الفهرس 
AnatomyOnly 14 pages are availabe for public view
 |  | from | 113 |  |  |
| | | from | 113 | | |
Abstract
Background: Breast cancer is the most common cancer type and the second cause of cancer-based mortality in women according to the 2020 global cancer statistics. Screening for breast cancer with mammography has shown a reduction in breast cancer mortality by many randomized trials and incidence-based mortality studies.
Aim of the Work: to emphasize the role of AI system in detection of breast cancer in digital mammogram. We hypothesise that the AI has accuracy that is comparable to human readers in breast cancer detection on Mammogram, and that integrating the AI into a standard screen-reading strategy increase the accuracy of cancer detection.
Patients and Methods: This is a prospective study (Diagnostic Accuracy Testing) conducted at the radiology department, Ain Shams University Hospitals. The main source of data for this study were the patients referred to the radiology department at Ain Shams university hospitals for screening mammography from March 2023 to August 2023.
Results: In our study 62.5% of the pathological lesions were on the right side, while 37.5% were on at the left side. - Regarding the agreement between radiologists and AI in the nature of the detected mass (Bengin / Malignant), the agreement between radiologist 1 Vs. radiologist 2 was 92.5%, while it was 86.5% between radiologist 1 Vs. AI. The agreement between radiologists 2 Vs. AI, it was 89.2%, So the best agreement was between radiologist 1 and radiologist 2,While the concordance between the analysis of the mammography and the pathological results was the highest by Radiologist 2 by accuracy 85% and sensitivity 100% followed by radiologist 1 as accuracy was 77.5% with 93.9% sensitivity, while the worst interpretation was by AI accuracy was 72.97% with 90% sensitivity, so the concordance between radiologists’ analysis and pathological results was stronger than that between AI and pathology.
Conclusion: We found that the unaided radiologist are having overall better performance than AI however, less experienced breast radiologists had a higher diagnostic performance with support from an AI system compared with reading unaided.