CE19 - Technologies pour la santé

Quantitative Optical Imaging for Liver Surgery – LiverSURG

Quantitative Optical Imaging for Liver Surgery

Despite significant progress in medical imaging technologies, there currently exist no tools capable of objectively helping healthcare professionals during liver transplantation surgeries. Instead, surgeons still rely on their own senses to determine whether a transplant is healthy either before, during or after the procedure. In this project we propose to develop innovative solutions to allow quantitative multispectral optical imaging in real-time during liver transplant surgery.

General objective of the project

Our general objective is to provide objective surgical guidance during liver transplant surgery. <br />The hypothesis underlying our study is that Near-Infrared (NIR) light travels relatively deeply into tissues and is capable of providing critical information during surgery. In particular, oxy- and deoxy-hemoglobin, water and lipids can provide functional information, while scattering can provide micro-structural information. We recently developed a novel method called Single Snapshot of Optical Properties (SSOP) that relies on the analysis of the tissue response in the Spatial Frequency Domain to extract its optical properties (absorption and scattering). Because SSOP works entirely in the frequency domain, it is today the first and only method amenable to provide video-rate quantitative images during surgery.<br />In this project we propose to develop innovative solutions to allow quantitative multispectral optical imaging in real-time during liver transplant surgery. To solve this challenging problem we have assembled a multidisciplinary team of scientists, engineers and surgeons. Pr Vibert from the Paul Brousse Hospital (HPB) in Paris is an expert in liver transplantation procedures. Dr. Diana from the University Hospital Institute (IHU) in Strasbourg is an expert experimental surgeon specialized in innovative surgical guidance techniques. Finally, the coordinator has invented SSOP, and our group and others at the ICube laboratory have a proven expertise in advanced photonics methods and the design of surgical guidance systems.

In order to achieve our long-term goal of providing objective surgical guidance during liver transplant surgery, we propose the following objectives:

Objective 1: To develop and validate novel acquisition and processing methods to perform multispectral SSOP imaging in real-time: Novel acquisition and processing methods must be developed to perform multi-spectral quantitative optical imaging in real-time. We will investigate an innovative solution that consists of modulating light both in space and time to perform multispectral SSOP at video rates (> 25 frames per second), to develop real-time processing and visualization methods, and then implement this solution either on FPGA, GPU or on a hybrid solution based on both components for real-time display.

Objective 2: To develop and validate a preclinical imaging system capable of performing real-time multispectral SSOP: We propose to develop a dedicated imaging system that integrates the technological development proposed in Objective 1. The system will be composed of a cart containing all power and electronics, an imaging head for performing real-time multispectral SSOP acquisitions and displays for real-time visualization during surgery. The imaging system will finally be validated onto tissue-mimicking phantoms, small animals (rodents) at ICube and large animals (pigs) at the IHU of Strasbourg.

Objective 3: To investigate the potential of this technology during preclinical liver surgery in large white pigs: In order to validate our hypothesis, we will study the relationship between tissue viability using the device and method developed in Objectives 1&2 in large white pigs, the closest animal model to human for liver surgery. Optical measurements will be compared to gold standard histological analysis. Finally, we will prepare all necessary documentation to translate our findings into the clinic during a first-in-human pilot trial at the Paul Brousse Hospital with Pr Vibert’s surgical team.

Objective 1: A new method for imaging large fields of view, quantitatively and in real time has been developed and tested. This method has the unique capability of offering exceptional image quality that has never been seen before, on images of 1 megapixel and with a processing time of 18 ms. To achieve these performances, we have developed a data acquisition and processing method which requires only one input image and allows to calculate the optical properties of the tissues (reduced scattering and absorption) while taking into account the correction of the profile of the sample. The associated processing was pushed to its maximum by the use of deep learning processing (deep learning / deep neural network) in combination with a low level GPU implementation allowing an extremely fast processing time (the fastest in existence today).

Objective 2: Design of a custom optical imaging head and of a dedicated light sources for the project. The optical head allows to combine 3 cameras in parallel, one color for the anatomical information necessary for surgery and 2 monochrome cameras which, associated with the imaging method mentioned above, allows imaging at several wavelengths (i.e. multispectral) the surgical field to extract vascular data (oxygenation and perfusion). A dedicated powerful light source has been designed from LASER diodes while complying with LASER safety standards (the system complies with class 3R).

Development of the most advanced real-time quantitative optical imaging method ever described.

Peer-review articles
1. Real-time, wide-field and high-quality single snapshot imaging of optical properties with profile correction using deep learning (in review)
2. Hyperspectral enhanced reality (HYPER) for anatomical liver resection Surgical Endoscopy Apr 2020

International conferences
1. Real-time processing and visualization of functional parameters
in living tissue with 3D profile correction– Photonics West 2020
2. Real-time, quantitative and wide-field oxygenation imaging
platform for surgery – Photonics West 2020
3. Real-time quantitative diffuse optical imaging for surgical
Guidance (invited) – Photonics West 2020
4. Real-time processing and visualization of functional and structural conditions of living tissue (accepted – COVID19) Photonics Europe 2020
5. Wide-field quantitative fluorescence imaging for safer oncologic surgery (accepted – COVID19) Photonics Europe 2020
6. Design of rigid endoscopic system for quantitative wide-field optical imaging of tissue optical properties (accepted – COVID19) Photonics Europe 2020
7. Hyperspectral imaging of liver ischemia in a porcine model: a proof of concept (iSMIT 2019, Heilbronn Germany)
8. Evaluation of a pocket-sized spectrometer for real-time assessment of liver graft macrosteatosis - A pilot study. European Association for the Study of the Liver (EASL), International Liver Congress, Paris 11-15 Avril 2018
9. Evaluation of a pocket-sized spectrometer for real-time assessment of liver graft macrosteatosis - A pilot study. International Liver Transplantation Society (ILTS), Lisbonne – Portugal, 23-26 Mai 2018

National conferences
1. Intérêt de la spectrométrie de poche pour l’évaluation extemporanée de la macrostéatose des greffons hépatiques. 13ème congrès francophone de chirurgie digestive et hépato-bilio-pancréatique (ACHBT-SFCD), Paris, 30 Novembre 2017

Despite significant progress in medical imaging technologies, there currently exist no tools capable of objectively helping healthcare professionals during liver transplantation surgeries. Instead, surgeons still rely on their own senses (vision and touch, primarily) to determine whether a transplant is healthy either before, during or after the procedure. In turn, surgery remains subjective and dependent on the experience of the surgeon, resulting in unacceptable failure, recurrence and morbidity rates, as well as in significant quality of care disparities across hospitals.

As of January 1st, 2018, 1437 patients are on the wait list for liver transplantation, a number that is continuously increasing, with only 574 patients a decade ago. Every year, 10 to 12% of these patients will not survive the wait-time for getting a transplant. By large, these numbers are the result of the lack of objective decision criteria to determine whether a donor liver is healthy enough to be transplanted. In addition, following liver surgery, 7 to 10% of the patients will suffer liver function deficiencies. While the reasons behind these deficiencies are well known (combination of defects in micro-circulation, venous congestion, arterial thrombosis, asynchronies in hepatocyte regeneration leading to physiological disorganization), no tool currently exists to detect these deficiencies early during surgery and intervene in a timely manner. In turn, these deficiencies lead to costly emergency re-operations, and a survival rate decrease as important as 15%.

Because these failures and complications can be mainly related to the lack of information regarding the liver, they could be avoided if tools were available to assist surgeons in visualizing the viability of the liver tissue both before and after transplantation allowing to select viable donors, as well as proper & timely intervention. Instead, surgeons subjectively rely on their own visual senses to assess the quality of the procedure, leading to an unacceptable morbidity and mortality rates.

The hypothesis underlying our study is that Near-Infrared (NIR) light travels relatively deeply into tissues and is capable of providing critical information during surgery. In particular, oxy- and deoxy-hemoglobin, water and lipids can provide functional information, while scattering can provide micro-structural information. We recently developed a novel method called Single Snapshot of Optical Properties (SSOP) that relies on the analysis of the tissue response in the Spatial Frequency Domain to extract its optical properties (absorption and scattering). Because SSOP works entirely in the frequency domain, it is today the first and only method amenable to provide video-rate quantitative images during surgery.

In this project we propose to develop innovative solutions to allow quantitative multispectral optical imaging in real-time during liver transplant surgery. To solve this challenging problem we have assembled a multidisciplinary team of scientists, engineers and surgeons. Pr Vibert from the Paul Brousse Hospital (HPB) in Paris is an expert in liver transplantation procedures. Dr. Diana from the University Hospital Institute (IHU) in Strasbourg is an expert experimental surgeon specialized in innovative surgical guidance techniques. Finally, the coordinator has invented SSOP, and our group and others at the ICube laboratory have a proven expertise in advanced photonics methods and the design of surgical guidance systems.

Project coordination

Michel de Mathelin (Laboratoire des sciences de l'Ingénieur, de l'Informatique et de l'Imagerie (UMR 7357))

The author of this summary is the project coordinator, who is responsible for the content of this summary. The ANR declines any responsibility as for its contents.

Partner

ICube Laboratoire des sciences de l'Ingénieur, de l'Informatique et de l'Imagerie (UMR 7357)
IHU FCS IHU MIXSURG
HPB Physiopathologie et traitement des maladies du foie

Help of the ANR 490,591 euros
Beginning and duration of the scientific project: December 2018 - 48 Months

Useful links

Explorez notre base de projets financés

 

 

ANR makes available its datasets on funded projects, click here to find more.

Sign up for the latest news:
Subscribe to our newsletter