Artificial Intelligence for chromosomal aberrations scoring in dosimetry – INCREASED

Exposure to ionizing radiation can have serious consequences for the health of exposed people and potentially impact many victims. There is now well-established scientific knowledge of the health effects induced by different dose levels and their associated kinetics of occurrence. Thus, once the first triage has been carried out to identify irradiated victims on the basis of clinical symptoms, it is necessary to refine the evaluation of the dose received in order to make a precise diagnosis / prognosis for the remaining asymptomatic victims. During insidious scenarios, When the assessement of exposure condition is difficult or impossible, the quantification, on samples taken from the victims, of radio-induced chemical or biological effects is more suitable for an individualized dose reconstruction compared to theoretical calculation or Monte-Carlo simulations in regards to theirs great uncertainties.
The standard and reference technique for biological dosimetry is the quantification of chromosomal aberrations in circulating lymphocytes. Its robustness and sensitivity range (from 150 mGy to 5 Gy) make it highly suitable both for the short-term diagnosis / prognosis phase of asymptomatic victims as well as for their long-term follow-up. However, the automated solutions proposed for the aberration recognition suffer up to day from lack of precision, especially in low doses due to the failure to eliminate false positive.
The INCREASED project aims to adapt the most powerful algorithms of modern artificial intelligence to the context of automatic detection of chromosomal aberrations in dosimetry. This project proposes to revisit the semi-automatic or automatic detection methods currently used for the recognition of dicentrics in GIEMSA imagery in the light of the most recent advances in artificial intelligence and deep learning. These modern methods, which demonstrated their indisputable superiority in other areas of computer vision, will be deployed on GIEMSA images for an exhaustive multi-class count not only of dicentric chromosomes but also of ring-centric aberrations, acentric fragments , and even tricentrics. Furthermore, the INCREASED initiative is also quite innovative in dosimetry based on FISH imaging. To date, this type of dosimetric reconstruction is based on manual counting of non-exhaustive and very simplified annotations of the different possible forms of translocations. This protocol is therefore as heavy as imprecise. The INCREASED project offers two major improvements in this context: firstly in terms of exhaustive annotations of the different forms of observable FISH translocations (3 colors) through an universal and rigorous scoring. Secondly, the use of the most modern artificial intelligence algorithms able to detect and interpret co-locations/co-neighborhoods as different kind of translocations.
The project shows a very strong dual character due to its applications targeting soldiers in operations, civilians in an accident context (NRBC-E, medical… etc) as well as the monitoring follow-up of military personnel, medical or industrial potentially exposed during their activity. Beyond the dosimetry framework, the results and methodologies developed during the INCREASED project could be applied to many fields based on massive cytogenetic images analyses.