Development of a label-free and multimodal optical spectroscopy imaging platform for biological tissue characterization and classification based on biophysical and statistical modeling techniques

点击翻译按钮获取中文摘要

Human diseases associated with solid tumors are usually discovered by computed tomography (CT), magnetic resonance imaging (MRI), positron emission tomography (PET), single positron emission tomography (SPECT) and ultrasound (US). While the resulting medical images can characterize the location(s) and anatomical relationships of the tumors, they are often unable to specifically identify their nature and detect their full spatial extent. Targeted needle biopsy approaches (direct tissue sampling) are often required for histological, molecular, and genomic characterization leading to a precise diagnosis and treatment planning. However, targeting errors, poor quality samples and disease heterogeneity can cause inaccurate sampling often leading to non-diagnostic specimens. After a diagnosis is made, surgery (resection of cancerous tissue) remains the first-line therapy for many cancers, but its effectiveness is reduced when all pathological cells are not detected, often leading to disease recurrences. ***Diseased and normal tissue exhibit different morphological, molecular and biochemical characteristics that can be quantified based on optical properties derived from spectroscopic optical data associated with absorption (vascular information), elastic scattering (tissue microstructure), fluorescence (metabolites, proteins) and inelastic Raman scattering (lipids, proteins, DNA/RNA). This Natural Sciences and Engineering project aims to develop new optical spectroscopy instruments and biophysical/statistical models allowing highly sensitive and accurate molecular tissue characterization with the perspective of improving cancer tissue detection and classification for various organ sites including brain, breast, lung, mouth & throat, ovarian, prostate and skin cancer. Specifically, this project will lead to the development of: 1) a non-contact macroscopic wide-field optical imaging platform and biophysical models for quantitative tissue characterization based on biomarker maps associated with Raman, fluorescence and diffuse reflectance, to be validated using tissue acquired during tumor resection surgeries, 2) integrated multimodal fiber optics needles for real time in situ tissue characterization, and 3) microscopic imaging techniques to discover the underlying cellular and extracellular origin of the signals detected with the macroscopic tools, leading to detailed tissue characterization across all scales. This proposal will set up the stage for a new generation of cancer detection technologies. It should contribute to improving patient survival prospects and health care cost-effectiveness by reducing inefficiencies and downstream costs of the standard procedures through a decrease of the number of repeat procedures, which are expensive and time-consuming.