PathAI logo
Poster

AI-based quantitation of cancer cell and fibroblast nuclear morphology reflects transcriptomic heterogeneity and correlates with survival in breast cancer

SABCS 2022

Study Background


  • Morphological features of cancer cell nuclei are routinely used to assess disease severity and prognosis, and prior work has linked cancer nuclear morphology to genomic alterations1-3. In addition, the cellular composition of cancer-associated stroma (CAS) has been linked to prognosis in several cancer types, including breast cancer4.

  • Quantitative analyses of 1) nuclear features of cancer cells and other tumor-resident cell types, such as cancer-associated fibroblasts (CAFs), and 2) composition of CAS may reveal novel biomarkers for prognosis and treatment response.

  • Here, we applied a nucleus detection and segmentation algorithm, a cell classification model, and a stromal subdivision model to hematoxylin and eosin (H&E)-stained whole slide images (WSIs) of breast cancer specimens, enabling the assessment of features related to nuclear morphology and stromal composition.


References:
1. Zink, D., et al. Nat Rev Cancer. 2004; 4:677–687.
2. Fischer, E.G. Acta Cytol. 2020; 64:511–519.
3. Chow, K.-H., et al. Nat Rev Cancer. 2012; 12:196–209.
4.Wu, S.Z., et al. The EMBO J. 2020; 39:e104063.
View Poster

Authors

Abel et al.