Towards rapid high-resolution mid-IR imaging for molecular spectral histopathological diagnosis of oesophageal cancers

摘要

Summary form only given. Modern western societies suffer from diseases of civilization which are mainly associated with smoking, fatty diets and obesity. One of those is the Gastroesophageal reflux disease (GERD) often also loosely called heartburn. GERD can in some cases lead to an abnormal change of cells (metaplasia) in the lining of the oesophagus. This syndrome is defined as Barrett's oesophagus (BO). If the oesophagus suffers from reflux of acid (heartburn) it can protect itself by replacing the normal stratified squamous epithelium by acid resistant columnar epithelium with goblet cells usually found lower in the gastrointestinal tract. These cells growing in an area where they usually do not belong, have a significantly higher risk to degenerate into cancers. These conditions are evaluated by performing endoscopy with harvest of biopsies. Biopsies are then cut and stained with haematoxylin and eosin and assessed by a pathologist. Patients diagnosed with BO undergo a regular surveillance in order to monitor if the metaplasia progresses into dysplasia or adenocarcinoma. The above described procedure is extremely time consuming and could in the future be replaced by hyperspectral imaging techniques utilising infrared (IR) radiation [1]. However, IR spectroscopic techniques are still not fast and sensitive enough to be used in daily clinical routines. A promising solution for this problem could be the use of more efficient detection systems and higher brightness IR light sources, such as quantum cascade lasers (QCL) or super continuum sources. Our work in the Marie Skłodowska-Curie Innovative Training Network Mid-TECH evaluates those light source technologies in combination with a new detection scheme called upconversion [2]. The signal in the IR is mixed with a visible or NIR wavelength laser to archive sum frequency generation and detect in the NIR or visible wavelength. This allows the use of silicon based CCDs which have a much better noise-equivalent power than conventional IR detectors while also being significantly cheaper.