Introduction: The Cell Population Data (CPD) generated from a modern blood analyzer can produce various blood cell parameters which can support the screening and diagnosis of many cellular abnormalities, including hematologic malignancies. CPD parameters provide the morphological and functional characteristics of the leukocytes which can be beneficial to help hematology experts in determining differential diagnosis. Recent advances Machine Learning (ML) technology offers the opportunity to deal with complex medical diagnostics fields using hematological parameters. Objective: This research presents a novel approach of using ML algorithms for screening hematologic malignancies using CPD parameters. Method: CPD parameters were collected from 882 subjects (457 with hematologic malignancy and 425 with hematologic non-malignancy) at Konkuk University Medical Center, Seoul. Seven machine learning models, (i.e., SGD, SVM, RF, DT, Linear model, Logistic regression, and ANN) were performed and model performance were evaluated using stratified 10-fold cross validation. The classification accuracy, precision, recall, and AUC were then compared. Result: We obtained outstanding performance by the ANN model as compared to other ML models. The diagnostic ability of ANN achieved the highest accuracy, precision, recall, and AUC ± Standard Deviation as follows: 82.8%, 82.8%, 84.9%, and 93.5%±2.6, respectively. Conclusion: ANN algorithm based on CPD appeared to be an efficient aid for clinical laboratory screening of hematologic malignancies. Our results encourage further work of applying ML to wider field of clinical practice.
Introduction: The Cell Population Data (CPD) generated from a modern blood analyzer can produce various blood cell parameters which can support the screening and diagnosis of many cellular abnormalities, including hematologic malignancies. CPD parameters provide the morphological and functional characteristics of the leukocytes which can be beneficial to help hematology experts in determining differential diagnosis. Recent advances Machine Learning (ML) technology offers the opportunity to deal with complex medical diagnostics fields using hematological parameters. Objective: This research presents a novel approach of using ML algorithms for screening hematologic malignancies using CPD parameters. Method: CPD parameters were collected from 882 subjects (457 with hematologic malignancy and 425 with hematologic non-malignancy) at Konkuk University Medical Center, Seoul. Seven machine learning models, (i.e., SGD, SVM, RF, DT, Linear model, Logistic regression, and ANN) were performed and model performance were evaluated using stratified 10-fold cross validation. The classification accuracy, precision, recall, and AUC were then compared. Result: We obtained outstanding performance by the ANN model as compared to other ML models. The diagnostic ability of ANN achieved the highest accuracy, precision, recall, and AUC ± Standard Deviation as follows: 82.8%, 82.8%, 84.9%, and 93.5%±2.6, respectively. Conclusion: ANN algorithm based on CPD appeared to be an efficient aid for clinical laboratory screening of hematologic malignancies. Our results encourage further work of applying ML to wider field of clinical practice.