Photoacoustic imaging of colorectal cancer and ovarian cancer

Photoacoustic (PA) imaging is an emerging hybrid imaging technology that uses a short-pulsed laser to excite tissue. The resulting photoacoustic waves are used to image the optical absorption distribution of the tissue, which is directly related to micro-vessel networks and thus to tumor angiogenesis, a key process in tumor growth and metastasis. In this thesis, the acoustic-resolution photoacoustic microscopy (AR-PAM) was first investigated on its role in human colorectal tissue imaging, and the optical-resolution photoacoustic microscopy (OR-PAM) was investigated on its role in human ovarian tissue imaging.Colorectal cancer is the second leading cause of cancer death in the United States. Significant limitations in screening and surveillance modalities continue to hamper early detection of primary cancers or recurrences after therapy. In the first phase of the study, benchtop co-registered ultrasound (US) and AR-PAM systems were constructed and tested in ex vivo human colorectal tissue. In the second phase of the study, a co-registered endorectal AR-PAM imaging system was constructed, and a pilot patient study was conducted on patients with rectal cancer treated with radiation and chemotherapy. To automate the data analysis, we designed and trained convolutional neural networks (PAM-CNN and US-CNN) using mixed ex vivo and in vivo patient data. 22 patients’ ex vivo specimens and five patients’ in vivo images (a total of 2693 US ROIs and 2208 PA ROIs) were used for CNN training and validation. Data from five additional patients were used for testing. A total of 32 participants (mean age, 60 years, range, 35-89 years) were evaluated. Unique PAM imaging markers of complete tumor response were found, specifically recovery of normal submucosal vascular architecture within the treated tumor bed. The PAM-CNN model captured this recovery process and correctly differentiated these changes from a residual tumor tissue. The imaging system remained highly capable of differentiating tumor from normal tissue, achieving an area under receiver operating characteristic curve (AUC) of 0.98 from the five patients tested. By comparison, US-CNN had an AUC of 0.71. As an alternative to CNN, a generalized linear model (GLM) was investigated for classification and results showed that CNN outperformed GLM in classification of both US and PAM images. Ovarian cancer is the leading cause of death among gynecological cancers but is poorly amenable to preoperative diagnosis. In the second project of this thesis, we have investigated the feasibility of “optical biopsy,” using OR-PAM to quantify the microvasculature of ovarian tissue and fallopian tube tissue. The technique was demonstrated using excised human ovary and fallopian tube specimens imaged immediately after surgery. Initially, a commercial software Amira was used to characterize tissue vasculature patterns, and later, an effective and easy-access algorithm was developed to quantify the mean diameter, total length, total volume, and fulfillment rate of tissue vasculature. Our initial results demonstrate the potential of OR-PAM as an imaging tool for quick assessment of ovarian tissue and fallopian tube tissue