Early Prediction of Breast Cancer Therapy Response using Multiresolution Fractal Analysis of DCE-MRI Parametric Maps

Archana Machireddy, Guillaume Thibault, Luminita (Alina) Tudorica, Aneela Afzal, May Mishal, Kathleen Kemmer, Arpana Naik, Megan Troxell, Eric Goranson, Karen Oh, Nicole Roy, Neda Jafarian, Megan Holtorf, Wei Huang, Xubo Song

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

We aimed to determine whether multiresolution fractal analysis of voxel-based dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) parametric maps can provide early prediction of breast cancer response to neoadjuvant chemotherapy (NACT). In total, 55 patients underwent 4 DCE-MRI examinations before, during, and after NACT. The shutter-speed model was used to analyze the DCE-MRI data and generate parametric maps within the tumor region of interest. The proposed multiresolution fractal method and the more conventional methods of single-resolution fractal, gray-level co-occurrence matrix, and run-length matrix were used to extract features from the parametric maps. Only the data obtained before and after the first NACT cycle were used to evaluate early prediction of response. With a training (N = 40) and testing (N = 15) data set, support vector machine was used to assess the predictive abilities of the features in classification of pathologic complete response versus non-pathologic complete response. Generally the multiresolution fractal features from individual maps and the concatenated features from all parametric maps showed better predictive performances than conventional features, with receiver operating curve area under the curve (AUC) values of 0.91 (all parameters) and 0.80 (Ktrans), in the training and testing sets, respectively. The differences in AUC were statistically significant (P < .05) for several parametric maps. Thus, multiresolution analysis that decomposes the texture at various spatial-frequency scales may more accurately capture changes in tumor vascular heterogeneity as measured by DCE-MRI, and therefore provide better early prediction of NACT response.

Original languageEnglish (US)
Pages (from-to)90-98
Number of pages9
JournalTomography (Ann Arbor, Mich.)
Volume5
Issue number1
DOIs
StatePublished - Mar 1 2019

Fingerprint

Fractals
Magnetic Resonance Imaging
Breast Neoplasms
Drug Therapy
Area Under Curve
Therapeutics
Blood Vessels
Neoplasms

Keywords

  • breast cancer
  • DCE-MRI
  • early prediction
  • multiresolution fractal analysis
  • neoadjuvant chemotherapy

Cite this

Early Prediction of Breast Cancer Therapy Response using Multiresolution Fractal Analysis of DCE-MRI Parametric Maps. / Machireddy, Archana; Thibault, Guillaume; Tudorica, Luminita (Alina); Afzal, Aneela; Mishal, May; Kemmer, Kathleen; Naik, Arpana; Troxell, Megan; Goranson, Eric; Oh, Karen; Roy, Nicole; Jafarian, Neda; Holtorf, Megan; Huang, Wei; Song, Xubo.

In: Tomography (Ann Arbor, Mich.), Vol. 5, No. 1, 01.03.2019, p. 90-98.

Research output: Contribution to journalArticle

@article{225d7b68008846b481e5753e669ad7f7,
title = "Early Prediction of Breast Cancer Therapy Response using Multiresolution Fractal Analysis of DCE-MRI Parametric Maps",
abstract = "We aimed to determine whether multiresolution fractal analysis of voxel-based dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) parametric maps can provide early prediction of breast cancer response to neoadjuvant chemotherapy (NACT). In total, 55 patients underwent 4 DCE-MRI examinations before, during, and after NACT. The shutter-speed model was used to analyze the DCE-MRI data and generate parametric maps within the tumor region of interest. The proposed multiresolution fractal method and the more conventional methods of single-resolution fractal, gray-level co-occurrence matrix, and run-length matrix were used to extract features from the parametric maps. Only the data obtained before and after the first NACT cycle were used to evaluate early prediction of response. With a training (N = 40) and testing (N = 15) data set, support vector machine was used to assess the predictive abilities of the features in classification of pathologic complete response versus non-pathologic complete response. Generally the multiresolution fractal features from individual maps and the concatenated features from all parametric maps showed better predictive performances than conventional features, with receiver operating curve area under the curve (AUC) values of 0.91 (all parameters) and 0.80 (Ktrans), in the training and testing sets, respectively. The differences in AUC were statistically significant (P < .05) for several parametric maps. Thus, multiresolution analysis that decomposes the texture at various spatial-frequency scales may more accurately capture changes in tumor vascular heterogeneity as measured by DCE-MRI, and therefore provide better early prediction of NACT response.",
keywords = "breast cancer, DCE-MRI, early prediction, multiresolution fractal analysis, neoadjuvant chemotherapy",
author = "Archana Machireddy and Guillaume Thibault and Tudorica, {Luminita (Alina)} and Aneela Afzal and May Mishal and Kathleen Kemmer and Arpana Naik and Megan Troxell and Eric Goranson and Karen Oh and Nicole Roy and Neda Jafarian and Megan Holtorf and Wei Huang and Xubo Song",
year = "2019",
month = "3",
day = "1",
doi = "10.18383/j.tom.2018.00046",
language = "English (US)",
volume = "5",
pages = "90--98",
journal = "Tomography (Ann Arbor, Mich.)",
issn = "2379-1381",
publisher = "Grapho Publications LLC",
number = "1",

}

TY - JOUR

T1 - Early Prediction of Breast Cancer Therapy Response using Multiresolution Fractal Analysis of DCE-MRI Parametric Maps

AU - Machireddy, Archana

AU - Thibault, Guillaume

AU - Tudorica, Luminita (Alina)

AU - Afzal, Aneela

AU - Mishal, May

AU - Kemmer, Kathleen

AU - Naik, Arpana

AU - Troxell, Megan

AU - Goranson, Eric

AU - Oh, Karen

AU - Roy, Nicole

AU - Jafarian, Neda

AU - Holtorf, Megan

AU - Huang, Wei

AU - Song, Xubo

PY - 2019/3/1

Y1 - 2019/3/1

N2 - We aimed to determine whether multiresolution fractal analysis of voxel-based dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) parametric maps can provide early prediction of breast cancer response to neoadjuvant chemotherapy (NACT). In total, 55 patients underwent 4 DCE-MRI examinations before, during, and after NACT. The shutter-speed model was used to analyze the DCE-MRI data and generate parametric maps within the tumor region of interest. The proposed multiresolution fractal method and the more conventional methods of single-resolution fractal, gray-level co-occurrence matrix, and run-length matrix were used to extract features from the parametric maps. Only the data obtained before and after the first NACT cycle were used to evaluate early prediction of response. With a training (N = 40) and testing (N = 15) data set, support vector machine was used to assess the predictive abilities of the features in classification of pathologic complete response versus non-pathologic complete response. Generally the multiresolution fractal features from individual maps and the concatenated features from all parametric maps showed better predictive performances than conventional features, with receiver operating curve area under the curve (AUC) values of 0.91 (all parameters) and 0.80 (Ktrans), in the training and testing sets, respectively. The differences in AUC were statistically significant (P < .05) for several parametric maps. Thus, multiresolution analysis that decomposes the texture at various spatial-frequency scales may more accurately capture changes in tumor vascular heterogeneity as measured by DCE-MRI, and therefore provide better early prediction of NACT response.

AB - We aimed to determine whether multiresolution fractal analysis of voxel-based dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) parametric maps can provide early prediction of breast cancer response to neoadjuvant chemotherapy (NACT). In total, 55 patients underwent 4 DCE-MRI examinations before, during, and after NACT. The shutter-speed model was used to analyze the DCE-MRI data and generate parametric maps within the tumor region of interest. The proposed multiresolution fractal method and the more conventional methods of single-resolution fractal, gray-level co-occurrence matrix, and run-length matrix were used to extract features from the parametric maps. Only the data obtained before and after the first NACT cycle were used to evaluate early prediction of response. With a training (N = 40) and testing (N = 15) data set, support vector machine was used to assess the predictive abilities of the features in classification of pathologic complete response versus non-pathologic complete response. Generally the multiresolution fractal features from individual maps and the concatenated features from all parametric maps showed better predictive performances than conventional features, with receiver operating curve area under the curve (AUC) values of 0.91 (all parameters) and 0.80 (Ktrans), in the training and testing sets, respectively. The differences in AUC were statistically significant (P < .05) for several parametric maps. Thus, multiresolution analysis that decomposes the texture at various spatial-frequency scales may more accurately capture changes in tumor vascular heterogeneity as measured by DCE-MRI, and therefore provide better early prediction of NACT response.

KW - breast cancer

KW - DCE-MRI

KW - early prediction

KW - multiresolution fractal analysis

KW - neoadjuvant chemotherapy

UR - http://www.scopus.com/inward/record.url?scp=85068624107&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85068624107&partnerID=8YFLogxK

U2 - 10.18383/j.tom.2018.00046

DO - 10.18383/j.tom.2018.00046

M3 - Article

VL - 5

SP - 90

EP - 98

JO - Tomography (Ann Arbor, Mich.)

JF - Tomography (Ann Arbor, Mich.)

SN - 2379-1381

IS - 1

ER -