2022-04406 – PhD Position F/M Spatio-temporal modeling and AI tools for dose-response analysis and prediction of radiation-induced brain damage and cognitive impairment in patients treated with radiotherapy for glioblastoma.

Contract type : Fixed-term contract

Level of qualifications required : Graduate degree or equivalent

Fonction : PhD Position

Context

Employer :  Institut de Radioprotection et de Sûreté Nucléaire (IRSN)   

Location :    IRSN, Fontenay-Aux-Roses (92) – Ile de France – France

Benefits & working conditions : The position corresponds to a 3-year 100% PhD appointment with competitive salary and working conditions. It is co-funded in the framework of the ANR RADIO-AIDE (2022-2026). The PhD candidate will be physically based at the laboratory of epidemiology (LEPID at IRSN, Fontenay-Aux-Roses) with regular trips to INRIA Grenoble (Statify team).

Applications should be sent to : Sophie Ancelet – sophie.ancelet@irsn.fr

                                                                    Florence Forbes – florence.forbes@inria.fr

Assignment

Background :

Radiotherapy (RT) is one of the most important treatments of brain tumours. However, its potential toxicity on the central nervous system is a highly relevant clinical issue [Soussain et al., 2009]. It is also part of the priority research questions in radiation protection, regarding the identification and the prevention of non-cancer side effects related to the use of ionizing radiation (IR) for therapeutic purposes [UNSCEAR, 2010]. Currently, the most frequent and threatening mid to long-term neurotoxic complication of brain RT is cognitive dysfunction mainly related to radiation-induced leukoencephalopathy (RIL). The main modifications observed in Magnetic Reasonance (MR) brain images following RT are diffuse supratentorial white matter hyperintensities (WMH), radionecrosis and brain atrophy (BA), resulting in ventricular dilatation. The symptoms related to cognitive impairments induced by brain RT are quite various (ex: dysfunctions in working memory, psychomotor speed, executive function and attention) and can significantly alter the quality of life of long-term survivors.  Currently, the potential association between the spatio-temporal (ST) progression of radiation-induced brain injuries (as observed in MR brain images following RT) and the occurrence of cognitive impairment is poorly understood. This is partly due to complex interactions between tumour-related factors (or possible tumour recurrence) and treatment-induced neurotoxicity [Bompaire et al., 2018]. Most of previous studies did not consider jointly these two dimensions, despite their high level of correlation.

This thesis is part of the ANR research project RADIO-AIDE (RADIation-induced neurOtoxicity assessed by ST modeling and Artificial Intelligence after brain raDiothErapy) which will start in spring 2022 for 4 years (2022-2026). Led by IRSN (Coordination : Sophie Ancelet, thesis supervisor), this data science project brings together 7 partners (INRIA Statify team led by Florence Forbes (thesis director), INSERM-GIN, Centre Giovanni Borelli, ICM, AP-HP Pitié-Salpêtrière, Institut de Cancérologie Strasbourg Europe (ICANS, Strasbourg) and the start-up company Pixyl). RADIO-AIDE aims at developing spatio-temporal (ST) models and Artificial Intelligence (AI) tools to better understand the neurotoxic mechanisms underlying the initial localization and ST evolution of brain tissue modifications induced by RT in patients treated with RT for high-grade glioma (glioblastoma thereafter). The developed tools will allow to study the potential links between these MRI lesions and the occurrence of cognitive disorders, to individually predict the cognitive side-effects at early stage after brain RT and to provide to clinicians a usable academic software to perform an automatic processing of patient’s MR brain images following brain RT, based on relevant image-based biomarkers of both RT-induced neurotoxicity and potential tumor progression.

Main activities

This thesis aims at developing probabilistic ST models and AI tools to: a) estimate the potential associations between radiation-induced brain tissue modifications (as observed in MR brain images following RT), the dose of IR absorbed to the brain at a fine spatial scale (e.g. voxels of MR brain images) and the occurrence of cognitive impairment in patients treated with RT for glioblastoma ; b) predict the individual occurrence of cognitive impairment in these patients at an early stage after RT, with a quantification of the associated prediction uncertainties The work will be based mainly on data from the EpiBrainRad cohort [Durand et al. (2015)], which is one of the largest international prospective cohorts of patients treated with chemo-radiotherapy for glioblastoma and who have undergone standardised neuropsychological assessments. This cohort currently includes 224 adults treated for glioblastoma at the Hôpital de la Pitié-Salpêtrière (Paris) or at ICANS (Strasbourg). Since 2015, data are collected prospectively during the clinical follow-up of patients. For each patient, they include: a) clinical characteristics; b) multimodal MR brain images (collected before surgery, before RT and then every 2-3 months after RT), c) CT images used in the planning of brain RT treatment; d) results of an extensive battery of cognitive tests performed before RT and at 12, 24 and 36 months after RT.

Skills

Applicants must have :

• validated a MSc’s degree (or an equivalent qualification in terms of credits)
• a background in statistics/machine learning/data science
• proficiency in modern high-level programming languages for data analysis (Python, R), probabilistic modelling and Bayesian statistics 
• an interest in applications in epidemiology/public health.
• Good writing and proficiency in oral and written English are also required.   

Benefits package

 Benefits and working conditions :

The position corresponds to a 3-year 100% PhD appointment with competitive salary and working conditions.

It is co-funded in the framework of the ANR RADIO-AIDE (2022-2026). The PhD candidate will be physically based at the laboratory of epidemiology (LEPID at IRSN, Fontenay-Aux-Roses) with regular trips to INRIA Grenoble (Statify team).

General Information

• Theme/Domain : Computational Neuroscience and Medicine Statistics (Big data) (BAP E)
• Town/city : Paris
• Inria Center : CRI Grenoble – Rhône-Alpes
• Starting date : 2022-10-03
• Duration of contract : 3 years
• Deadline to apply : 2022-09-15

Contacts

• Inria Team : STATIFY
• PhD Supervisor : Forbes Florence / forbes@inria.fr

The keys to success

The successful candidate is expected to be willing to work collaboratively in an interdisciplinary team of mathematicians, data scientists, neuroimaging experts, clinicians, and epidemiologists. He or she is also expected to be independent, highly motivated and able to communicate in scientific terms.

About Inria

Inria is the French national research institute dedicated to digital science and technology. It employs 2,600 people. Its 200 agile project teams, generally run jointly with academic partners, include more than 3,500 scientists and engineers working to meet the challenges of digital technology, often at the interface with other disciplines. The Institute also employs numerous talents in over forty different professions. 900 research support staff contribute to the preparation and development of scientific and entrepreneurial projects that have a worldwide impact.

Instruction to apply

Defence Security :
This position is likely to be situated in a restricted area (ZRR), as defined in Decree No. 2011-1425 relating to the protection of national scientific and technical potential (PPST).Authorisation to enter an area is granted by the director of the unit, following a favourable Ministerial decision, as defined in the decree of 3 July 2012 relating to the PPST. An unfavourable Ministerial decision in respect of a position situated in a ZRR would result in the cancellation of the appointment.

Recruitment Policy :
As part of its diversity policy, all Inria positions are accessible to people with disabilities.