Potential role of coregistered photoacoustic and ultrasound imaging in ovarian cancer detection and characterization

Abstract

Currently, there is no adequate technology to detect early stage ovarian cancers. Most of the cancers in the ovary are detected when the cancer has already metastasized to other parts of the body. As a result, ovarian cancer has the highest mortality of all gynecologic cancers with a 5-year survival rate of 30% or less. Thus, there is an urgent need to improve the current diagnostic techniques. Photoacoustic imaging (PAI) is an emerging modality with a great potential to assist ultrasound for detecting ovarian cancer noninvasively. In this article, we report the first study of coregistered ultrasound and PAI of 33 ex vivo human ovaries. An assessment of the photoacoustic images has revealed light absorption distribution in the ovary, which is directly related to the vasculature distribution and amount. Quantification of the light absorption levels in the ovary has indicated that, in the postmenopausal group, malignant ovaries showed significantly higher light absorption than normal ones (P = .0237). For these two groups, we have obtained a sensitivity of 83% and a specificity of 83%. This result suggests that PAI is a promising modality for improving ultrasound diagnosis of ovarian cancer.

Figure 1

Experimental setup used for imaging ex vivo ovaries with the coregistered ultrasound and PAI system. Reproduced from Aguirre et al. [28] with permission.

Figure 2

Imaging of a normal premenopausal ovary. (A) Ultrasound image. (B) Coregistered ultrasound and photoacoustic image. (C) H&E staining of the corresponding area (x1). (D) Immunostaining (CD31, x100) revealing microvessels in the theca around the follicles from the rectangular region indicated in B. White bar, 5 mm.

Figure 3

Imaging of an abnormal postmenopausal ovary from a patient with endometrial cancer. (A) Ultrasound image. (B) Coregistered ultrasound and photoacoustic image. (C) Immunostaining (CD31) of the corresponding area (x1). (D) Immunostaining (x100) of the highly vascularized area showing a cluster of small arteries from the rectangular region indicated in B. White bar, 5 mm.

Figure 4

Imaging of a malignant postmenopausal ovary. (A) Coregistered ultrasound and photoacoustic images of two different locations are shown in B and C. (B1 and C1) H&E stains (x40) of the corresponding areas showing extensive high-grade tumors. (B2 and C2) CD31 stains (x100) of the corresponding areas showing extensive thin-walled microvessels. White bar, 5 mm.

Figure 5

Scatter plot of the measured AMRFS values of four groups. The vertical axis is the AMRFS, and the horizontal axis presents four groups. A threshold is chosen as the average of all the AMRFS measurements of all ovaries. For the postmenopausal group, the sensitivity and specificity between normal and malignant ovaries is 83%.

Figure 6

Statistics of AMRFS of all subgroups of postmenopausal ovaries. Statistical significance between normal andmalignant ovaries is shown, as well as between normal and abnormal and malignant.

Figure 7

Statistics of AMRFS of all normal ovaries.

Figure 8

Average AMRFS measured with PAI versus absorption coefficient of ovaries imaged with a diffused optical tomography system (n = 24). Linear regression analysis obtained a 0.59 correlation coefficient, which is statistically significant (P < .05).