Unsupervised Cross-Domain Feature Extraction for Single Blood Cell Image Classification
This addresses the challenge of reusing machine learning models across different medical labs without costly expert labeling, though it is incremental in applying unsupervised feature extraction to a specific domain.
The paper tackled the problem of domain shifts in white blood cell image classification by proposing an unsupervised cross-domain adapted autoencoder, which enabled a simple random forest classifier to outperform existing methods on unseen datasets.
Diagnosing hematological malignancies requires identification and classification of white blood cells in peripheral blood smears. Domain shifts caused by different lab procedures, staining, illumination, and microscope settings hamper the re-usability of recently developed machine learning methods on data collected from different sites. Here, we propose a cross-domain adapted autoencoder to extract features in an unsupervised manner on three different datasets of single white blood cells scanned from peripheral blood smears. The autoencoder is based on an R-CNN architecture allowing it to focus on the relevant white blood cell and eliminate artifacts in the image. To evaluate the quality of the extracted features we use a simple random forest to classify single cells. We show that thanks to the rich features extracted by the autoencoder trained on only one of the datasets, the random forest classifier performs satisfactorily on the unseen datasets, and outperforms published oracle networks in the cross-domain task. Our results suggest the possibility of employing this unsupervised approach in more complicated diagnosis and prognosis tasks without the need to add expensive expert labels to unseen data.