Radiological markers or clips are an essential diagnostic and surgical tool. Radiological clips made of titanium or gold are widely used, compatible with MRI and have proven safe for use in human patients. These clips come in many shapes and sizes and are visible with both X-ray and ultrasound. The clips provide large ultrasonic scattering signals because they are made of gold or titanium, which have large impedance mismatch with tissue. For example, patients with locally advanced breast cancer (LABC) undergoing neoadjuvant chemotherapy have clips placed in the tumors so that surgeons can locate and excise, with accuracy, residual cancer tissue. Over the last fifteen years, we have been developing quantitative ultrasound (QUS) imaging techniques for a number of applications including monitoring of therapy response. We have demonstrated that QUS techniques can detect the response of tumors to therapy because QUS is sensitive to the presence of cell death. Recently, we were able to implement this approach in a small clinical study, where we demonstrated the ability of QUS to definitively detect and predict the response of LABC patients to chemotherapy between one and four weeks from therapy onset. We verified that QUS could identify LABC therapy responders and nonresponders. However, we hypothesize that accuracies of QUS estimates in humans can be dramatically improved by incorporating a novel calibration procedure that utilizes appropriate radiological clips as an in situ calibration target. Clinically, these clips are already being placed in tumors for various diagnostic tasks. By inserting an appropriate type of clip, the clip can be used as an in situ calibration providing a reference signal for QUS estimates. The in situ reference will automatically correct for attenuation and transmission losses from overlying tissue layers. Current reference phantom techniques are incomplete in providing corrections for attenuation and transmission losses. Therefore, the use of appropriate radiological clips as calibration targets will provides superior bias and variance of QUS estimates resulting in improved accuracy for identifying LABC response. The scientific premise of the proposed research is that radiological markers used for X-ray and conventional ultrasound can also be used as an in situ calibration target for improving QUS estimates in vivo. Novel calibration procedures will be tested and refined in phantom studies, in animals models of cancer and finally validated in human breast cancer patients undergoing neoadjuvant chemotherapy using QUS to identify early responders to therapy. To verify this scientific premise we propose three specific aims. 1) Develop and refine an in situ calibration approach in phantom studies that accounts for layering effects on QUS estimates thereby improving QUS estimate accuracy. 2) Validate the in situ calibration approach in animal models of cancer and quantify calibration properties in a longitudinal study. 3) Collect longitudinal QUS patient data at early time points during the course of neoadjuvant chemotherapy using a novel in situ calibration procedure and quantify the improvements of QUS classifier accuracy with the in situ calibration approach.
In this proposal, use of radiological clips that can also serve as a calibration target for spectral-based quantitative ultrasound will be assessed for their ability to account for transmission and attenuation losses and thereby improving the accuracy of quantitative ultrasound to identify the early response of breast cancer patients to chemotherapy. Radiological clips are already an integral part of treatment course for patients undergoing neoadjuvant chemotherapy in breast cancer. We will evaluate multiple clip materials and physical sizes through phantom studies, studies in animal models of breast cancer and validate the process in a human study of locally advanced breast cancer patients undergoing neoadjuvant chemotherapy.
