Job offer

• Building a preclinical database using murine models (mice/rats) to study the impact of respiratory neuromuscular stimulation.
• Developing predictive models (supervised/unsupervised learning, deep learning) to identify responders/non-responders and optimize stimulation programs.
• Validating models through experimental protocols on animals, combining biology, signal processing, and data science.

The project is structured around two main axes:

1. Animal Experimentation: Acquisition of physiological data (EMG, respiration, SpO2, etc.) from murine models.
2. Predictive Modeling: Development of advanced machine learning methods to analyze this data and propose optimal stimulation programs.

Main Responsibilities:

1. Animal Experimentation (30 à 40% of time) * Design and implementation of protocols on murine models: o Measurements at D0, D2, D3, D15: weight, respiratory parameters (Vt, Fr, Ti, Te), EMG (diaphragm/intercostal/abdominal), SpO2, lean mass index, respiratory variability. o Compliance with good practices (ethics, regulations on animal experimentation).

• Data acquisition and preprocessing: o Cleaning, structuring, and annotating data to make it usable for machine learning. o Automation of data collection pipelines (Python scripts).

1. Predictive Modeling (60 à 70% of time) * Development of supervised and unsupervised learning models: o Classification (SVM, Random Forest, KNN) to predict response to stimulation. o Regression to estimate the effectiveness of stimulation programs.

• Application of deep learning methods: o Use of neural networks (CNN, possibly RNN/Transformers) to analyze biological signals (EMG, respiration). o Integration of multimodality (combining numerical data and raw signals).

• Statistical processing and biological data modeling: o Analysis of signal variability (ECG, EEG, EMG). o Selection of relevant features and model interpretability (SHAP, LIME).

• Model validation: o Performance evaluation (AUC-ROC, precision/recall, RMSE).

NeAR Team, Environment: 9 Principal Investigator researchers, 5 PhD students, 3 study engineers; a specific respiratory research group composed of 3 PhD students and 1 study engineer. Shared experimental areas, teamwork highly valued, weekly working meetings, and weekly lab and/or unit seminars.

PhD or equivalent

Research Field Biological sciences, Mandatory Skills in Neuroscience and Modeling

• Experience in developing supervised and unsupervised learning methods: o Mastery of classification algorithms (SVM, Random Forest, KNN) and regression. o Experience in feature selection and model validation.

• Knowledge of deep learning methods: o Experience with CNNs (for signal analysis) and possibly other architectures (RNN, Transformers). o Ability to adapt these methods to complex biological data.

• Knowledge of biological data in signal theory: o Experience with physiological signals (ECG, EEG, EMG): acquisition, preprocessing, feature extraction. o Understanding of specific challenges related to the analysis of these signals (noise, artifacts, inter-individual variability).

• Knowledge of statistical processing and biological data modeling: o Mastery of statistical tests (ANOVA, regression, non-parametric tests). o Experience in predictive modeling applied to biology or medicine.

Mandatory Skills in Animal Experimentation

• Proven experience in handling murine models (mice/rats): o Knowledge of neuromuscular stimulation protocols. o Mastery of physiological data acquisition techniques (EMG, respiration, SpO2).

• Compliance with ethical and regulatory standards related to animal experimentation.

Computer Development Skills

• Confirmed experience in Python: o Mastery of libraries: scikit-learn, tensorflow/pytorch, pandas, numpy, matplotlib/seaborn. o Ability to develop robust scripts for preprocessing, analysis, and modeling.

• Use of collaborative tools: o Git (version control), Jupyter Notebooks, development environments (VS Code, etc.). Additional comments