| Summary: | 碩士 === 國立清華大學 === 電機工程學系 === 102 === In medicine, cancer has been the leading cause of death in Taiwan for decades; meanwhile breast cancer is among the top 10 cause of cancer death. Therefore, this research tries to explore domestic breast cancer survival analysis by SPSS using the valuable clinical data provided by Doctor Yi-Hong Chou, the Taipei Veterans General Hospital. Life table, Kaplan-Meier analysis, and Cox-regression on the data reveals that axillary metastasis, lymph node metastasis, tumor size and progesterone receptor (PR) are the important factors leading to breast cancer. The majority of patients died in the first four years, and the mortality rate is significantly reduced if survived more than five years.
The research of phalanx has been the subject that is discussed widely as well. With X-ray image of phalanx, the condition of the growth, and the estimate of height can be interpreted. And whether the age of phalanx is in accordance with real age is the basis for doctor to diagnose whether the patient might have the endocrine disorders or other disease. Tanner and Whitehouse (TW) is a method to estimate the age of phalanx. 20 parts of RUS and Carpal bones in the X-ray image of left hand are scored against corresponding bone maturity stages templates, respectively. RUS system score and Carpal system score are then formed by summing up individual bone part maturity scores and the bone age is prognosed by looking up the table of maturity score vs. bone age. This research makes reference to TW method and try to do bone age assessment by using SPSS multiple regression analysis with increased parameters including the length, width, and area of phalanges and thickness of epiphysis. PCA is applied first before multiple regression analysis. We found that bone age and the phalanx’s length have the positive correlation in growth, and thickness of epiphysis has the negative correlation with bone age. In assessing bone age using the multiple regression analysis, the results in girls are always better than boys in terms of assessment accuracy. Perhaps it is because the growth development of girls is earlier than boys, and the boy’s development spans over relatively long period of time.
|
|---|
