Deep Learning-Based Label-Free Hematology Analysis Framework Using Optical Diffraction Tomography

Hematology analysis, a common clinical test for screening various diseases, has conventionally required chemical staining process that is time-consuming and labor-intensive. To reduce the costs of staining, label-free imaging can be utilized in hematology analysis. In this work, we exploit optical diffraction tomography fully convolutional one-stage object detector or FCOS, deep learning architecture detection, to develop analysis framework. Detected cells are classified into four groups: red blood cell, abnormal platelet, white cell. results, trained detection model showed superior performance refractive index tomograms (0.977 mAP) also high accuracy four-class classification (0.9708 weighted F1 score, 0.9712 total accuracy). For further verification, mean corpuscular volume (MCV) hemoglobin (MCH) were compared with values obtained from reference equipment, our results showing reasonable correlation both MCV (0.905) MCH (0.889). This study provides successful demonstration proposed framework detecting classifying using