Title

Improving the light quantification of near infrared (NIR) diffused light optical tomography with ultrasound localization

Date of Completion

January 2010

Keywords

Engineering, Biomedical|Engineering, Electronics and Electrical|Health Sciences, Radiology|Physics, Optics

Degree

Ph.D.

Abstract

According to the statistics published by the American Cancer Society, currently breast cancer is the second most common cancer after skin cancer and the second cause of cancer death after lung cancer in the female population. Diffuse optical tomography (DOT) using near-infrared (NIR) light, guided by ultrasound localization, has shown great promise in distinguishing benign from malignant breast tumors and in assessing the response of breast cancer to chemotherapy. ^ Our ultrasound-guided DOT system is based on reflection geometry, with patients scanned in supine position using a hand-held probe. For patients with chest-wall located at a depth shallower than 1 to 2cm, as in about 10% of our clinical cases, the semi-infinite imaging medium is not a valid assumption and the chest-wall effect needs to be considered in the imaging reconstruction procedure. ^ In this dissertation, co-registered ultrasound images were used to model the breast-tissue and chest-wall as a two-layer medium. The effect of the chest wall on breast lesion reconstruction was systematically investigated. The performance of the two-layer model-based reconstruction, using the Finite Element Method, was evaluated by simulation, phantom experiments and clinical studies. The results show that the two-layer model can improve the accuracy of estimated background optical properties, the reconstructed absorption map and the total hemoglobin concentration of the lesion. ^ For patients' data affected by chest wall, the perturbation, which is the difference between measurements obtained at lesion and normal reference sites, may include the information of background mismatch between these two sites. Because the imaging reconstruction is based on the perturbation approach, the effect of this mismatch between the optical properties at the two sites on reconstructed optical absorption was studied and a guideline for imaging procedure was developed to reduce these effects during data capturing. ^ To reduce the artifacts caused by the background mismatch between the lesion and reference sites, two solutions were introduced. The first solution uses a model-based approach and the second method uses an exogenous contrast agent. The results of phantom and animal studies show that both methods can significantly reduce artifacts generated by the background mismatch. ^