RESUMO
Introduction: To compare the accuracy of Artificial Intelligent Breast Ultrasound (AIBUS) with hand-held breast ultrasound (HHUS) in asymptomatic women and to offer recommendations for screening in regions with limited medical resources. Methods: 852 participants who underwent both HHUS and AIBUS were enrolled between December 2020 and June 2021. Two radiologists, who were unaware of the HHUS results, reviewed the AIBUS data and scored the image quality on a separate workstation. Breast imaging reporting and data system (BI-RADS) final recall assessment, breast density category, quantified lesion features, and examination time were evaluated for both devices. The statistical analysis included McNemar's test, paired t-test, and Wilcoxon test. The kappa coefficient and consistency rate were calculated in different subgroups. Results: Subjective satisfaction with AIBUS image quality reached 70%. Moderate agreements were found between AIBUS with good quality images and HHUS for the BI-RADS final recall assessment (κ = 0.47, consistency rate = 73.9%) and breast density category (κ = 0.50, consistency rate = 74.8%). The lesions measured by AIBUS were statistically smaller and deeper than those measured by HHUS (P < 0.001), though they were not significant in clinical diagnosis (all < 3 mm). The total time required for the AIBUS examination and image interpretation was 1.03 (95% CI (0.57, 1.50)) minutes shorter than that of HHUS per case. Conclusion: Moderate agreement was obtained for the description of the BI-RADS final recall assessment and breast density category. With image quality comparable to that of HHUS, AIBUS was superior for the efficiency of primary screening.
RESUMO
Abstract Objective To compare hand-held breast ultrasound (HHBUS) and automated breast ultrasound (ABUS) as screening tool for cancer. Methods A cross-sectional study in patients with mammographically dense breasts was conducted, and both HHBUS and ABUS were performed. Hand-held breast ultrasound was acquired by radiologists and ABUS by mammography technicians and analyzed by breast radiologists. We evaluated the Breast Imaging Reporting and Data System (BI-RADS) classification of the exam and of the lesion, as well as the amount of time required to perform and read each exam. The statistical analysis employed was measures of central tendency and dispersion, frequencies, Student t test, and a univariate logistic regression, through the odds ratio and its respective 95% confidence interval, and with p<0.05 considered of statistical significance. Results Atotal of 440 patientswere evaluated. Regarding lesions,HHBUS detected 15 (7.7%) BI-RADS 2, 175 (89.3%) BI-RADS 3, and 6 (3%) BI-RADS 4, with 3 being confirmed by biopsy as invasive ductal carcinomas (IDCs), and 3 false-positives. Automated breast ultrasound identified 12 (12.9%) BI-RADS 2, 75 (80.7%) BI-RADS 3, and 6 (6.4%) BI-RADS 4, including 3 lesions detected by HHBUS and confirmed as IDCs, in addition to 1 invasive lobular carcinoma and 2 high-risk lesions not detected by HHBUS. The amount of time required for the radiologist to read the ABUS was statistically inferior compared with the time required to read the HHBUS (p<0.001). The overall concordance was 80.9%. A total of 219 lesions were detected, from those 70 lesions by both methods, 126 only by HHBUS (84.9% not suspicious by ABUS) and 23 only by ABUS. Conclusion Compared with HHBUS, ABUS allowed adequate sonographic study in supplemental screening for breast cancer in heterogeneously dense and extremely dense breasts.
Resumo Objetivo Comparar a ultrassonografia convencional das mamas (US) com a ultrassonografia automatizada das mamas (ABUS) no rastreio do câncer. Métodos Realizamos um estudo transversal com pacientes com mamas mamograficamente densas, sendo avaliadas pela US e pela ABUS. A US foi realizada por radiologistas e a ABUS por técnicos de mamografia e analisada por radiologistas especializados em mama. A classificação Breast Imaging Reporting and Data System (BIRADS) do exame e das lesões o tempo de leitura e de aquisição foram avaliados. A análise estatística foi realizada através de medidas de tendência central, dispersão e frequências, teste t de Student e regressão logística univariada, através do odds ratio, com intervalo de confiança de 95%, e com p<0,05 sendo considerado estatisticamente significante. Resultados Foram avaliadas 440 pacientes. Em relação às lesões, a US detectou 15 (7,7%) BI-RADS 2, 175 (89,3%) BI-RADS 3 e 6 (3%) BI-RADS 4, das quais 3 foram confirmadas, por biópsia, como carcinomas ductais invasivos e 3 falso-positivos. A ABUS identificou 12 (12,9%) BI-RADS 2, 75 (80,7%) BI-RADS 3 e 6 (6,4%) BI-RADS 4, incluindo 3 lesões detectadas pela US e confirmadas como carcinomas ductais invasivos, além de 1 carcinoma lobular invasivo e 2 lesões de alto risco não detectadas pela US. O tempo de leitura dos exames da ABUS foi estatisticamente inferior ao tempo do radiologista para realizar a US (p<0,001). A concordância foi de 80,9%. Um total de 219 lesões foram detectadas, das quais 70 por ambos os métodos, 126 observadas apenas pela US (84,9% não eram lesões suspeitas no ABUS) e 23 apenas pela ABUS. Conclusão Comparado à US, a ABUS permitiu adequado estudo complementar no rastreio do câncer de mamas heterogeneamente densas e extremamente densas.
