module. The GradCAM method was used to obtain insights into the model decision basis through a quantitative analysis of the importance maps on the validation set. Moreover, specific model choices were motivated by ablation studies. Results: The AI model reached an AUC of 0.83 (95% CI: 0.63–1.00) (test set). The spatial correlation of importance maps for input volumes from the same patient but at different time points was high, possibly indicating that the model focuses on the same areas during decision-making. Conclusions: We demonstrate a high discriminative performance of our algorithm for predicting pCR/non-pCR. Availability of larger datasets would permit more comprehensive training of the models and more rigorous evaluation of their prediction performance for future patients.

(Less)
Please use this url to cite or link to this publication:
author
; ; ; and
organization
publishing date
type
Contribution to journal
publication status
published
subject
keywords
Artificial intelligence, Breast cancer, Breast tomosynthesis, Imaging, Mammography, Neoadjuvant chemotherapy
in
European Journal of Radiology
volume
178
article number
111624
publisher
Elsevier
external identifiers
  • pmid:39029241
  • scopus:85198925932
ISSN
0720-048X
DOI
10.1016/j.ejrad.2024.111624
language
English
LU publication?
yes
id
18ca84d7-6be3-4c49-86b1-e321606516e8
date added to LUP
2024-09-03 15:09:19
date last changed
2024-09-03 15:09:39
@article{18ca84d7-6be3-4c49-86b1-e321606516e8,
  abstract     = {{<p>Purpose: Different imaging tools, including digital breast tomosynthesis (DBT), are frequently used for evaluating tumor response during neoadjuvant chemotherapy (NACT). This study aimed to explore whether using artificial intelligence (AI) for serial DBT acquisitions during NACT for breast cancer can predict pathological complete response (pCR) after completion of NACT. Methods: A total of 149 women (mean age 53 years, pCR rate 22 %) with breast cancer treated with NACT at Skane University Hospital, Sweden, between 2014 and 2019, were prospectively included in this observational cohort study (ClinicalTrials.gov: NCT02306096). DBT images from both the cancer and contralateral healthy breasts acquired at three time points: pre-NACT, after two cycles of NACT, and after the completion of six cycles of NACT (pre-surgery) were analyzed. The deep learning AI system used to predict pCR consisted of a backbone 3D ResNet and an attention and prediction module. The GradCAM method was used to obtain insights into the model decision basis through a quantitative analysis of the importance maps on the validation set. Moreover, specific model choices were motivated by ablation studies. Results: The AI model reached an AUC of 0.83 (95% CI: 0.63–1.00) (test set). The spatial correlation of importance maps for input volumes from the same patient but at different time points was high, possibly indicating that the model focuses on the same areas during decision-making. Conclusions: We demonstrate a high discriminative performance of our algorithm for predicting pCR/non-pCR. Availability of larger datasets would permit more comprehensive training of the models and more rigorous evaluation of their prediction performance for future patients.</p>}},
  author       = {{Förnvik, Daniel and Borgquist, Signe and Larsson, Måns and Zackrisson, Sophia and Skarping, Ida}},
  issn         = {{0720-048X}},
  keywords     = {{Artificial intelligence; Breast cancer; Breast tomosynthesis; Imaging; Mammography; Neoadjuvant chemotherapy}},
  language     = {{eng}},
  publisher    = {{Elsevier}},
  series       = {{European Journal of Radiology}},
  title        = {{Deep learning analysis of serial digital breast tomosynthesis images in a prospective cohort of breast cancer patients who received neoadjuvant chemotherapy}},
  url          = {{http://dx.doi.org/10.1016/j.ejrad.2024.111624}},
  doi          = {{10.1016/j.ejrad.2024.111624}},
  volume       = {{178}},
  year         = {{2024}},
}