AIpolypNET: A National Network for the Development and Validation of AI Support Systems for the Detection and Diagnosis of Colorectal Cancer

AIpolypNET: A National Network for the Development and Validation of AI Support Systems for the Detection and Diagnosis of Colorectal Cancer

  • The AIpolypNET network aims to develop and validate artificial intelligence systems designed to improve colorectal cancer diagnosis.
  • Coordinated by the CVC and led by Dr. Jorge Bernal, the network consists of 8 institutions—4 with a clinical focus and 4 with a technical focus.
  • AIpolypNET is funded by the State Research Agency, part of the Ministry of Science and Innovation.

The national AIpolypNET network, dedicated to improving the diagnosis of colorectal cancer through Artificial Intelligence (AI), held its first working meeting in Ourense, Galicia. Comprising 8 institutions, including 4 clinical and 4 technical profiles, this network represents a collective effort to address one of the most prevalent diseases globally.

Colorectal cancer is one of the most common forms of cancer, ranking as the second most frequent in women and the third in men in Europe. Globally, approximately 1.93 million new cases are diagnosed every year, with mortality reaching 916,000 patients in 2020. Despite its high incidence, over 90% of cases can be cured if the precancerous polyp, the precursor lesion, is detected and treated in time.

Although various detection techniques exist, colonoscopy is the most effective method as it allows for the detection and treatment of the lesion in a single intervention. However, this technique has limitations, and about 22% of lesions are not detected during the examination, leading to a missed opportunity for effective prevention and treatment.

In recent years, advancements in machine learning and computer vision have driven the development of AI-based methods to support clinical staff in detection and diagnosis. However, the scarcity of high-quality annotated data and the limited ability to predict histology have been significant challenges in this field.

In this context, AIpolypNET emerges as a national network dedicated to the development and validation of intelligent systems for the detection and diagnosis of colorectal cancer. This initiative seeks to establish common protocols for image acquisition and annotation, as well as uniform validation systems, to ensure the good performance of the developed methods. Additionally, it aims to define a set of best practices for using these systems in the examination room, including the selection of necessary equipment and compliance with current regulations.

The collaboration among the 8 participating research groups, spread across Catalonia, Extremadura, Galicia, and the Basque Country, promises to create synergies and improve the outcomes of individual initiatives, optimizing existing research resources. The creation of a comprehensive joint database with high-quality images representing different histological categories and the integration of various methodologies are key steps toward advancing the development of computational methods for efficient and effective detection and diagnosis that can assist clinical staff in real time in the examination room.

The AIpolypNET network is funded by the State Research Agency of the Ministry of Science and Innovation and coordinated by the CVC, led by Dr. Jorge Bernal. In addition to the CVC, the network includes the CVC, the Gastrointestinal and Pancreatic Oncology Research Group at Hospital Clínic de Barcelona, the Jesús Usón Minimally Invasive Surgery Centre Foundation (CCMIJU), the Digestive System Service at the University Hospital of Cáceres, the New Generation Computer Systems Group (SING) at the University of Vigo, the Digestive Oncology Group of Ourense (GIODO) from the Galician Health Service, eVIDA Research Group from the University of Deusto, and Osakidetza.

The AIpolypNET network held its inaugural working meeting on Tuesday, April 16th, at the Higher School of Computer Engineering in Ourense. This session brought together various network activities and provided a forum to identify clinical challenges that can be tackled using AI. Discussions also covered future steps to enhance the network's objectives.

AI for Colorectal Cancer Detection at CVC

CVC has been dedicated to the research of AI systems for the detection and diagnosis of colorectal cancer for over 15 years. The Image Sequence Evaluation lab (ISE lab) is actively involved in six research projects related to endoscopic image processing on both regional and national levels, including projects such as ALETEHIA, iVENDIS, and APODEC. These research efforts have yielded significant outputs, including 25 articles published in Q1-Q2 indexed journals, participation in over 20 international conferences, and the organization of 5 challenges at major conferences. Furthermore, the lab has developed three distinct databases—CVC-ColonDB, CVC-ClinicDB, and CVC-VideoClinicDB— which have been downloaded by more than 300 research groups worldwide.

AI4polypNET is a RED2022-134964-T grant funded by MCIN/AEI/10.13039/501100011033