Tissue elasticity properties as biomarkers for prostate cancer

Abstract

In this paper we evaluate tissue elasticity as a longstanding but qualitative biomarker for prostate cancer and sonoelastography as an emerging imaging tool for providing qualitative and quantitative measurements of prostate tissue stiffness. A Kelvin-Voigt Fractional Derivative (KVFD) viscoelastic model was used to characterize mechanical stress relaxation data measured from human prostate tissue samples. Mechanical testing results revealed that the viscosity parameter for cancerous prostate tissue is greater than that derived from normal tissue by a factor of approximately 2.4. It was also determined that a significant difference exists between normal and cancerous prostate tissue stiffness (p < 0.01) yielding an average elastic contrast that increases from 2.1 at 0.1 Hz to 2.5 at 150 Hz. Qualitative sonoelastographic results show promise for cancer detection in prostate and may prove to be an effective adjunct imaging technique for biopsy guidance. Elasticity images obtained with quantitative sonoelastography agree with mechanical testing and histological results. Overall, results indicate tissue elasticity is a promising biomarker for prostate cancer.

Fig. 1

(a) Stress relaxation curves from normal (black) and cancerous (gray) tissue specimens obtained from the same prostate. The KVFD fit is shown (line) for the cancerous specimen. (b) The complex Young’s moduli of the two types of prostate tissue are plotted as a function of frequency revealing a relationship between tissue stiffness and cancer percentage.

Fig. 2

Frequency-dependent average magnitude plots for complex Young’s moduli corresponding to normal (black circles) and cancerous (gray squares) prostate tissue.

Fig. 3

Experimental setup for qualitative (left) and quantitative (right) sonoelastographic imaging experiments. Both illustrations depict the position of the (a) US transducer and the (b) Material under investigation. For the qualitative sonoelastographic experimental setup, the (c) double bar material contact and (d) mechanical source are shown. Conversely, for the qualitative sonoelastographic setup the (e) bending piezoelectric actuators are represented.

Fig. 4

Matched (a) B-mode ultrasound, (b) Sonoelastographic, and (c) Histological images. A deficit in the sonoelastogram (red arrows) was verified as a cancerous mass by histology (blue outline). Note that sonoelastographic regions where the vibration amplitude is low are shown as dark green, while regions with high vibration are depicted as bright green.

Fig. 5

Three-dimensional reconstruction of a prostate gland from (a) Histological and (b) US and sonoelastographic images. For both results, the surface of the gland is shown in blue with histological and sonoelastographic tumors depicted as red and green, respectively. The registered 3D image detailing sonoelastographic (green) and histological (red) findings is shown in (c). The intersection of both is presented in white.

Fig. 6

Matched (a) B-mode US and (b) Sonoelastographic images from an in vivo prostate study. The sonoelastographic image reveals a stiff (cancerous) mass (denoted by arrows) in the middle of the image.

Fig. 7

Results from an in vivo study illustrating (a) 3D reconstruction from the prostate scan and (b) Histological image taken from the midgland region. Sonoelastography depicts two stiff cancerous masses (green) that are corroborated by histology (blue outline). Note a small tumor was missed by sonoelastography in the anterior right part of the gland.

Fig. 8

Summary of quantitative elasticity measurements from crawling wave sonoelastographic images (black circles) and mechanical testing using a KVFD model (gray squares). Results are depicted for two different prostate cases and experimental conditions.

Fig. 9

Experimental quantitative sonoelastographic imaging results depicting (a) B-mode US image, (b) Crawling wave sonoelastogram, (c) Quantitative (i.e., Young’s modulus) sonoelastogram (units of kPa), and (d) Histological image.