Summary: | 碩士 === 國立中興大學 === 資訊管理學系所 === 100 === The modern diagnosis of anemia symptom is typically determined by the observation of red blood cells under microscopes. Stained blood smears of blood samples taken from patients are diagnosed by the area percentage of biconcave disk in a red blood cell by professionally trained doctors. The measurement of erythrocyte volume hematocrit is an important indicator of the severity and type of anemia. Manually microscopic examination of the number and size of red blood cells is a time-consuming and error-prone process. It also requires experienced hematologists to accurately interpret the results. Therefore, development of an automatic red blood cell segmentation method is essential to assist the diagnosis of anemia symptom.
The images of blood smear were used in this study. The intact red blood cells were first defined through image analyses of red blood cells, overlapped red blood cells, biconcave disks, and character calculations . Furthermore, the number and size of red blood cells and the percentage of biconcave disk within one smear slide can be calculated. All these parameters provide quantitative measures to ameliorate the accuracy of anemia diagnosis.
|