Selbstorganisierende neuronale Netze zur automatischen Detektion und Klassifikation von Kontrast(mittel)-verstärkten Läsionen in der dynamischen MR-Mammographie

[Quantitative parametric analysis of contrast-enhanced lesions in dynamic MR mammography]

PURPOSE

The aim of the study was an evaluation of the quantitative parametric analysis of contrast-enhanced lesions in dynamic MR mammography.

MATERIAL AND METHODS

In 137 patients, 183 contrast-enhanced lesions were identified in dynamic MR mammography. In 82 lesions histopathology was performed and in 101 lesions follow-up MR mammography was carried out. The contrast kinetics of lesions was analyzed quantitatively, on a pixel-by-pixel basis. The initial signal enhancement was coded by color intensity (bright, medium, dark), the post-initial signal enhancement was coded by color hue (blue, green, red). ROC analysis and logistic regression were performed.

RESULTS

Malignant lesions showed a significantly higher number of bright red, medium red and dark red, bright green and medium green pixels than benign lesions. Benign lesions showed a significantly higher number of bright blue, medium blue and dark blue pixels than malignant lesions. The highest areas under the ROC curves (AUC) were found for medium red (AUC = 0.782) and medium green pixels (AUC = 0.733). A regression model with medium red and medium green pixels allows diagnosis of malignant lesions with a sensitivity of 60.7% and a specificity of 83.6%.

CONCLUSIONS

The quantification of contrast-enhanced lesions allows objective analysis of the signal intensities in malignant and benign lesions. Therefore, this method might increase the specificity of MR mammography. Further developments are necessary before this method can be used for routine analysis of contrast-enhancing lesions in MR mammography.

Quantitative 2- and 3-dimensional analysis of pharmacokinetic model-derived variables for breast lesions in dynamic, contrast-enhanced MR mammography

PURPOSE

2- and 3-dimensional evaluation of quantitative pharmacokinetic parameters derived from the Tofts model modeling dynamic contrast enhancement of lesions in MR mammography.

MATERIALS AND METHODS

In 95 patients, MR mammography revealed 127 suspicious lesions. The initial rate of enhancement was coded by color intensity, the post-initial enhancement change is coded by color hue. 2D and 3D analysis of distribution of color hue and intensity, vascular permeability and extracellular volume were performed.

RESULTS

In 2D, malignant lesions showed significant higher number of bright red, medium red, dark red, bright green, medium green, dark green and bright blue pixels than benign lesions. In 3D, statistical significant differences between malignant and benign lesions was found for all this parameters. Vascular permeability was significant higher in malignant lesions than in benign lesions. Regression model using the 3D data found that the best discriminator between malignant and benign lesions was combined number of voxels and medium green pixels, with a sensitivity of 79.4% and a specificity of 83.1%.

CONCLUSIONS

Quantitative analysis of pharmacokinetic variables of contrast kinetics showed significant differences between malignant and benign lesions. 3D analysis showed superior diagnostic differentiation between malignant and benign lesions than 2D analysis. The parametric analysis using a pharmacokinetic model allows objective analysis of contrast enhancement in breast lesions.

Classification of small contrast enhancing breast lesions in dynamic magnetic resonance imaging using a combination of morphological criteria and dynamic analysis based on unsupervised vector-quantization

PURPOSE

To evaluate the diagnostic value of breast magnetic resonance imaging (MRI) in small focal lesions using dynamic analysis based on unsupervised vector quantization in combination with a score for morphologic criteria.

MATERIALS AND METHODS

We examined 85 mammographically indetermintate lesions (BIRADS 3-4; 47 malignant, mean lesion size 1.2 cm; 38 benign, mean lesion size 1.1 cm). MRI was performed with a dynamic T1-weighted gradient echo sequence (1 precontrast and 5 postcontrast series). Lesions with an initial contrast enhancement >/=50% were selected with semiautomatic segmentation. For conventional dynamic analysis, we calculated the mean initial signal increase and postinitial course of all voxels included in a lesion. Secondly, all voxels within the lesions were assigned to 4 clusters using minimal-free-energy vector quantization. Dynamic and morphologic criteria were summarized in a diagnostic score and evaluated by receiver operating characteristic analysis.

RESULTS

In the present collection of small lesions, morphologic criteria [area under the curve (AUC) = 0.610] were inferior to dynamic criteria in the detection of breast cancer. Dynamic analysis with vector quantization (AUC = 0.760) presented slightly better results compared with standard dynamic analysis (AUC = 0.693). There was no benefit for combined morphologic and dynamic analysis.

CONCLUSION

In small MR-mammographic lesions, dynamic analysis with vector quantization alone tends to result in a higher diagnostic accuracy compared with combined morphologic and dynamic analysis.

Cluster analysis of signal-intensity time course in dynamic breast MRI: does unsupervised vector quantization help to evaluate small mammographic lesions?

We examined whether neural network clustering could support the characterization of diagnostically challenging breast lesions in dynamic magnetic resonance imaging (MRI). We examined 88 patients with 92 breast lesions (51 malignant, 41 benign). Lesions were detected by mammography and classified Breast Imaging and Reporting Data System (BIRADS) III (median diameter 14 mm). MRI was performed with a dynamic T1-weighted gradient echo sequence (one precontrast and five postcontrast series). Lesions with an initial contrast enhancement >or=50% were selected with semiautomatic segmentation. For conventional analysis, we calculated the mean initial signal increase and postinitial course of all voxels included in a lesion. Secondly, all voxels within the lesions were divided into four clusters using minimal-free-energy vector quantization (VQ). With conventional analysis, maximum accuracy in detecting breast cancer was 71%. With VQ, a maximum accuracy of 75% was observed. The slight improvement using VQ was mainly achieved by an increase of sensitivity, especially in invasive lobular carcinoma and ductal carcinoma in situ (DCIS). For lesion size, a high correlation between different observers was found (R(2) = 0.98). VQ slightly improved the discrimination between malignant and benign indeterminate lesions (BIRADS III) in comparison with a standard evaluation method.