Mehdi Ech-Chouini Email & Phone Number

Paris, Île-De-France, France

• Supported Biologists in Developmental Imaging Analysis: Provided key support to biologists by analyzing developmental time‑lapse movies of Drosophila cells, identifying biomechanical differences between animal groups, and contributing to the understanding of tissue morphogenesis.• Developed Advanced Computational Tools: Designed and implemented computational tools for segmentation correction and cell tracking, including a user‑friendly GUI to streamline the analysis of computer vision pipelines used in biological research.• Enhanced Cell Tracking with Deep Learning: Improved cell tracking accuracy by integrating temporal context into segmentation processes using advanced Deep Learning models, leading to more precise identification and tracking of cells over time.• Innovated in 3D Cell Segmentation: Led efforts in 3D cell segmentation, advancing the quantitative analysis of live‑imaging data and providing new insights into cellular behavior during tissue formation and development.

Toulouse, Occitanie, France

Topic: Feature engineering and Deep Learning on microscopic images for molecular classification of breast cancer• Designed, implemented, and compared Deep Learning and Machine Learning models: Improved breast cancer classification accuracy against the St. Gallen Consensus using a range of models, from simple classifiers to advanced deep learning approaches.• Conducted a state‑of‑the‑art review on feature extraction: Examined techniques from tumor nuclei and cells, guiding the development of more effective classification models.• Studied the explainability of classification methods (XAI): Analyzed methods like LIME and SHAP, enhancing transparency in breast cancer prediction models.• Developed a web application for domain experts: Enabled easy analysis and interpretation of AI model results, featuring interactive visualizations and automated report generation to aid clinical decision‑making.

Paris, Île-De-France, France

• Predicted asset CAPEX values to accurately quantify client exposure.• Created and implemented recommendation algorithms to match client portfolios with assets in the company's database, providing clients with a better understanding of their risks and exposure.• Engaged in continuous development of the Back application using Django.

Paris, Île-De-France, France

• Designed a tool for automatic evaluation of pathology detection algorithms, streamlining the model validation process and improving overall development workflow.• Implemented a time series forecasting algorithm to predict resource needs, such as drugs and beds. This project involved conducting a literature review of the best methods in the field, implementing the chosen method on company data, and comparing the performance of different methods.