Advancing Ki67 hotspot detection in breast cancer: a comparative analysis of automated digital image analysis algorithms
2025

Improving Ki67 Detection in Breast Cancer with Automated Analysis

Sample size: 135 publication 10 minutes Evidence: high

Author Information

Author(s): Mieke C Zwager, Shibo Yu, Henk J Buikema, Geertruida H de Bock, Thomas W Ramsing, Jeppe Thagaard, Timco Koopman, Bert van der Vegt

Primary Institution: University Medical Center Groningen

Hypothesis

Can automated digital image analysis algorithms improve the detection and scoring of Ki67 hotspots in breast cancer compared to manual assessment?

Conclusion

Automated Ki67 hotspot detection and analysis correlated strongly with manual assessment and provided higher proliferation indices compared to manual scoring.

Supporting Evidence

  • Automated detection was successful in 73% of cases for VDS and 100% for DL.
  • Automated methods provided higher Ki67 hotspot proliferation indices than manual methods.
  • Correlation between manual and automated methods was substantial (r=0.90 for VDS and r=0.95 for DL).

Takeaway

Finding the most active parts of breast cancer tumors is hard for doctors, but using computers can help them do it better and faster.

Methodology

The study involved analyzing tissue sections from 135 breast cancer cases using two automated algorithms and comparing their results with manual assessments.

Limitations

The study was retrospective, had a heterogeneous sample, and lacked follow-up data to assess prognostic power.

Participant Demographics

{"gender":{"female":132,"male":3},"age":{"under_60":63,"60_and_over":72,"mean_age":60.4},"histological_type":{"invasive_carcinoma_NST":114,"invasive_lobular_carcinoma":18,"mixed_invasive_carcinoma":3}}

Digital Object Identifier (DOI)

10.1111/his.15294

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication