Data Matters: The Impact of Data Curation in the Classification of Histopathological Datasets
Preprint
Abstract
In this work, the impact of outliers on the performance of machine learning and deep learning models is investigated, specifically for the case of histopathological images of colorectal cancer stained with Haematoxylin and Eosin. The evaluation of the impact is done through the systematic comparison of one machine learning model (Random Forests) and one deep learning model (ResNet-18). Both models were trained with the popular NCT-CRC-HE-VAL-100K dataset and tested on the CRC-HE-VAL-7K companion set. Then, a curation process was performed by analysing the divergence of patches based on chromatic, textural and topological features of the training set and removing outliers to repeat the training with a cleaned dataset. The results showed that machine learning models, can benefit more from improvements in the quality of data, than deep learning models. Further, the results suggest that deep learning models are more robust to outliers as, through the training process, the architectures can learn features other than those previously mentioned.Keywords
This publication has 19 references indexed in Scilit:
- STHELAR, a multi-tissue dataset linking spatial transcriptomics and histology for cell type annotationScientific Data, 2026
- Persistent Homology and Gabor Features Reveal Inconsistencies Between Widely Used Colorectal Cancer Training and Testing Datasets2025
- NCT-CRC-HE: Not All Histopathological Datasets are Equally Useful2025
- A large histological images dataset of gastric cancer with tumour microenvironment annotation for AIScientific Data, 2025
- Publicly available datasets of breast histopathology H&E whole-slide images: A scoping reviewJournal of Pathology Informatics, 2024
- BRACS: A Dataset for BReAst Carcinoma Subtyping in H&E Histology ImagesDatabase: The Journal of Biological Databases and Curation, 2022
- Pan-tumor CAnine cuTaneous Cancer Histology (CATCH) datasetScientific Data, 2022
- Gland segmentation in colon histology images: The glas challenge contestMedical Image Analysis, 2017
- A random forest guided tourTEST, 2016
- Random ForestsMachine Learning, 2001