Towards Interpretable Classification of Leukocytes based on Deep Learning
This work addresses interpretability for healthcare professionals in blood cell analysis, but it is incremental as it builds on existing deep learning methods.
The paper tackled the problem of integrating machine learning into clinical workflows by improving confidence calibration and comparing visual explanation methods for leukocyte classification, achieving high accuracy where human observers struggle.
Label-free approaches are attractive in cytological imaging due to their flexibility and cost efficiency. They are supported by machine learning methods, which, despite the lack of labeling and the associated lower contrast, can classify cells with high accuracy where the human observer has little chance to discriminate cells. In order to better integrate these workflows into the clinical decision making process, this work investigates the calibration of confidence estimation for the automated classification of leukocytes. In addition, different visual explanation approaches are compared, which should bring machine decision making closer to professional healthcare applications. Furthermore, we were able to identify general detection patterns in neural networks and demonstrate the utility of the presented approaches in different scenarios of blood cell analysis.