Discriminating healthy from tumor tissue in breast lumpectomy specimens using deep learning-based hyperspectral imaging

Lynn-Jade S Jong; Naomi de Kruif; Freija Geldof; Dinusha Veluponnar; Joyce Sanders; Marie-Jeanne T F D Vrancken Peeters; Frederieke van Duijnhoven; Henricus J C M Sterenborg; Behdad Dashtbozorg; Theo J M Ruers

doi:10.1364/BOE.455208

Discriminating healthy from tumor tissue in breast lumpectomy specimens using deep learning-based hyperspectral imaging

Biomed Opt Express. 2022 Apr 4;13(5):2581-2604. doi: 10.1364/BOE.455208. eCollection 2022 May 1.

Authors

Lynn-Jade S Jong^{1

2

3}, Naomi de Kruif^{4

3}, Freija Geldof^{1

2}, Dinusha Veluponnar^{1

2}, Joyce Sanders⁵, Marie-Jeanne T F D Vrancken Peeters¹, Frederieke van Duijnhoven¹, Henricus J C M Sterenborg^{1

6}, Behdad Dashtbozorg^{1

4}, Theo J M Ruers^{1

2}

Affiliations

¹ Department of Surgery, Netherlands Cancer Institute, Plesmanlaan 121, 1066 CX Amsterdam, The Netherlands.
² Faculty of Science and Technology, University of Twente, Drienerlolaan 5, 7522 NB Enschede, The Netherlands.
³ Equal contributors.
⁴ Department of Biomedical Engineering, Eindhoven University of Technology, PO Box 513, 5600 MB Eindhoven, The Netherlands.
⁵ Department of Pathology, Netherlands Cancer Institute, Plesmanlaan 121, 1066 CX Amsterdam, The Netherlands.
⁶ Department of Biomedical Engineering and Physics, Amsterdam University Medical Center, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands.

Abstract

Achieving an adequate resection margin during breast-conserving surgery remains challenging due to the lack of intraoperative feedback. Here, we evaluated the use of hyperspectral imaging to discriminate healthy tissue from tumor tissue in lumpectomy specimens. We first used a dataset obtained on tissue slices to develop and evaluate three convolutional neural networks. Second, we fine-tuned the networks with lumpectomy data to predict the tissue percentages of the lumpectomy resection surface. A MCC of 0.92 was achieved on the tissue slices and an RMSE of 9% on the lumpectomy resection surface. This shows the potential of hyperspectral imaging to classify the resection margins of lumpectomy specimens.