Elsevier

Clinical Imaging

Volume 30, Issue 3, May–June 2006, Pages 160-165
Clinical Imaging

Original contribution
Evaluation of the three-time-point method for diagnosis of breast lesions in contrast-enhanced MR mammography

https://doi.org/10.1016/j.clinimag.2005.11.005Get rights and content

Abstract

Purpose

To evaluate the three-time-point (3TP) method for diagnosis of breast lesions detected on contrast-enhanced MR mammography.

Materials and Methods

MR imaging was performed in 40 women with 120 suspected breast lesions in mammography and/or sonography. The contrast kinetics was converted by 3TP software on a pixel-by-pixel basis into color-coded images. Lesion diagnosis was made by analysis of color intensity and color hue. The 3TP results were compared with the results of the region-of-interest (ROI) method. In 16 patients, we were able to correlate the results with histopathological findings.

Results

The 3TP method could successfully be performed in all MR mammographies. Forty (33%) lesions had a diameter of less than 5 mm, 56 (47%) lesions between 5 and 10 mm, and 24 (20%) lesions were greater than 1 cm. Of all 120 lesions, 65 (54%) showed heterogeneous contrast enhancement. In 117 (97%) of all 120 lesions the results of ROI and the automated 3TP method were considered equivalent. However, in three lesions the manual ROI differed from the 3TP method. After a second, repeated manual ROI placement, we were able to confirm equivalent results with the 3TP images as well.

Conclusions

The 3TP method automatically and reliably converts contrast kinetic information of the entire breast into a color-coded image. The 3TP method presents kinetic information of the entire dynamic series in an easy-to-interpret format and this automated method may allow to forego time-consuming and sometimes subjective manual ROI placements. This method displays the heterogeneity of the contrast enhancement pattern often observed in malignant lesions and makes it usable as diagnostic criterion.

Introduction

Contrast-enhanced MR mammography is a useful complementary method to conventional diagnostic techniques on certain clinical indications. The clinical applicability of MR mammography is based on the advantages of its very high sensitivity on breast cancer (above 90%), but in contrast, the literature reports varying results concerning specificity (37–86%) [1], [2], [3], [4], [5], [6].

During the last years, a wide range of contrast-enhanced protocols and interpretation criteria were developed and evaluated with the attempt to increase the specificity and standardization of MR mammography [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15]. In addition, Degani et al. [9] and Furman-Haran et al. [10] developed a parametric method for diagnosis breast lesions. This method is based on using high-spatial-resolution images while scanning the images at judiciously selected three time points (one precontrast and two postcontrast time points). The enhancement rate defined by the intensity difference in the first two time points is coded by color intensity, and the change in enhancement between the second and third time points is coded by color hue (red, green, blue).

This method may provide an automated means of beast lesion identification and classification and should enable to more reliably differentiate benign vs. malignant breast lesions [12]. However, the 3TP method also enables an automated and pixel-by-pixel display of the lesion's contrast enhancement pattern. It was our hypothesis that the 3TP method may allow to avoid manual region-of-interest (ROI) placements that are time-consuming and have a large interreader variability and subjectivity.

The aim of our prospective study is to evaluate the three-time-point (3TP) method for its ability to produce equivalent results to manual ROI placement in terms of dynamic enhancement pattern for the diagnosis of breast lesions in contrast-enhanced MR mammography.

Section snippets

Patients

MR imaging was performed in 40 female patients (age range, 32–78 years; mean, 52 years) who were seen in our radiology department for mammographic and/or sonographic abnormalities.

MR imaging

MR imaging was performed on a 1.5-T scanner (Magnetom Sonata; Siemens, Erlangen, Germany), using a dedicated receive-only breast coil (phased-array breast coil; Siemens).

A three-dimensional axial gradient-echo acquisition was performed using the following parameters: TR/TE, 11/4,76 ms; flip angle, 15°; field of view,

Results

We included in the analysis 120 lesions in 40 patients. The average number of lesions was three with a range of 0–11. In 19 patients, lesions were also present in the contralateral breast.

Forty (33%) lesions had a diameter of less than 5 mm, 56 (47%) lesions 5–10 mm, and 24 (20%) lesions were greater than 1 cm (range, 0.3–3.7 cm; mean, 0.8 cm).

Of all 120 lesions, 65 (54%) showed heterogeneous color hue and color intensity with mixed blue, green, or red pixels. Twenty-one of these lesions were

Discussion

Time-intensity curves by manual placement of ROI are known to be useful in the measurement of dynamic enhancement of signal intensity over time following injection of contrast agent [5], [7]. The concentration change over time is predominantly determined by two pathophysiologic parameters that differ in malignant tumors and benign lesions. These parameters are the product to the blood vessel surface area (often referred to as vascular permeability) and the extracellular volume fraction

Conclusions

  • 1.

    The 3TP method automatically and reliably converts contrast kinetic information of the entire breast into a color-coded image.

  • 2.

    The 3TP method presents kinetic information of the entire dynamic series in an easy-to-interpret format and this automated method may allow to forego time-consuming and sometimes subjective manual ROI placements.

  • 3.

    With the 3TP method the entire lesion was analyzed on a pixel-by-pixel basis. Thus, this method displays the heterogeneity of the contrast enhancement pattern

References (15)

  • SH Heywang et al.

    MRI of the breast: histopathologic correlation

    Eur J Radiol

    (1987)
  • R Gilles et al.

    Assessment of breast cancer recurrence with contrast-enhanced subtraction MR imaging: preliminary results in 26 patients

    Radiology

    (1993)
  • U Fischer et al.

    Prognostic value of contrast-enhanced MR mammography in patients with breast cancer

    Eur Radiol

    (1997)
  • F Kelcz et al.

    Application of a quantitative model to differentiate benign from malignant lesions detected by dynamic, gadolinium-enhanced MRI

    J Magn Reson Imaging

    (1996)
  • CK Kuhl et al.

    Dynamic breast MR imaging: are signal intensity time course data useful for differential diagnosis of enhancing lesions?

    Radiology

    (1999)
  • CK Kuhl

    MRI of breast tumors

    Eur Radiol

    (2000)
  • U Fischer et al.

    Signalverhalten maligner und benigner Läsionen in der dynamischen 2D-MRT der Mamma

    Röfo

    (1993)
There are more references available in the full text version of this article.

Cited by (39)

  • Breast DCE-MRI segmentation for lesion detection using Chimp Optimization Algorithm

    2022, Expert Systems with Applications
    Citation Excerpt :

    The analysis of clinically relevant regions of interest showed both a decrease of error (15%–62% reduction following registration) in tissue time-intensity curves and improved Areas Under the Curve (AUC60) at early enhancement. Hauth et al. (2006) described a method for voxel-wise curve categorization of the entire tumor, including voxel washing and washout. On a pixel-by-pixel basis, Three-Time Point (3TP) software translates the contrast kinetics into color-coded graphics.

  • A geometrical perspective on the 3TP method in DCE-MRI

    2015, Biomedical Signal Processing and Control
    Citation Excerpt :

    The three time points and the pseudo-colour mapping were chosen in order both to emphasise the benign, malignant or uncertain nature of the tissue (which is derived from the time-course information) and to optimise the spatial resolution in breast DCE-MRI. Since then the 3TP method has been largely used in DCE-MRI clinical literature [2,6,8,18]. The fact that only three time-points are used to catch the information contained in the TCC makes the method sensitive to noise: one could argue that moderate noise levels superimposed to DCE images can cause benign/malignant tissues to be erroneously assigned to the uncertain class or vice-versa, while higher level of noise could lead to benign tissues confused with malignant and vice-versa.

View all citing articles on Scopus
View full text