Pranab Dey1, Amit Lamba, Savita Kumari, Neelam Marwaha. 1. Department of Cytology, Postgraduate Institute of Medical Education and Research, Chandigarh 160012, India. deypranab@hotmail.com
Abstract
OBJECTIVE: To use a commercially available artificial neural network (ANN) software program to distinguish prognostically good and bad cases of chronic myeloid leukemia (CML). STUDY DESIGN: A total of 40 patients with CML who developed blast crisis or proceeded in the accelerated phase were selected. They formed two groups, group 1 and group 2, of 20 patients each, who developed accelerated phase or blast crisis within 18 months or 30 months, respectively, after the initial diagnosis of the chronic phase of CML. The detailed clinical, hematologic, and morphometric data were collected in all these cases. A suitable ANN software program was used to analyze these data. RESULTS: All cases were randomly distributed automatically by the program into three groups: training set (28), validation set (4), and test set (8). In the test set, the ANN program successfully classified all group I and group II patients. CONCLUSION: We successfully used a commercially available ANN software program to develop a model able to classify prognostically good and bad cases of CML.
OBJECTIVE: To use a commercially available artificial neural network (ANN) software program to distinguish prognostically good and bad cases of chronic myeloid leukemia (CML). STUDY DESIGN: A total of 40 patients with CML who developed blast crisis or proceeded in the accelerated phase were selected. They formed two groups, group 1 and group 2, of 20 patients each, who developed accelerated phase or blast crisis within 18 months or 30 months, respectively, after the initial diagnosis of the chronic phase of CML. The detailed clinical, hematologic, and morphometric data were collected in all these cases. A suitable ANN software program was used to analyze these data. RESULTS: All cases were randomly distributed automatically by the program into three groups: training set (28), validation set (4), and test set (8). In the test set, the ANN program successfully classified all group I and group II patients. CONCLUSION: We successfully used a commercially available ANN software program to develop a model able to classify prognostically good and bad cases of CML.
Authors: Antonio Rivero-Juárez; David Guijo-Rubio; Francisco Tellez; Rosario Palacios; Dolores Merino; Juan Macías; Juan Carlos Fernández; Pedro Antonio Gutiérrez; Antonio Rivero; César Hervás-Martínez Journal: PLoS One Date: 2020-01-10 Impact factor: 3.240
Authors: Jens P E Schouten; Christian Matek; Luuk F P Jacobs; Michèle C Buck; Dragan Bošnački; Carsten Marr Journal: Sci Rep Date: 2021-04-12 Impact factor: 4.379