The role of ultrasound examination in breast cancer

Introduction

Breast ultrasound is frequently used to evaluate breast problems found during a mammography, especially in women with dense breast tissue or with palpable lesions on physical exam. Ultrasound helps to distinguish normal findings like cysts or fat lobules from suspicious breast changes which need to be biopsied. Moreover, ultrasound can be used to evaluate the enlargement of lymphatic nodes. The test is painless and does not expose the patient to radiation. The use of ultrasound for breast cancer screening is limited by the skill of the examiner, resolution of the device and uncertainty, whether the whole breast was scanned. Mammography provides almost complete breast pictures at two projections with high sensitivity for microcalcifications. The aim of our study is to compare the ultrasound and mammography findings, as the basis for improving the ultrasound results using 3D technique. Supported by Grant Technological agency Czech republic TA04011392

Materials/methods

67 breast cancer patients and 71 healthy women took part in this study. The median age was 61 years, minimum 30 years and maximum 90 years. 55% of patients were between 50 and 70 years. The probands were routinely clinically examined and underwent mammography every two years and hand held ultrasound in 6 months periods. During a breast exam, we record ultrasound images and the position and orientation of the ultrasound probe. For this purpose we use an off-the-shelf grabbing PC card in combination with the tracking device. The hardware set-up used to acquire all the necessary data during ultrasound examination consist of a computer with a digital video grabber and an electro-magnetic six degrees of freedom tracking system (TrakSTAR, Ascension Technology Corp., Shelburne, VT, USA). We are developing the 3D breast ultrasound technique to ensure the whole breast exam and better depicting of lesions using electromagnetic tracking technology and Bayesian inference algorithm for reconstructing 3D volumes.

Results and conclusions

From the tumor cohort 37% of patients were diagnosed at screening mammography and 63% of them came with a known nodule, found by self-examination. Mammography revealed 93% true positivity, 7% false negativity. The hand held ultrasound revealed 90% true positivity but 10% false negativity. In the group of 70 healthy women the mammography revealed 1% false positivity and 99% true negativity, ultrasound had 3% false positivity and 97% true negativity. The sensitivity of mammography was 93 %. The sensitivity of the ultrasound was 90 %. The clinical evaluation had 31% false negative rate, 69% sensitivity. During a breast exam, we record ultrasound images and the position and orientation of the ultrasound probe. We can thus monitor the quality of the exam and regions that were examined and minimize the risk of not detecting harmful lesions. The proposed method thus ultimately reconstructs the 3D volume from 2D images while removing noise, blur, increasing resolution and surpassing the quality of the original 2D images. This technique could lead to lower false negative rate of breast hand held ultrasound.