Volumetric mapping of tissue microstructure with OCT for enhanced dysplasia detection

点击翻译按钮获取中文摘要

Project summary Better imaging-based screening and surveillance methods stand to improve patient outcomes in cancers that are typically diagnosed at late stages, such as esophageal adenocarcinoma (EAC). Although success- ful proactive treatment options are available to patients diagnosed with low- or high-grade dysplasia, many at-risk patients are diagnosed at later stages with few effective treatment options, leading to high mortality rates. Dysplasia is diagnosed based on histological evaluation of esophageal biopsies taken un- der the guidance of white light endoscopy (WLE), which in itself does not have the resolution to detect the key morphological changes, namely increases in nuclear size and nuclear cytoplasmic ratio (NCR), that characterize dysplastic progression. Thus, missed diagnoses are common in WLE-screened patients. Recently, optical coherence tomography (OCT), a high-speed, cross-sectional imaging technique was in- troduced as an alternative to WLE based on its ability to visualize glandular anomalies associated with EAC. Although it offers higher resolution than WLE and sub-surface tissue interrogation, OCT remains unable to directly visualize nuclear morphology, limiting its ability to capture or grade low- or high-grade dysplasia. This proposal seeks to amplify current OCT capabilities with novel physics-based signal pro- cessing methods, and accompanying custom optical hardware, to visualize early dysplastic change while retaining a wide field of view. Firstly, we will develop a framework for quantifying nuclear size and NCR in healthy excised tissues based on the measurement of tissue optical properties with OCT. Previous work has fallen short of trans- lating these properties into meaningful dysplastic biomarkers, but we accomplish this with new, more accurate methodology leveraging additional sample information and hardware-based advances. Unlike previous approaches, our new methods will be suitable for characterizing dysplastic tissues with high epi- thelial densities of large nuclei and significant fibrosis. Secondly, we will redesign our multi-measure- ment OCT system for clinical compatibility, enabling full esophageal scans in under ten minutes with en- doscopic imaging hardware. A proof-of-concept pilot study will assess feasibility of deploying our imag- ing platform in patients undergoing upper gastrointestinal endoscopy. Our volumetric tissue microstruc- ture sensing platform will not be limited to investigation of esophageal dysplasia, but generalizable to other malignancies with opportunity for early detection based on morphological changes. We anticipate that our technology will offer significant improvement over traditional imaging methods for biopsy guid- ance, translating into earlier diagnoses, greater opportunity for proactive treatment of disease, and im- proved outcomes in at-risk patients.