Time Resolved Breast Imaging Using a Combined MRI and Optical Tomography Approach

原发性全身（新辅助）治疗通常用于局部晚期乳腺癌患者， 手术来降低疾病的分期，增加成功的机会。许多患者虽然 对新辅助治疗没有反应，并且可能最好切换到不同的治疗方案，或 马上进行手术然而，治疗监测是困难的，因为目前的临床 检查和X线/超声乳腺摄影随访与最终治疗病理学相关性较差 结果。磁共振成像（MRI）和正电子发射断层扫描（PET）都已被证实。 评估为反应的早期预测因子，研究显示18F-FDG PET以及弥散 加权（DW）MRI和胆碱化合物MR波谱（tCho-MRS）测量结果符合良好 在治疗开始后的几周内治疗成功。不幸的是，可用性（PET）， 复杂性（tCho-MRS）和特异性（MRI）限制了这些方法的适用性。因此，有一个 需要开发非侵入性的特异性早期预测方法，更容易整合到医疗领域。 实践一个潜在的答案可能是近红外光谱和断层扫描（NIRS-DOT）。 NIRS-DOT是一种新的功能成像技术，可以提供组织发色团的图像，如氧合酶， (HbO)和脱氧血红蛋白（HbR），水和脂质，小型研究暗示其预测的潜力 早在治疗开始后一周就能以高准确度获得治疗结果。此外，最近 技术的进步使DOT达到了高时间分辨率（> 1Hz），允许新类型的 通过动态成像探测功能信息。特别是，我们的小组已经获得了有希望的初步 结果监测乳房组织对外部压迫的反应。组织粘弹性反应 压缩导致血流动力学（血容量）变化，双相时间曲线可能 区分健康组织和乳腺病变。此外，血液动力学和组织氧的相互作用 新陈代谢导致血红蛋白氧合瞬变，这提供了估计组织氧的机会 消耗（OC）和血流量（BF）。 该提案的总体目标是将联合收割机MRI和NIRS-DOT结合起来，以表征预测值 组织血流动力学、血流量和耗氧量的压缩测量， 生物标志物敏感的治疗进展，并量化其与最终病理结果的关系。 来自MRI扫描的结构信息将用作光学重建的先验信息， 动态光学和HbR相关MR血氧水平依赖（BOLD）图像将同时 采集的数据可用于重建时间分辨血流动力学图和BF/OC的融合方法 分布。将根据相应的HbR图交叉验证差异BOLD图像。这项工作将 最终进行临床试验，以评估新生物标志物的早期治疗结果预测能力。

Primary systemic (neoadjuvant) therapy is routinely used for locally advanced breast cancer patients before surgery to down-stage the disease and increase the chances of a successful outcome. Many patients though do not respond to neoadjuvant therapy, and may be better off switching to a different treatment regime, or progressing to surgery immediately. However, therapy monitoring is difficult because current clinical examination and x-ray/ultrasound mammography follow-ups correlate poorly with final therapy pathological outcome. Both magnetic resonance imaging (MRI) and positron emission tomography (PET) have been evaluated as early predictors of response, with studies showing that 18F-FDG PET as well as diffusion weighted (DW) MRI and choline-compounds MR spectroscopy (tCho-MRS) measurements correspond well with therapy success within a few weeks from the beginning of treatment. Unfortunately, availability (PET), complexity (tCho-MRS), and specificity (MRI) limit the applicability of these methods. Consequently, there is a need to develop non-invasive specific early prediction approaches that more easily integrate into medical practice. A potential answer may be offered by near infrared spectroscopy and tomography (NIRS-DOT). NIRS-DOT is a novel functional imaging technique that can offer images of tissue chromophores such as oxy (HbO) and deoxy hemoglobin (HbR), water and lipids, and small studies have hinted at its potential to predict therapy outcome with high accuracy as early as one week after the start of treatment. Further, recent technological advancements have permitted DOT to reach high time resolution (> 1Hz), allowing new types of functional information to be probed by dynamic imaging. In particular, our group has obtained promising initial results monitoring the response of breast tissue to external compression. Tissue viscoelastic response to compression causes hemodynamic (blood volume) changes with bi-phasic temporal profiles likely to differentiate healthy tissues from breast lesions. Further, the interplay of hemodynamics and tissue oxygen metabolism leads to hemoglobin oxygenation transients that offer the opportunity to estimate tissue oxygen consumption (OC) and blood flow (BF) from time-resolved optical data. The overall goal of this proposal is to combine MRI and NIRS-DOT to characterize the predictive value of compression-enabled measurements of tissue hemodynamics, blood flow and oxygen consumption as new biomarkers sensitive to therapy progress and quantify their relationship to final pathological outcome. Structural information from the MRI scans will be used as prior information for the optical reconstructions and dynamic optical and HbR-related MR blood oxygen level dependent (BOLD) images will be simultaneously acquired enabling a fusion approach for reconstructing time-resolved hemodynamic maps and BF/OC distributions. Difference BOLD images will be cross-validated against corresponding HbR maps. The work will culminate with a clinical trial to assess the early therapy outcome predictive ability of the new biomarkers.