| 摘要 |
<p>The treatment planning, surgical planning and dose estimation of the abnormal tissues and lesion is barred by the lack of a well-established tumor volume calculation technique noninvasive^ from the patient's dataset. The technique that is being commonly used for estimation of volume of abnormalities is the ABC/2 method. This method however is not exact and just gives the bounding box volume for the tumor. It is simple but its efficacy is limited by intraobserver variability. To address issues like accuracy, level of trained intervention, time for computation and complexity for the calculation of volume of tumors and others abnormal tissues or lesions, we have developed a method which is a semi-automatic (single user initialization), non-invasive, accurate and fast with least intervention from a well-trained radiologist. We traced exact boundaries of tumors slice by slice automatically by utilizing the fused low and high frequency components embedded in an image which consequently was used in surveyor's algorithm for the calculation of volume of tumors, cysts, clots and other abnormal tissues and lesions.</p> |
