Performed at one institution, Diagnostic or prognostic study, Prospective, Cancer, Screening, Diagnostic procedure, Ultrasound, Mammography, Breast
M. Ershtein; Moscow/RU
Results were categorized by BI-RADS and nosology. Diagram 1 (Figure 2) shows that the majority of patients were classified as healthy - 26% and with benign non-mass changes – 41% (cysts, etc.). Benign mass lesions (fibroadenomas, etc.) were detected in 21% of cases. Malignant lesions were detected in 12%, including one DCIS and two IDC up to 0.5 cm.
Compliance of the results of HHUS and ABUS was 95.45% (p<0.05). The discrepancies concerned the differential diagnosis of only benign lesions less than 0.8 cm, in 3 cases BI-RADS4a ABUS and BI-RADS3 HHUS confirmed leaf-shaped tumor and fibroadenomas. Comparing the results for categories BI-RADS 4 bc, 5 (46 cases, 19% of respondents), no discrepancies were found. Histological confirmation of cancer was obtained in 30 cases (65.2% of cases of BI-RADS4bc, 5), among which one DCIS and two IDC up to 0.5 cm.
In the study of multiple lesions (multicentric cancer - 3 cases, multiple fibroadenomas - 4 cases), ABUS allowed to reliably evaluate their relative position, it helped to plan surgical treatment better (Figure 3).
In a complex study with cytological and histological verification, an increase in the sensitivity of the examination is noted with a combination of two methods against one mammography with a slight decrease in specificity, as shown in Diagram 2 (Figure 4). The decrease in specificity is due to the receipt of more data and does not significantly affect the overall increase in research efficiency.
ROC analysis curves show the effectiveness of the studied methods. In the analysis of prognostic value, the area under the curve (AUC) was estimated: MG = 0.807687, MG + HHUS = 0.89075, MG + ABUS = 0.89275. Thus, HHUS and ABUS have the same added value to MG. The use of MG in combination with ultrasound techniques increases the efficiency of diagnosis.
Due to standardization and automation, the time spent by a doctor on one study was reduced by an average of 2 times (3,5-7 minutes vs. 14).