Skip to main content

Multimodal highly-sensitive PhotonICs endoscope for improved in-vivo COLOn Cancer diagnosis and clinical decision support

Periodic Reporting for period 2 - PICCOLO (Multimodal highly-sensitive PhotonICs endoscope for improved in-vivo COLOn Cancer diagnosis and clinical decision support)

Reporting period: 2018-07-01 to 2020-09-30

Colorectal cancer ranks as one of the predominant cancers, affecting approximately 1:10 people during their life and causing almost 880,792 annual deaths globally in 2018. Specifically, adenocarcinoma represents more than 95% of all these cases. It is estimated that 20-40% of patients undergoing colonoscopy present polyps. Of these detected polyps 29-42% are hyperplastic, whereas the rest are neoplastic. Clinicians demand new diagnostic technologies to assist detection and discrimination of hyperplastic and neoplastic polyps and reduce recurrence rates.
The main objective of the project is the development of a fully-functional wide-field fluorescence based, OCT and MPT photonics endoscope for improving colorectal cancer diagnosis providing in-vivo image-guided biopsy capabilities and higher sensitivity and specificity than current diagnostic methods. The development of this fully functional photonics-based endoscope will:
-Obtain unprecedented sensitivity and specificity (>99%, >95%) on the optical analysis of neoplastic colorectal lesions.
-Reduce by 40% the needs of biopsying of the polyps detected on a colonoscopy.
-Reduce the number of missed polyps by 30% including the most challenging flat lesions.
-Reduce re-interventions by 90% by assessing lesion infiltration and resection assessment in situ.
-Reduce the intervention costs by 30% per patient outcome.
-Facilitate and augment the presence of EU Photonics SMEs and companies into the demanding biomedical diagnosis market via the exploitation of the PICCOLO endoscope and its components.
PICCOLO consortium has worked through all the project pursuing these objectives. On the hardware side, enhanced wide-field imaging and probe integrating OCT and MPT technologies have been delivered. On the software side, a computer aided diagnosis (CAD) system that makes use of advanced deep learning models for automatic image diagnosis and virtually stains MPT images into gold-standard Haematoxylin-Eosin (H&E) images with high sensitivity and specificity has been delivered. Prototype and algorithms have been fully tested on murine models and validation on human models has been partially achieved.
Medical and commercial needs (WP1) have been identified. Requirements and high-level design have been cyclically revisited and updated considering unforeseen risks. The first version of the prototype (WP2 &WP4) was delivered in November 2018. The probe has been designed to be used through the working channel of a colonoscope, so it can be sold separately as an additional advanced component. Further improvements of the probe have been integrated and, in parallel, a new (confidential) technology that provides enhanced wide-field imaging has been developed and successfully demonstrated within the project. Generation of new open knowledge has been fully achieved thanks to database management and algorithm implementation (WP3). Various databases have been collected, organized and annotated in the scope of this work package, three of them made publicly available. A Wide-field Polyp Segmentation algorithm approach, which automatically detects lesions on colonoscopy videos, has been proposed and delivered. Optical biopsy capabilities have been achieved combining various algorithms developments. Advanced strategies based on deep learning able to deal with few samples (a common problem in machine learning), named few-shot learning, have been proposed. Apart from the OCT and MPM classification algorithms and innovative solution that virtually stains MPM images into H&E images has been generated and patent application filled. This innovation aims at facilitating the adoption of advanced imaging techniques by clinicians and ease the learning curve, as it is able to translate unknown images to a known knowledge space by clinicians, as it is the gold-standard H&E.
Fully validation with animal models (WP5) has been achieved through laboratory tests, validation trials, safety tests, etc. for the different prototype hardware components and CAD software algorithms. A database of OCT images using a commercial device has been generated with murine samples. Same specimens have also been used for the development of Colorectal Cancer Molecular Biomarkers based on microRNA analysis. Part of the specimens were healthy animals were hyperplasia has been induced with a model developed and validated in this work package.
Various validation actions and generation of new knowledge have been achieved with human samples (WP6). Colonoscopy videos recorded have been annotated by project clinicians, generating as a result an openly available database. Additionally, another openly published database of multi-photon microscopy (MPM) images have been acquired previous selection of samples. Clinical validation of wide-field polyp segmentation algorithm and optical biopsy algorithms have been achieved.
Ethical implications (WP7) of animal and human models’ samples acquisition and validation have been closely monitored during the whole project, data management plan defined, and ethical committee approvals obtained.
With respect to exploitation and dissemination (WP8), project achievements, publications, events, results have been shared through the project website and twitter on regular basis. A confidential business plan has been elaborated for an endoscope integrating the Photonics and Optical imaging technologies with the CAD Software for colon lesion diagnosis support. Various business models are considered upon commercialization and detailed exploitation plans are defined. Action plans for hardware components and plans to use for software components have been fully defined.
As a result of all this work, different individual results have been made publicly available and presented in a catalogue.
Considering project objectives, two important achievements are remarkable. On the one side, that the delivered probe fits the working channel of a colonoscope and it’s no exclusive of a specific device. The colonoscopy industry can benefit from this design solution and offer the optical biopsy probe as an additional component that can be sold separately from the colonoscope. Additional value can be offered if CAD system algorithms are added to the probe purchase. This benefits the Health systems, clinicians and patients, as it would make possible to update existing hospital colonoscopes with least monetary investment. Besides, optical biopsy adoption would lead health systems to reduce diagnosis costs, time and patient trauma.
On the other side, the CAD system algorithms. Advanced image processing methods based on deep learning have been implemented on the algorithms. With this introduction, the market forecast for clinical software is expected to rapidly growth in the short term. In this sense, the resulting metrics of the different algorithms developed, suggests that the realization of the optical biopsy is closer than ever. Besides this, during the PICCOLO project, it has been demonstrated that the future of clinical devices relies on the combination of hardware and software. The virtual staining algorithm approach proposed, which automatically transforms MPT images into gold-standard H&E images, can facilitate the adoption of new imaging techniques by clinicians and at the same time easy the dreaded learning curve.
Patients Leaflet, Side A
summary of public results
Clinicians leaflet_Page2
Project Leaflet, Side A
Patients Leaflet, Side B
Clinicians leaflet_Page1
Project Leaflet, Side B