2020
DOI: 10.1001/jamanetworkopen.2020.0265
|View full text |Cite|
|
Sign up to set email alerts
|

Abstract: IMPORTANCE 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. OBJECTIVE To evaluate whether AI can overcome human mammography interpretation limitations with a rigorous, unbiased evaluation of machine learning algorithms. DESIGN, SETTING, AND PARTICIPANTS In this diagnostic accuracy study conducted between September 2016 and November 2017, an int… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

5
166
2
2

Year Published

2020
2020
2022
2022

Publication Types

Select...
8

Relationship

3
5

Authors

Journals

citations
Cited by 248 publications
(175 citation statements)
references
References 30 publications
5
166
2
2
Order By: Relevance
“…The specificity at 80% recall improved from 0.768 to Figure 4D), which is lower than community-practice radiologists' specificity of 90.5% on this dataset. 32 However, for both AUC and specificity this algorithm was a co-best performer, and an ensemble of this and other top algorithms with the radiologist's algorithm can further improve on the radiologist's performance. 32 In Figure 4E, we plotted the partial AUC above different levels of recall.…”
Section: Assembling Individual Features Reveals That Integrating Infomentioning
confidence: 94%
See 2 more Smart Citations
“…The specificity at 80% recall improved from 0.768 to Figure 4D), which is lower than community-practice radiologists' specificity of 90.5% on this dataset. 32 However, for both AUC and specificity this algorithm was a co-best performer, and an ensemble of this and other top algorithms with the radiologist's algorithm can further improve on the radiologist's performance. 32 In Figure 4E, we plotted the partial AUC above different levels of recall.…”
Section: Assembling Individual Features Reveals That Integrating Infomentioning
confidence: 94%
“…32 However, for both AUC and specificity this algorithm was a co-best performer, and an ensemble of this and other top algorithms with the radiologist's algorithm can further improve on the radiologist's performance. 32 In Figure 4E, we plotted the partial AUC above different levels of recall. This result, together with the comparison results from the mass model (single breast versus paired breasts), supports the value of the information from the opposite breast in predicting breast cancers with deep learning.…”
Section: Assembling Individual Features Reveals That Integrating Infomentioning
confidence: 94%
See 1 more Smart Citation
“…The model from Ribli et al achieved an auROC of 0.48 when tested on UKy data. Of note, this model came in second in the international Digital Mammography Dialogue on Reverse Engineering Assessment and Methods Challenge (vide infra), with an auROC of 0.85 achieved on the final validation set [6,9].…”
Section: The Challenge Of Developing Generalizable Ai Enabled Analytimentioning
confidence: 99%
“…14 There are several AI tools on the 'roadmap' to full diagnostic use and some have regulatory clearance for such use, for example, in prostate cancer detection. 15 With further evidence, AI could provide double reporting such as has been outlined as a possibility in mammography screening, 16 providing resilience to services when pathologists are not available, and creating further efficiency gains. Collation of evidence to support and validate such tools does however require a digital workflow, and pathologists with time and facility to support such validation processes and integration of such tools into the routine pathology workflow.…”
Section: Digital Pathology and Routine Pathology Servicesmentioning
confidence: 99%