Automated Breast Volume Scanning 3D Ultrasound of the Breast

More documents

Recommendations

Info

shown in 3 directions. Bottom right: lesionkin and nipple shown underneath.ht: sagittal.Figure 5. Post-lumpectomy exam of a specimen for verification. The coronal aspect (left)clearly shows the hypoechogenic tumor, bounded by spiculae. Above right: transverse;bottom right: sagittal.evaluation of the results of a 3D MRIbreast examination. As the A-P diameterof the breast is much smaller than the L-Ror cranio caudal diameter, the entire setof data can be worked through muchmore quickly with the coronal view.First experiences with the 3D ultrasoundsystem show that evaluation of thegenerated images takes 2-4 minutesper breast, much faster than hand heldultrasound.Is ultrasound a usefuladdition to mammographyscreening?Interest in a combined mammography/ultrasound screening protocol was raisedby the results of the American College ofRadiology Imaging Network (ACRIN) 6666study (Berg et al., JAMA 2008), whichshowed that combined use of the twomodalities allows more tumors to betraced in women with a high risk ofbreast cancer and a dense glandularbreast tissue structure. The study notonly showed that over half of the womenaged under 50 have more than 50%glandular tissue (in women over 50 thisnumber is 30%), but also that the sensitivityof mammography with thesewomen levels out at 30-48%. Berg et al.(JAMA 2008) showed that the addition ofultrasound to mammography increasedtumor detection sensitivity to 77.5%.Their proposal is to introduce ultrasoundas a supplementary screening modality.A drawback of the revised examinationprotocol is the extra time required.The ACRIN report states that the averageexamination time for a bilateral ultrasoundbreast examination is 19 minutes,which does not include comparison withearlier examinations and reporting. At arough estimate, 35% of the population ofpatients with more than 50% glandulartissue will need additional ultrasoundexaminations. This will mean a significantincrease in workload and hence a lack ofmanpower. Automation of procedures,however, means that the actual examinationcan be performed by radiologicaltechnicians, and data acquisition andreporting can be performed separately,both in place and time.Partly as a result of the outcome of theACRIN 6666 study in October 2009, theAmerican State of Connecticut has passeda bill that grants women the right to beinformed on the amount of glandulartissue in their breast and its implicationsfor their screening. It is expected that thiswill strongly increase interest in breastultrasound, with the rest of the worldsoon to follow this lead.
Roel Mus, radiologist UMC St RadboudMatthieu Rutten, radiologist JBZScientific researchThe St. Radboud University NijmegenMedical Centre and the Jeroen BoschHospital are about to begin furtherclinical research using the ACUSONS2000 ABVS.Although the first results are promising,much research still needs to be doneregarding the value of 3D ultrasound.Multiple studies have already been set upin Germany, the United States of America,Japan and the Netherlands (RUNMC andJBZ). These studies specifically addressthe sensitivity, specificity and positivepredictive value of 3D ultrasoundcompared with 2D ultrasound and MRI.Improved specificity could help reducethe number of biopsies performed.However, the added value of 3D ultrasoundseems to be lesion detectionrather than lesion characterization.Automated volume ultrasound willprobably make it easier to find smallertumors (3-4 mm). The potential to reviewthe 3D digital data sets of ultrasoundimages using a computer-aided detection(CAD) system should be expanded in thefuture.Women with an average breast cancerrisk, i.e. a 10 to 20% lifetime risk ofdeveloping the disease, could probablygain the most from automated ultrasoundscreening for breast cancer.Presently in Europe and the UnitedStates, MRI is recommended as anadditional screening modality for womenwith a high risk (20% or up) of breastcancer. Women with an intermediatebreast cancer risk do not qualify foranything other than standard mammography,but they may also have adense glandular breast tissue structurethat negatively impacts the precision ofmammographic examination. Womenwith more than 75% glandular breasttissue have a 4 to 5 times higher risk ofbreast cancer than women with little tono glandular tissue in their breasts. Thisresults in a higher percentage of intervalcarcinomas and a poorer prognosis forany clinically diagnosed tumors.Especially for young women with aBRCA1 or BRCA2 gene mutation anddense breasts, additional examinationsthrough ultrasound could well be muchmore reliable than conventional mammography,with the added advantagethat no radiation is used. Presentlythis category of patients is screenedaccording to a protocol involving a yearlymammogram and an MRI of the breasts.Using the ACUSON S2000 ABVS, theradiology departments of the RadboudUniversity Nijmegen Medical Centre(Roel Mus, Henkjan Huisman, NicoKarssemeijer) and the Jeroen BoschHospital in ‘s-Hertogenbosch (MatthieuRutten, Mathijn de Jong, Ivo Dubelaar,Thomas Fassaert) will this fall start aclinical study among women who carrya BRCA gene mutation. In this studythe current protocol (combined yearlymammography + MRI) will be comparedwith an alternative protocol (biannualABVS + yearly mammography + MRI).Also the results of automated breastvolume ultrasound and mammographywill be compared.We hypothesize that the use ofautomated breast volume scanning willdetect more tumors than mammography,and that the incidence of intervalcarcinomas will decrease as the ABVSexamination will take place every sixmonths.In regular screening, the distinctionbetween a cyst and a hypoechogenicfibroadenoma is of less importance (bothbeing BI-RADS 2, i.e. benign). However,in examining BRCA gene carriers thisdistinction is very important as thegrowth rate of tumors in this group ismuch higher than in “regular” patients.The fast-growing tumor pushes thesurrounding tissue away, leaving littletime for spicula-like ingrowth. Hence,BRCA tumors are often characterizedby an echographically clear, “benign”delineability.The entire study will take approximatelytwo years. By that time we expect to havegathered sufficient data to assess the roleof automated breast volume scanning indetecting breast cancer.
Page 6: Local Contact InformationSiemens Me

Automated Breast Volume Scanning 3D Ultrasound of the Breast

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?