Dokuz eylül University Faculty of Medicine Radiology Department's full field digital mammography experience/ predictive value of digital mammography

SUMMARY:Purpose: Our purpose was to evaluate the diagnostic value of full-field digital mammography(FFDM) based upon the positive predictive value (PPV) for breast cancer depending on theexperience of our department of Radiology in Dokuz Eylül University Faculty of Medicine inscreen-film mammography (SFM) between the years of 2000-2003 and FFDM between the yearsof 2004-2005 .Materials and Methods: A total of 20.324 women evaluated by SFM and 6.024 womenevaluated by FFDM between the age of 32-93 years (mean, 48.9 years) are grouped depending onthe BI-RADS assesment cathegories. The patients categorized as BI-RADS 4 or BI-RADS 5were evaluated. PPVs are calculated for both groups . Breast cancer status was ascertained on thebases breast biopsy. The subgroups (microcalcifications, mass or asymetric density) of the lesionswere also evaluated seperately . For both methods we compared the time spent during the core-needle biopsy. We used the Ki square test and Fischer exact test in the SPSS 11.0 programme forstatistical analysis.Results:Out of all the cases exposed to SFM evaluation, 128 (%0.55) of them were diagnosed asBI-RADS 4 and 104 (0.44%) as BI-RADS 5. On the other hand among all the cases, subject toFFDM (full-field digital mammography) evaluation, 76 cases (1.26%) were diagnosed as BI-RADS 4, and 56 (0.93 %) were diagnosed as BI-RADS 5. The PPVs have been calculated as57.0 % for the BI-RADS 4 lesions, 99% for BI-RADS 5 lesions, which were both evaluatedthrough conventional mammography. On the other hand, PPVs for the BI-RADS 4 was read as57.8% and as %98.2 for BI-RADS 5 lesions, which were subject to FFDM evaluation. Ninety-three lesions recomended for biopsy as a result of SFM readings and 30 lesions recomended forbiopsy as a result of FFDM readings were all defined as non-palpable lesions, on which thebiopsy was carried out by means of ?core-needle biopsy? method. Lesions BI-RADS 4nad BI-RADS 5 evaluated by either SFM or FFDM have not showed any considerable statisticaldifference between their histopathological allocations. Any considerable statistical difference hasalso not been observed among the histopathological allocation of the non-palpable lesions, whichhad gone through the core- needle biopsy. Core- needle biopsy process takes averagely 30minutes in SFM whereas it is averagely 15 minutes in FFDM process.Discussion and Conclusion:Our study has not traced a considerable numerical differentiation inthe PPVs of SFM and FFDM. Besides, any statistically considerable difference has also not beenobserved between the histopathological allocation of the lesions proposed for biopsy either bySFM or FFDM. Core-needle biopsy method in FFDM reduced the time of the exposure to aconsiderable extend, which helps to improve patients comfort and prevents of time. In addition,manipulation of the images on the soft-copy monitors results in improvement of the lesionsdetective ability; lower dose exposure, better archiving of the images in digital environment andsmoother consultation facility among the central points are the reasons that help the digitalmammography to attract more attention.