In the Vietnamese population, a notably higher incidence of dense breast tissue in women presents significant challenges for mammographic screening, as the dense tissue can obscure underlying lesions, thereby reducing the sensitivity of detection [1]. The American College of Radiology has developed the Breast Imaging-Reporting and Data System (BI-RADS), which provides a uniform system for categorizing findings from breast imaging studies, including mammography, ultrasound, and MRI [2]. This research assesses the effectiveness of BI-RADS assessments when radiologists employ the Transpara Breast AI in mammographic screenings,...

A prospective study was carried out at Hanoi Medical University Hospital from November 2022 to April 2023, including female patients who were scheduled for mammography following ultrasound or manual breast examinations. The selection process was rigorous, employing exclusion criteria to maintain data integrity: Definitive prior diagnostics. Small breast size. Presence of breast implants. A BI-RADS assessment of below 1. After applying these criteria, the study included a final cohort of 1,119 eligible patients. To ensure the study's robustness: Initial mammograms were analyzed by an experienced...

The evaluation of screening sensitivities within mammographic assessments, comparing the AI-supported workflow of Reader 2 against the traditional approach of Reader 1, was meticulously conducted through Receiver Operating Characteristic (ROC) curve assessments, utilizing the BI-RADS scoring system as a standardized measure. Figure 1 illustrates the Receiver Operating Characteristic (ROC) curve comparison analysis, which includes the ROC curves for individual BI-RADS scores.[Fig 1]The results of this comparative study are summarized as follows: Utilizing Reader 1's evaluations as the baseline, the integration of AI with Reader 2's...

In conclusion, we observe that incorporating AI assistance into mammography screening has demonstrated increased sensitivity, a development particularly advantageous for the Vietnamese and broader Southeast Asian populations, characterized by a higher prevalence of dense breast tissue. This study underscores the significant potential of AI-supported mammographic analysis to enhance the accuracy and reliability of breast cancer detection within these groups. This encourages the broader use of AI in breast cancer screening, particularly where dense breast tissue makes diagnosis more difficult.

A. Bhandary Panambur: Other: Siemens Healthineers A. Hoang Dinh: Nothing to disclose L. Chung: Nothing to disclose L. T. Le: Nothing to disclose A. Maier: Nothing to disclose A. Rodriguez Ruiz: Employee: ScreenPoint Medical M. Andrews: Nothing to disclose B. Schmitt: Employee: Siemens Healthineers S. Bayer: Employee: Siemens Healthineers

Trieu PD, Mello-Thoms C, Peat J, Do TD, Brennan PC. Characterization of breast density in Vietnam and its association with demographic, reproductive and lifestyle factors. In Medical Imaging 2017: Image Perception, Observer Performance, and Technology Assessment 2017 Mar 10 (Vol. 10136, pp. 423-428). SPIE. Sickles, E. A., D’Orsi, C. J., Bassett, L. W., et al. (2013). ACR BI-RADS Mammography. In ACR BI-RADS® Atlas, Breast Imaging Reporting and Data System (pp. 121–140). Reston, VA: American College of Radiology.