Artificial intelligence versus radiologists in detecting early-stage breast cancer from mammograms: a meta-analysis of paradigm shifts.
Hashim Talib Hashim, Ahmed Qasim Mohammed Alhatemi, Motaz Daraghma, Hossam Tharwat Ali, Mudassir Ahmad Khan, Fatimah Abdullah Sulaiman, Zahraa Hussein Ali, Mohanad Ahmed Sahib, Ahmed Dheyaa Al-Obaidi, Ammar Al-Obaidi
Author Information
Hashim Talib Hashim: University of Warith Al-Anbiyaa, College of Medicine, Karbala, Iraq.
Ahmed Qasim Mohammed Alhatemi: Al-Nassiryah Teaching Hospital, Thi Qar, Iraq.
Motaz Daraghma: Faculty of Medicine, Arab American University of Palestine, Jenin, Palestine.
Hossam Tharwat Ali: Qena Faculty of Medicine, South Valley University, Qena, Egypt.
Mudassir Ahmad Khan: Department of Physiology, Shalamar Institute of Health Sciences, Lahore, Pakistan.
Fatimah Abdullah Sulaiman: Al-Nassiryah Teaching Hospital, Nassiryah, Iraq.
Zahraa Hussein Ali: University of Baghdad, College of Medicine, Baghdad, Iraq.
Mohanad Ahmed Sahib: Department of Radiological Techniques, College of Health and Medical Techniques, Al-Mustaqbal University, Babylon, Iraq.
Ahmed Dheyaa Al-Obaidi: MBChB, University of Baghdad, College of Medicine, Baghdad, Iraq.
Ammar Al-Obaidi: Department of Hematology/Oncology, University of Missouri-Kansas City, MO, USA.
Purpose: Early detection of breast cancer is crucial for improving patient outcomes. With advancements in artificial intelligence (AI), there is growing interest in its potential to assist radiologists in interpreting mammograms for early cancer detection. AI algorithms offer the promise of increased accuracy and efficiency in identifying subtle signs of breast cancer, potentially complementing the expertise of radiologists and enhancing the screening process for early-stage breast cancer detection. Material and methods: A systematic literature review was conducted to identify and select original research reports on breast cancer diagnosis by artificial intelligence versus conventional radiologists in using mammograms in accordance with the PRISMA guidelines. Data were analysed with Review Manager version 5.4. -value and were used to test the significance of differences. Results: This systematic review and meta-analysis included 8 studies with data from a total of 120,950 patients. Regarding the sensitivity of AI, the pooled analysis of 6 studies with sensitivities ranging from 0.70 to 0.89 yielded a sensitivity of 0.85. However, the sensitivity of the radiologists ranged from 0.63 to 0.85, with an overall sensitivity of 0.77. As for specificity, both radiologists and AI groups had closer results. Conclusions: The comparison between AI systems and radiologists in detecting early-stage breast cancer from mammograms highlights the potential of AI as a valuable tool in breast cancer screening. While AI algorithms have shown promising results in terms of accuracy and efficiency, they should be viewed as complementary to radiologists rather than replacements.
