法国国家健康与医学研究院（INSERM）2022年招聘博士后（医疗算法用身份证）

Post-Doctoral Researcher - Identity Card For Algorithms In Healthcare

Universities And Institutes Of France

Description

Organisation/Company: METHODS team of the Centre of Research in Epidemiology and StatisticS (CRESS - UMR1153 Université Paris Cité / INSERM)

Research Field: Medical sciences › Medicine

Researcher Profile: First Stage Researcher (R1)

Application Deadline: 08/09/2022 21:00 - Europe/Brussels

Location: France › Paris

Type Of Contract: Temporary

Job Status: Full-time

Hours Per Week: 35

The METHODS team of the Centre de RechercheEpidémiologie et Statistiques (CRESS - UMR1153 Université Paris Cité / INSERM) is looking for a group of four post-doctoral fellows to work on the @Hotel-Dieu project (Scientific director: Pr. Philippe Ravaud).

Candidates should have a doctoral degree in clinical epidemiology, statistics, methods research or meta research.

@Hotel-Dieu is an ambitious public-private partnership aimed at developing an interface between public organizations (Assistance Publique - Hôpitaux de Paris and Université Paris Cité) and technology startups (Lifen, Withings, Nouveal, Nabla and Implicity). Its objective is to accelerate the uptake of digital medicine tools, such as biometric wearable devices and AI, in both hospitals and outpatient care. The program received 17 million Euros from the French Public Investment Bank (“BanquePubliqued'Investissement” - BPI).

Within the @Hotel-Dieu project, the METHODS team is responsible of an important research program aimed at:

Improving the development of algorithms in healthcare, especially to ensure their robustness and external validity;

Conceptualizing a new care model blending digital technologies in the current workflow of physicians and patients, through the use of participative sciences;

Developing methodological and mathematical tools for real-time personalized care and “Just-in Time Adaptive Interventions”;

Refining the methods to accelerate and lower the cost of the evaluation of the effectiveness of digital health interventions

The recruited post-doctoral fellows will be part of the METHODS Team of CRESS, located at Hôtel-Dieu hospital in the center of Paris. The team is affiliated to the Université Paris Cité and Inserm (French national health research institute). It offers a dynamic multidisciplinary research environment with four tenured university hospital professors, two tenured INSERM research, one junior professor, two assistant professors, 5 post-docs and about 10 doctoral students.

Mission:

The following section is specific to a position aimed at designing an identity card for algorithms in healthcare. Algorithms based on artificial intelligence (AI) techniques are about to play a decisive role in the production of medical solutions, whether for monitoring, prevention, diagnosis or therapy. Yet, there are currently no consensus recommendations for evaluating the performance of AI tools. Numerous pitfalls can undermine the reliability of the claimed performance of AI-based algorithms: the validation methods of statistical models, their robustness, the metrics chosen for performance measurement, the composition of datasets, as well as problems of clinical methodology, which can undermine the transparency, quality or generalizability of results.

The objective of the project is to leverage on an existing scoping review of the key elements to be reported to understand the performance of an algorithm to set-up and conduct a large international Delphi study involving regulators, AI specialists and end-users (researchers / clinicians) to develop the identity card.

For this project, we are looking for a highly motivated candidate with a doctoral degree in clinical epidemiology, statistics, methods research or meta research.

The position is open and to be filled as soon as a suitable candidate will be identified.

Activities:

Identification of the elements that should be included in the identity card of algorithms for healthcare (literature search, interviews with stakeholders, etc.)

Set-up and conduct of a large international Delphi survey with stakeholders (regulators, developers of algorithms, end users, etc.)

Organization of a consensus meeting to finalize the identity card

Designing the identity card (organization of the items, graphical presentation, implementation in an online tool)

Pilot testing the identity card with algorithms provided by partners from the @Hotel-Dieu program

Skills:

Knowledge in the development and validation of AI models

Conduct of systematic reviews, knowledge in reviews of prediction models / AI models

Conduct of Delphi surveys

Organizational skills

Education level: Doctoral degree in Clinical Epidemiology statistics, methods research or meta research.

Duration: 24 months (depending on the starting date)

Salary: 2 400 to 2 700