Journal Basic Info

  • Impact Factor: 1.995**
  • H-Index: 8
  • ISSN: 2474-1647
  • DOI: 10.25107/2474-1647
**Impact Factor calculated based on Google Scholar Citations. Please contact us for any more details.

Major Scope

  •  Oral and Maxillofacial Surgery
  •  Pediatric Surgery
  •  Vascular Surgery
  •  Transplant Surgery
  •  Robotic Surgery
  •  Gastroenterological Surgery
  •  General Surgery
  •  Breast Surgery

Abstract

Citation: Clin Surg. 2021;6(1):3196.Research Article | Open Access

Computer Assisted Detection (CAD) - Negative Recalls in Screening Mammography: What are the Characteristics of Cancers Missed by CAD?

Pasteur Rasuli1*, Gerard Collin2 , Allan Gitterman3 , Stavroula Karmanos3 and Jean Seely1

1 Departments of Radiology, The Ottawa Hospital, Canada 2 Departments of Radiology, Montfort Hospital, Canada 3 Departments of Radiology, Queensway-Carleton Hospital, Canada

*Correspondance to: Pasteur Rasuli 

 PDF  Full Text DOI: 10.25107/2474-1647.3196

Abstract

Purpose: The purpose of this prospective study was to determine the characteristics of cancers missed by CAD and to determine if the missed cancers were significant. Methods: The study included 37,752 consecutive women aged 50 or older undergoing screening mammography through an organized screening program in Ontario Canada. All mammograms obtained with Hologic units read using R2 CAD Image Checker with intermediate sensitivity. All CAD-negative recalls were investigated with appropriate diagnostic workup and biopsies as needed. Results: 37,752 women were screened in the study period, with an invasive cancer detection rate of 5.9/1000, and the abnormal call rate was 4.7%. Of the 1,789 recalled cases, 109 were true positive cancers. The interval cancer rate was 15 per 12,000 in 12 months. 108 (6%) of the recalled cases were CAD-negative. Of these, 7 (6%) were found to be malignant representing a Negative Predictive Value (NPV) of 93.5% (101/108) for CAD-negative recalls. Comparatively, among the 1,681 CAD positive-recalls, 368 (21%) were found to represent breast cancer. Of the 7 missed cancers by CAD, four lesions were 7 mm or less in diameter, of which three were invasive and three were in situ ductal carcinomas. The seventh lesion was an invasive lobular carcinoma measuring 9 mm ? 15 mm ? 17 mm in diameter. All missed lesions were masses without any mammographically perceptible calcifications. One lesion was discovered on the first screen and six were on rescreens. Conclusion: CAD has a high NPV, and CAD negative recalls are likely to yield small cancers at early stages.

Keywords

Cite the article

Rasuli P, Collin G, Gitterman A, Karmanos S, Seely J. Computer Assisted Detection (CAD) - Negative Recalls in Screening Mammography: What are the Characteristics of Cancers Missed by CAD?. Clin Surg. 2021; 6: 3196..

Search Our Journal

Journal Indexed In

Articles in PubMed

RAF Kinase Inhibitory Protein Expression and Phosphorylation Profiles in Oral Cancers
 PubMed  PMC  PDF  Full Text
Automated Sagittal Craniosynostosis Classification from CT Images Using Transfer Learning
 PubMed  PMC  PDF  Full Text
View More...

Articles with Grants

Laparoscopic Versus Robot Techniques for the Choledochal Cyst in Children: An Initial Meta-Analysis
 Abstract  PDF  Full Text
Operation Field Contamination during Intraoperative Fluoroscopy
 Abstract  PDF  Full Text
View More...