Limits...
Computer-aided detection (CAD) for breast MRI: evaluation of efficacy at 3.0 T.

Meeuwis C, van de Ven SM, Stapper G, Fernandez Gallardo AM, van den Bosch MA, Mali WP, Veldhuis WB - Eur Radiol (2009)

Bottom Line: For the same two radiologists the mean sensitivity and specificity for CAD-based interpretation was 90.4% (not significant) and 81.3% (significant at p < 0.05), respectively.With one-way ANOVA no significant differences were found between the two breast radiologists and the two residents together, or between any two readers separately.Automated analysis at 50% and 100% thresholds showed a high sensitivity and specificity for readers with varying levels of experience.

View Article: PubMed Central - PubMed

Affiliation: Department of Radiology, Alysis Zorggroep, Rijnstate Hospital, Wagnerlaan 55, 6815 AD Arnhem, The Netherlands. CMeeuwis@alysis.nl

ABSTRACT

Objective: The purpose of the study was to evaluate the accuracy of 3.0-T breast MRI interpretation using manual and fully automated kinetic analyses.

Material and methods: Manual MRI interpretation was done on an Advantage Workstation. Retrospectively, all examinations were processed with a computer-aided detection (CAD) system. CAD data sets were interpreted by two experienced breast radiologists and two residents. For each lesion automated analysis of enhancement kinetics was evaluated at 50% and 100% thresholds. Forty-nine malignant and 22 benign lesions were evaluated.

Results: Using threshold enhancement alone, the sensitivity and specificity of CAD were 97.9% and 86.4%, respectively, for the 50% threshold, and 97.9% and 90%, respectively, for the 100% threshold. Manual interpretation by two breast radiologists showed a sensitivity of 84.6% and a specificity of 68.8%. For the same two radiologists the mean sensitivity and specificity for CAD-based interpretation was 90.4% (not significant) and 81.3% (significant at p < 0.05), respectively. With one-way ANOVA no significant differences were found between the two breast radiologists and the two residents together, or between any two readers separately.

Conclusion: CAD-based analysis improved the specificity compared with manual analysis of enhancement. Automated analysis at 50% and 100% thresholds showed a high sensitivity and specificity for readers with varying levels of experience.

Show MeSH

Related in: MedlinePlus

Example of a lesion that was not colour-coded by CAD as possibly malignant, because the absolute pre-contrast signal intensity was determined to be “too low”. Manual analysis revealed a washout-type curve for this lesion, with positive threshold-enhancement at both 50% and 100%. Pathology showed an invasive ductal carcinoma
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC2822230&req=5

Fig3: Example of a lesion that was not colour-coded by CAD as possibly malignant, because the absolute pre-contrast signal intensity was determined to be “too low”. Manual analysis revealed a washout-type curve for this lesion, with positive threshold-enhancement at both 50% and 100%. Pathology showed an invasive ductal carcinoma

Mentions: Using threshold enhancement alone, the sensitivity of CAD is high for both the 50% threshold and 100% threshold. There was only one malignant lesion that was false negative at CAD and did not demonstrate enhancement at the 50% and 100% thresholds. This lesion was described by three readers as a lobular-shaped mass with irregular margin and heterogeneous enhancement. One reader described this lesion as a round-shaped mass with regular margin and heterogeneous enhancement. Placing a manual ROI showed a malignant curve (Fig. 3). Three out of the four readers classified this lesion as a BI-RADS 5 lesion, highly suggestive of malignancy, and one reader as a BI-RADS 4 lesion, suspicious. Histopathological evaluation revealed an invasive ductal carcinoma. There are some technical limitations to the CAD program that can cause negative enhancement at malignant lesions which underscore the importance of using CAD as a complement to but not as a replacement for the radiologist’s assessment [5].Fig. 3


Computer-aided detection (CAD) for breast MRI: evaluation of efficacy at 3.0 T.

Meeuwis C, van de Ven SM, Stapper G, Fernandez Gallardo AM, van den Bosch MA, Mali WP, Veldhuis WB - Eur Radiol (2009)

Example of a lesion that was not colour-coded by CAD as possibly malignant, because the absolute pre-contrast signal intensity was determined to be “too low”. Manual analysis revealed a washout-type curve for this lesion, with positive threshold-enhancement at both 50% and 100%. Pathology showed an invasive ductal carcinoma
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC2822230&req=5

Fig3: Example of a lesion that was not colour-coded by CAD as possibly malignant, because the absolute pre-contrast signal intensity was determined to be “too low”. Manual analysis revealed a washout-type curve for this lesion, with positive threshold-enhancement at both 50% and 100%. Pathology showed an invasive ductal carcinoma
Mentions: Using threshold enhancement alone, the sensitivity of CAD is high for both the 50% threshold and 100% threshold. There was only one malignant lesion that was false negative at CAD and did not demonstrate enhancement at the 50% and 100% thresholds. This lesion was described by three readers as a lobular-shaped mass with irregular margin and heterogeneous enhancement. One reader described this lesion as a round-shaped mass with regular margin and heterogeneous enhancement. Placing a manual ROI showed a malignant curve (Fig. 3). Three out of the four readers classified this lesion as a BI-RADS 5 lesion, highly suggestive of malignancy, and one reader as a BI-RADS 4 lesion, suspicious. Histopathological evaluation revealed an invasive ductal carcinoma. There are some technical limitations to the CAD program that can cause negative enhancement at malignant lesions which underscore the importance of using CAD as a complement to but not as a replacement for the radiologist’s assessment [5].Fig. 3

Bottom Line: For the same two radiologists the mean sensitivity and specificity for CAD-based interpretation was 90.4% (not significant) and 81.3% (significant at p < 0.05), respectively.With one-way ANOVA no significant differences were found between the two breast radiologists and the two residents together, or between any two readers separately.Automated analysis at 50% and 100% thresholds showed a high sensitivity and specificity for readers with varying levels of experience.

View Article: PubMed Central - PubMed

Affiliation: Department of Radiology, Alysis Zorggroep, Rijnstate Hospital, Wagnerlaan 55, 6815 AD Arnhem, The Netherlands. CMeeuwis@alysis.nl

ABSTRACT

Objective: The purpose of the study was to evaluate the accuracy of 3.0-T breast MRI interpretation using manual and fully automated kinetic analyses.

Material and methods: Manual MRI interpretation was done on an Advantage Workstation. Retrospectively, all examinations were processed with a computer-aided detection (CAD) system. CAD data sets were interpreted by two experienced breast radiologists and two residents. For each lesion automated analysis of enhancement kinetics was evaluated at 50% and 100% thresholds. Forty-nine malignant and 22 benign lesions were evaluated.

Results: Using threshold enhancement alone, the sensitivity and specificity of CAD were 97.9% and 86.4%, respectively, for the 50% threshold, and 97.9% and 90%, respectively, for the 100% threshold. Manual interpretation by two breast radiologists showed a sensitivity of 84.6% and a specificity of 68.8%. For the same two radiologists the mean sensitivity and specificity for CAD-based interpretation was 90.4% (not significant) and 81.3% (significant at p < 0.05), respectively. With one-way ANOVA no significant differences were found between the two breast radiologists and the two residents together, or between any two readers separately.

Conclusion: CAD-based analysis improved the specificity compared with manual analysis of enhancement. Automated analysis at 50% and 100% thresholds showed a high sensitivity and specificity for readers with varying levels of experience.

Show MeSH
Related in: MedlinePlus