Conclusion
Radiologists must make every effort to reduce the false-negative rate of mammography without unnecessarily increasing the false-positive rate. Perception and interpretation problems account for many missed breast cancers. One must pay careful attention to film review; searching for developing densities, focal asymmetries, and masses; and perform complete diagnostic evaluation of potential lesions detected.
Dr. Shaw de Paredes is Professor of Radiology and Chief of Breast Imaging in the Department of Radiology at the Medical College of Virginia of Virginia Commonwealth University, Richmond, VA.
Acknowledgement
The author gratefully thanks Ms. Louise Logan for her preparation of the manuscript.
12 3 4
Let AuntMinnie.com know what you think about this story.
References
- Baker LH: Breast cancer detection demonstration project: Five-year summary report.
CA Cancer J Clin 32:194-225, 1982.
- Bird RE, Wallace TW, Yankaskas BC: Analysis of cancers missed at screening mammography.
Radiology 184:613-617, 1992.
- Harvey JA, Fajardo LL, Innis CA: Preview mammograms on patients with impalpable breast carcinoma: Retrospective vs blind interpretation.
AJR Am J Roentgenol 161:1167-1172, 1993.
- Bird RE: Professional quality assurance for mammography screening programs.
Radiology 177:587, 1990. Letter.
- Thurfjell EL, Lernvall KA, Taube AAS: Benefit of independent double reading in a population based mammography screening program.
Radiology 191:241-244, 1994.
- Anderson EDC, Muir BB, Walsh JJ, Kirkpatrick AE: The efficacy double reading mammograms in breast screening.
Clin Radiology 49:248-251, 1994.
- Chan HP, Doi K, Vyborny CJ, et al: Improvement in radiologists' detection of clustered microcalcifications on mammograms: The potential of computer-aided diagnosis.
Invest Radiol 15:1102-1110, 1990.
- Wu Y, Giger ML, Coi K, et al: Artifical neural networks in mammography: Application to decision making in the diagnosis of breast cancer.
Radiology 187:81-87, 1993.
- Hendrick RE, Bassett L, Botsco MA, et al: Mammography Quality Control Manual.
Reston, VA, American College of Radiology, 1999.
- Goergen SK, Evans J, Colen GPB, MacMillan JH: Characteristics of breast carcinomas missed by screening radiologists.
Radiology 204:131-135, 1997.
- Kinne DW: Management of the contralateral breast.
In: Harris JR, Hellman S, Henderson P, et al (eds): Breast Diseases. pp. 620-621. Philadelphia: JD Lippincott,1987.
- Huynh PT, Jarolimek AM, Daye S: The False-negative mammogram.
Radiographics 18:1137-1154, 1998.
- Swann CA, Kopans DB, Koerner FC, et al: The halo sign and malignant breast lesions.
AJR Am J Roentgenol 149:1145-1147, 1987.
- Mendelson EB, Harris KM, Doshi N, Tobon H: Infiltrating lobular carcinoma mammographic patterns with pathologic correlation.
AJR Am J Roentgenol 153:265-271, 1989.
- Fournier DV, Weber E, Hoeffken W, et al: Growth rate of 147 mammary carcinomas.
Cancer 45:2198-2207, 1980.
- Lev-Toaff AS, Feig SA, Saitas VL, et al: Stability of malignant breast microcalcifications.
Radiology 192:153-156, 1994.


![A normal mammogram confirmed by three-year radiologic follow-up illustrates reader-marked regions of interest (ROIs) during (A) unaided (round 1) and (B) artificial intelligence (AI)–assisted (round 2) reading. Each colored dot represents an ROI for recall by a human reader. Readers could mark more than one ROI per case, represented by multiple dots of the same color. During AI-assisted reading, the AI system displayed three visible prompts: two with suspicion of malignancy scores of 35% (left mediolateral oblique [L MLO] and craniocaudal [L CC]) and one with a suspicion of malignancy score of 10% (right craniocaudal [R CC]), shown as polygonal overlays. Without AI, six of 10 readers (60%) marked a false-positive ROI. With AI assistance, this fell to two of 10 (20%). R MLO = right mediolateral oblique.](https://img.auntminnie.com/mindful/smg/workspaces/default/uploads/2026/07/2026-07-14-radiology-mammogram-ai-auto-bias.H0bYO8QlWs.jpg?auto=format%2Ccompress&fit=crop&h=100&q=70&w=100)







![A normal mammogram confirmed by three-year radiologic follow-up illustrates reader-marked regions of interest (ROIs) during (A) unaided (round 1) and (B) artificial intelligence (AI)–assisted (round 2) reading. Each colored dot represents an ROI for recall by a human reader. Readers could mark more than one ROI per case, represented by multiple dots of the same color. During AI-assisted reading, the AI system displayed three visible prompts: two with suspicion of malignancy scores of 35% (left mediolateral oblique [L MLO] and craniocaudal [L CC]) and one with a suspicion of malignancy score of 10% (right craniocaudal [R CC]), shown as polygonal overlays. Without AI, six of 10 readers (60%) marked a false-positive ROI. With AI assistance, this fell to two of 10 (20%). R MLO = right mediolateral oblique.](https://img.auntminnie.com/mindful/smg/workspaces/default/uploads/2026/07/2026-07-14-radiology-mammogram-ai-auto-bias.H0bYO8QlWs.jpg?auto=format%2Ccompress&fit=crop&h=112&q=70&w=112)







