TNBC-PDKi67-Uganda v1.0: A Public Digital Pathology Dataset of PD-L1 and Ki-67 Expression in Triple-Negative Breast Cancer from Uganda

Abstract

TNBC-PDKi67-Uganda v1.0 is a publicly available digital pathology dataset comprising representative immunohistochemistry (IHC) images of Programmed Death-Ligand 1 (PD-L1) and Ki-67 expression in Triple-Negative Breast Cancer (TNBC) from Uganda. The dataset was curated from archived formalin-fixed paraffin-embedded (FFPE) specimens obtained at Mbarara Regional Referral Hospital (MRRH) and utilized in the Master of Medicine dissertation, “PD-L1 Expression and Its Association with Ki-67 in Triple-Negative Breast Cancer at Mbarara Regional Referral Hospital.” Version 1.0 contains five representative JPEG photomicrographs demonstrating a spectrum of biomarker expression patterns, including PD-L1 combined positive scores (CPS) of 0%, 15%, and 33%, and Ki-67 expression levels of 5% and 95%. Each image is accompanied by structured metadata, including anonymized case identifiers, biomarker labels, expression percentages, scoring methodology, and specimen information. The dataset is intended to support educational activities, biomarker interpretation, digital pathology benchmarking, and exploratory computational pathology research. It provides a valuable open-access resource from an underrepresented African population and contributes to ongoing efforts to improve the availability of publicly accessible cancer pathology datasets from low- and middle-income countries. Dataset Characteristics Dataset Version: v1.0 Total Images: 5 PD-L1 Images: 3 Ki-67 Images: 2 Disease: Triple-Negative Breast Cancer (TNBC) Specimen Type: Formalin-Fixed Paraffin-Embedded (FFPE) Institution: Mbarara Regional Referral Hospital (MRRH) / Mbarara University of Science and Technology (MUST) Country of Origin: Uganda Image Format: JPEG License: Creative Commons Attribution 4.0 International (CC BY 4.0) This dataset promotes open science, supports the dissemination of African digital pathology resources, and serves as a foundation for future expansions incorporating additional TNBC cases, whole-slide imaging, expert annotations, and artificial intelligence applications in precision oncology.