![]() | Only 14 pages are availabe for public view |
Abstract Breast cancer is the most common malignancy in women worldwide. Information on the incidence and mortality of breast cancer is essential for planning health measures. Breast cancer remains a leading cause of cancer deaths among women in many parts of the world. Early detection of breast cancer through periodic screening has noticeably improved the outcome of the disease. Detection and diagnoses of this disease as early as possible and getting state-of-theart cancer treatment are the most important strategies to prevent deaths from breast cancer. Breast cancer that’s found early, when it’s small and has not spread, is easier to treat successfully. Getting regular screening tests is the most reliable way to find breast cancer early. The American Cancer Society has screening guidelines for women at average risk of breast cancer, and for those at high risk for breast cancer. Regular physical breast exams done by either a health professional (clinical breast exams) or by the female itself (breast self-exams), may be useful in this respect. Nowadays, numerous tests are available to look for and diagnose breast cancer. If the doctor finds an area of concern on a test (a mammogram), or if the female has symptoms that could mean breast cancer, the patient will need more tests to know for sure if it’s cancer. Newer types of tests are being developed for breast imaging, including Optical imaging tests, Molecular Breast Imaging, Positron Emission Mammography, Electrical impedance imaging, and Elastography. The most familiar tests available todays in the majority of clinics are: Mammograms, Breast Ultrasound, Breast Magnetic Resonance Image (MRI) Scans. Each of these modalities has its advantages and disadvantages. The aim of this study is to compare between these three modalities by employing two analyzing software programs for manipulating the images obtained by each modality as a trial to obtain solid conclusion of using either of these modalities for the better differentiation between benign and malignant breast tumors. A total of 60 female breast patients were enrolled in this study, in addition to 10 control female with no symptoms of breast disease. The breast patients were divided into two main groups; each group included 20 female breast patient. One group was of benign breast tumors and the other group of malignant breast changes, based on large core needle biopsy. All groups, including the control group, was submitted to mammographic, ultrasound and magnetic resonance breast examination. In the mammography, two routine views; a top-to-bottom view (CC view) and an oblique side view (MLO view),were recorded. In ultrasonography, and magnetic resonance imaging procedures, were followed. Photo Shop 7.0 ME Software, and MATLAB Software were employed, based on the graylevel histogram, to describe the obtained data. Some validation measures were used to evaluate the validity of each modality , including; sensitivity, specificity, and accuracy, in addition to some other statistical features, extracted from the regions of interest including; Summary 112 mean, standard deviation, coefficient of variation, median, mode value, minimum value, maximum value, kurtosis, entropy, Contrast, Correlation, Homogeneity, IDM, and RMS. To interpret the results, two routes were followed, namely; the qualitative and the quantitative comparisons. The qualitative comparisons revealed that in normal breast tissues, the histogram appears as a sharp narrow peak in any of the used modalities. However, the histogram peak in each modality occupies different location, i.e., in mammography, (whether it is of Cranio-Cuadal or MedioLateral oblique view) , the histogram peaks of normal breast tissues are localized at the high end of the gray-scale level (to the right), while both ultrasonography and MRI images the normal peaks are localized at the lower end of the gray-scale level (to the left). The distribution does not skewed to left or right, in case of mammography,i.e., it is leptokurtic (often its central peak is higher and sharper) while , it is shifted to the left, in case of both ultrasonography, and MRI, i.e., it is (-) skewed. Skewness is a measure for the degree of symmetry in the distribution curve in which normal breast tissue can be distinguished from other breast tissues. In case of benign breast tissues, the mammograms (whether it is of Cranio-Cuadal or MedioLateral oblique view), and the gray level histogram appears as a broad peak at the middle of the gray-scale, with broad base and sharp top with more range and more area is occupied by the benign peak tissue due to the wide variations in the benign breast tissues, especially because the types of benign breast tumors are numerous ranging from areas of thickening to fibro adenomas passing by Lobular neoplasia, and fibrocystic changes. In case of ultrasonography, benign breast tissue appears as a peak with broad base and relatively sharp top with peaks and vales, with the histogram main peak extends from the middle and shifted to the left, i.e. the distribution curve is skewed to left which is called (-) skewed. In case of MRI, the benign histogram gray level distribution is characterized by two main peaks at the beginning and at the end of the distribution with the distribution is located in the middle of the gray-scale with peaks and valley. Malignant breast tissue in mammography, whether it is of Cranio-Cuadal or MedioLateral oblique views, appears as a broad peak of high gray-level at the middle of the histogram, with broad base and broad top, occupies more area in the middle of the histogram gray-level. The distribution curve is skewed to left which is called (-) skewed. In case of ultrasonography, the histogram of malignant breast tissue appears as a peak with sharp top extends from the middle of the gray-level scale shifted to the left, i.e. the distribution curve is skewed to left which is called (-) skewed. In case of MRI, the histogram appears as a peak of gradual increase in intensity forming a sharp peak of high gray-level at the lower end of the histogram but declined sharply towards the high end of the histogram forming an extended tail towards the lower end of the histogram. The distribution curve is skewed to left which is called (-) skewed. In conclusion: Each modality gives specific shape for each imaged ROI of each tissue category. So, the qualitative shape and location of the histogram gray-level distribution can be used to differentiate, roughly, between the breast tissue category . In the quantitative comparisons, and due to the numerous statistical features extracted from the gray-level histograms, and as a trial to obtain a solid conclusion Summary 113 concerning the most suitable modality to be able to choose the modality that gives the most accurate breast lesion diagnosis, a score table was established, where it is scored to point 0 if no significant difference exists between any two breast tissue categories, it is scored to point 1 if a significant difference between two breast tissue categories exists, e.g., normal/benign. Also, it is scored to point 2 if two significant differences exist, e.g., normal/benign, benign/malignant, and it is scored to point 3 if three significant differences exist, e.g., normal/benign, normal/malignant, and benign/malignant. This, is of course, was considered with each software programe; photoshop 7.0 ME and MATLAB software. According to this table, MRI has maximum scores, followed by ultrasonography, and ended by mammography in case differentiation between the three categories of breast tissues. In addition, MRI, proved according to this table , that it is the best modality that can differentiates between benign and malignant breast tissues. However, in considering the sensitivity and specificity, it is clear that, sensitivity increases on the expense of specificity, and vice versa. The data of this study, revealed that, both mammography and MRI has high sensitivity, with ultrasonography is lower. This is because the acoustic characteristics of benign and malignant lesions are overlapping. As a final conclusion, no single imaging method has perfect sensitivity and perfect specificity for the diagnosis of breast cancer. Accordingly, the choice depends on many factors that take into considerations the limitations, advantages and disadvantages of each modality that is coinside with the patient abilities. |