Loïc Le Bescond

Loïc Le Bescond is a research scientist at Institut Curie specializing in the integration of artificial intelligence with oncology and digital pathology. His work bridges computational methods and clinical applications, particularly in breast cancer diagnostics and treatment optimization through advanced machine learning techniques.

His primary research domains include Machine Learning, Digital Pathology, and Computational Oncology, with a focus on developing algorithms for nuclei segmentation, out-of-distribution detection in medical imaging, and biomarker analysis for targeted cancer therapies. Recent projects involve leveraging sparse convolutions for efficient whole slide image classification and creating deep learning models for fine-needle aspiration cytology analysis.

Le Bescond's publication trajectory reveals a strong emphasis on translating AI innovations to clinical oncology, notably through contributions to the DAISY trial investigating trastuzumab deruxtecan in metastatic breast cancer. His work combines technical AI advancements with practical clinical insights, particularly regarding HER2 expression variability and treatment resistance mechanisms.

Scientific contributions:

• Lead developer of memory-efficient pathological image analysis frameworks
• Key contributor to DAISY trial biomarker analyses
• Innovator in unsupervised segmentation techniques for histopathology

His research is conducted within Institut Curie's collaborative oncology ecosystem, working alongside clinicians and computational biologists to advance precision medicine. Current efforts focus on improving model robustness for clinical deployment and exploring multi-modal data integration for treatment response prediction.