Instrument design of FFOCT based endoscope using high-resolution imaging for non-invasive biopsy assessment

Adimec together with five partners, as part of the CAReIOCA consortium are developing & validating a high-resolution high-speed imaging based medical device to perform non-invasive optical biopsy for cancer assessment. Adimec contributes with a specific high-performance CMOS CoaXPress camera.

The goal of CAReIOCA is to provide pathologists and/or surgeons non-invasive optical imaging at the cellular level within the human body in real time. The optical technology used in the end system is based on Full Field Optical Coherence Tomography (FFOCT), a technique that enables volumetric image capture on semi-transparent tissue at micron resolution in 3D. The technology intends to assist in diagnosis, particularly of cancer, in skin, breast, prostate, brain etc, as well as quality control of biopsies.

The project was started at the beginning of 2013 and key achievements have started to be delivered in the first half of 2014, including:

The release of high-speed, high dynamic range CMOS sensor prototypes, ready for first camera integration
Recent improvements in the instrumental design of FFOCT endoscopic devices
Advances in building an atlas of clinical data, as a basis for the definition of FFOCT-based reading and diagnosis criteria

Release of high-speed, high dynamic range CMOS sensor prototypes

At the beginning of March 2014, assembled image sensor prototypes designed by CMOSIS and manufactured by an external CMOS foundry were delivered.

Adimec has fully defined the camera design, in terms of electronic circuits, firmware, thermal performance and casing. Adimec will embed the sensor in a camera platform with adequate high-speed precision imaging and interfacing in order to plug and use it easily in LLTech FFOCT devices.

The integration of the camera to LLTech will be realized this summer.

Improvements in the instrumental design of FFOCT endoscopic devices

The singular principle of FFOCT endoscopic imaging is incompatible with an implementation on classical endoscopes. For example, a significant difference with standard endoscopes is that the light transmitted to the sample and the signal coming back from the sample follow the same path in the system, which brings parasitic light issues. Thus, the endoscopic system, composed of a light source, an optical probe and a detection unit, has to be built from scratch.

The major stake of endoscopic systems is to optimize the energy transfer and limit the losses, which are inevitable when long distances are traveled through small diameter optics, not to mention the source-probe coupling efficiency. This is all the more important as the endoscope is dedicated to high-speed imaging. Another challenge is the acquisition and transfer of high-quality images, which requires specific and restrictive properties of the optical probe.

A layout for the high-resolution FFOCT endoscopic system has been completed.

Building an atlas of clinical data

With the final goal of assessing the FFOCT technique on cancer imaging, the clinical study is divided in 3 steps: 

Identify non-pathological and pathological features on the FFOCT images of tissues;
Train the pathologists in reading FFOCT images;
Evaluate the correlation with conventional histopathology and deduce sensitivity and specificity values.

This study has started with LLTech commercial microscope so as to maximize the data volume in the time limit of the project, define a first set of diagnosis criteria, and ease the clinical usability of future devices such as the endoscope. The clinical partners in this project are Leiden University Medical Center (LUMC) in Leiden, the Netherlands and Gustave Roussy Institute (IGR) in Villejuif (Paris), France. LUMC perform studies on breast cancer while IGR is focused on studying head & neck cancer.

The image below is an example of a neck tissue sample (epiglottis) imaged in depth using FFOCT (Light-CT scanner – LLTech), before histology slide preparation. Raw biopsies have been imaged in depth without the need for tissue slicing or staining. Many images of head and neck tissue were collected to demonstrate the very good correlation between FFOCT and histology in terms of morphology. The image below also illustrates the unique resolution capability of FFOCT: individual cells or nuclei are clearly visible when zooming on specific areas without the need for additional contrast agents.

These results have been obtained as a preliminary ex-vivo phase of the CAReIOCA project, which will deliver a novel in-vivo FFOCT endoscope, for intraoperative diagnosis of tumor margins, in particular in the field of Head & Neck Cancer. At the moment, a total of 80 biopsies have been imaged. Much more are planned for the future, to build atlases as complete as possible given the diversity of tissue anatomical structures.

This project has received funding from the European Union Seventh Framework Program FP7-ICT-2011-8 under grant agreement number 318729.

To get more details on the CAReIOCA project, click here. http://careioca.eu/index.php?option=com_content&view=article&id=47:careioca-second-newsletter&catid=10:news&Itemid=4