Bone marrow aspirate (BMA) differential cell counts (DCCs) are critical for classification of hematologic disorders. Manual DCCs are still considered the gold standard as a reliable automated method is yet to be developed. Digital pathology and machine learning represent a highly promising technology for this purpose. The authors report their experience developing machine learning algorithms to detect and classify BMA cells. Promising early results signify an important initial step in the effort to devise a reliable, objective method to automate DCCs.
- Ramraj Chandradevan
- Ahmed A. Aljudi
- David L. Jaye
