<![CDATA[This work investigates the use of sophisticated image analysis methods to differentiate between benign and cancerous blood cells directed on different phases of pro-B and pre-B lymphoblast growth. Binary image processing, segmentation, and masking techniques were used for 500 blood cell pictures. One hundred eighty (180) were determined to be benign and 320 to be malignant, with considerable morphological differences such as alterations in cytoplasmic ratios and aberrant nuclear structure. With 95% accuracy, these characteristics are made between benign and malignant cells to distinguish. Significant morphological variations, including anomalies in the atom form and changes in the cytoplasmic ratios, were detected, and they could extricate between malignant and benign cells with 95% accuracy. More features could be extracted from the images based on segmentation, especially when identifying cancerous cells early in their development. These results imply that automated techniques can be invaluable in helping pathologists identify hematopoietic malignancies such as acute lymphoblastic leukemia (ALL) at an early stage. Better therapy results could result from increased diagnostic speed and accuracy brought about by this automation. Further study is necessary to improve the generalizability of these systems across datasets.]]>
Copyrights © 2025
