||Project frame and team’s works
The development of in vivo reflectance imaging systems featuring wide field mapping of biological tissue optical properties is one of the current key challenges to provide clinicians with new extended diagnostic capabilities.
The CRAN research team has a strong experience in the fields of (i) autofluorescence and diffuse reflectance spectroscopy, biological tissues’ optical properties modelling, (iii) image registration and mosaicing and (iv) multimodal in vivo imaging applied to bladder (cystoscopy) and skin photodiagnosis. We have recently developed a unique multimodality reflectance imaging system [CP1] for skin pathology telediagnosis (benign, malignant lesions, chronic wounds, etc.) including technical innovations such as bimodal video sequence acquisitions (white light and 2 wavelengths light diffusion/absorption), hand-held and communicating facilities, and addressing scientific challenges for robust and precise image automatic mosaicing. In the framework of the regional collaborative project « InnovaTICs-Dépendance » (2014-2018) gathering two industrial partners (SEFAM and SD Innovation), the « Centre d’accueil et de Ressources pour Entreprises du Pays du Val de Lorraine » (CAREP) and the CRAN, this new device is under clinical trial for the diagnosis and care of cutaneous pathologies of dependent people in several long-term care nursing houses (EHPAD Etablissements d’Hébergement pour Personnes Agées Dépendantes) where highly resolved and extended field of view image mosaics are acquired by nurses and transferred to dermatologists located at the hospital for tele-expertise and medical instruction.
The “Instrumentation for Health” group led by Dr. Gioux at the ICube Laboratory in Strasbourg has extensive experience in designing and translating to the clinic novel imaging devices based on the first principles of light-tissue interactions. More particularly, the group has been historically involved in the design and translation of imaging systems for fluorescence-guided surgery and translated this technology to clinical practice on over 1000 patients wordwide. More recently, the group has been at the forefront of the development of novel technology that enables wide-field quantitative optical imaging of endogenous parameters in real-time. Contrary to fluorescence imaging that requires the injection of a contrast agent, and therefore increases risk and time in translating innovations to clinical practice, endogenous imaging relies solely onto the fundamental interactions between photons and tissues to extract information relative to its function (oxygenation, water content, and metabolism) and its structure (scattering). The “Instrumentation for Health” group has pioneered this approach by developing the only method to date allowing to make such measurements in real-time and within the constraints of surgery. Recently the group was awarded an ERC grant to further these development for colorectal surgery.
The PhD subject proposed aims at combining the experience and expertise of both CRAN and ICube teams to push forward their respective works. A co-supervision of the PhD student is therefore required between UL and Unistra.
Scientific/methodological issues – PhD subject
The scientific issues are related to finding diagnostic biomarkers from skin subsurface scattering and absorption information while taking into account the complexity of tissue and image features in terms of textures, skin hairiness, cutaneous phototype (implying different optical properties and reflectance), vascular activity (inflammation, wounding), surface forms according to the anatomic site (back, face, arm, leg, foot, etc.), and relief (wounds, lesions).
The originality of the problem we will tackle, in collaboration with ICube, lies in obtaining a non-invasive spatial characterization of the skin tissue condition by building a panoramic-field reflectance imaging method exploiting light absorption, scattering and polarization at macroscopic levels. Compared to current wide-field imaging approaches, panoramic-field (obtained through image mosaicking) can provide images of large areas of skin with higher resolution. Compared to other multispectral Diffuse Reflectance Imaging (DRI) systems under investigation, such as the teledermatology device developed by the CRAN and providing image maps of skin oxygenation levels based on main chromophores’ concentration estimation (HbO2, Hb and melanin), a higher specificity and sensitivity of a large number of skin conditions (from inflammation to cancer) should be extracted from combining additional information such as scattering and polarization.
More precisely, functional imaging provided by the measurement of chromophores (hemoglobin, water, lipids, melanin) provides a set of information that can be largely enriched by the combination with structural imaging that related to the cell composition and nuclear density at the cellular level and to the layered organization at an histological level through scattering and polarization. The potential of each modality (functional and structural) has been independently evaluated and validated by many investigators, including our respective groups, but the combination of these two parameters not exploited to increase the sensitivity and specificity in detecting diseases in living tissues.
The objective of the proposed work is to combine both modalities onto a single imaging method, to fabricate a device and to implement the proposed solution into an imaging system (hardware and software). This system will then be validated onto living tissues and the results correlated to standard-of-care histopathological analysis.
More specifically, the specific aims of the thesis are scheduled as follows:
- Development of instrumentation, metrological validation and tests on optical phantoms, multilayer modelling, study of the influence of a larger number of wavelengths and on the information combination from different modalities (absorption, scattering and polarization)
- Experimental validations on synthetic and hybrid (including ex vivo biological tissue) phantoms
- Finalization of a device to conduct clinical trials (integrating mosaicking algorithms developed elsewhere in the lab i.e. outside the present thesis)