الفهرس 
Title:Efficacy of mammographic artificial intelligence in detecting different histopathological subtypes of breast cancer
AbstractOnly 14 pages are availabe for public view
 |  | from | 139 |  |  |
| | | from | 139 | | |
Abstract
Background:Mammography screening currently relies on subjective human interpretation. Artificial intelligence (AI) advances could be used to increase mammography screening accuracy by reducing missed cancers and false positives.AI-based algorithms can also increase the efficiency of interpretation workflows by reducing workload and interpretation time Methods: The study was prospectively carried on 123 female patients with 134 pathologically proven malignant breast lesion (between December 2020 to June 2021); the mean age was 53.6 ± SD 12.0 years old.Females coming to breast imaging unit either for screening or with breast complaint, basic sono-mammography was done. Artificial intelligence images were automatically generated by AI software (Lunit INSIGHT for mammography) from mammographic images. Biopsieswere done for suspicious breast lesions.Artificial intelligence results as well as mammography results were correlated to the pathology as the gold reference standard. Results: Artificial intelligence has higher sensitivity than mammography in detecting malignant breast lesions; sensitivity of the two methods (AI and mammography) was 96.6% vs 87.3% and false negative rate 3.4% vs 12.7% respectively. Also AI was more sensitive to detect cancers with suspicious mass 95.2% vs 75%, suspicious calcifications 100% vs 86.5% as well as asymmetry and distortion 100% vs 84.6%.AI has better performance in detecting different histopathological subtype of breast malignancy as DCIS, IDC and ILC than mammography with sensitivity (100%, 96.7%, 96.6%) vs (88.9%, 89%, 82.2%) respectively.While in other rare types of breast malignancy both AI and mammography showed the same sensitivity 80%