AI Applied scientist in continual learning for medical imaging

SYNDIAG SRL; Torino; On-site; EUR 60,000 - 85,000; Full time

Job summary

SYNDIAG SRL in Torino, Italy, is seeking an AI Applied Scientist to advance continual learning for OvAI, our EU MDR-certified SaMD platform for ovarian cancer. You will research, design, develop, and evaluate continual learning-based strategies for training computer vision models on ultrasound imaging data, specifically for ovarian tumor segmentation and classification.

This is a high-impact role at the interface of AI, medical imaging, and clinical practice, offering cutting-edge challenges and

Responsibilities

• Research continual learning strategies for training CV models on ultrasound data.
• Design and develop training protocols for ovarian tumor segmentation and classification.
• Evaluate performance and mitigate catastrophic forgetting in deployment.

Skills

Deep learning

Continual learning

Medical imaging

Computer vision

Job description

AI Applied Scientist

Unlike conventional Machine Learning, where models are trained on large, annotated, and static datasets, practical healthcare scenarios demand a more dynamic approach. Annotations are scarce, data arrives in batches, datasets grow incrementally, and models must evolve continuously to meet new business needs and incorporate new capabilities, such as classifying or detecting novel types of organs and tumors.

Continual learning offers potential solutions for these challenges. Under this paradigm, resource-efficient training strategies enable models to introduce new information without requiring complete retraining on previous data. The core objective is to preserve and exploit previously acquired knowledge while seamlessly integrating new data, thereby mitigating the phenomenon of "catastrophic forgetting" prevalent in conventional ML settings.

We are seeking a highly skilled **AI Applied Scientist** with expertise in deep learning to join our R&D team. In this role, you will play a crucial part in advancing the model training protocols that underpin OvAI, our EU MDR-certified SaMD platform for ovarian cancer. Your responsibilities will include researching, designing, developing, and evaluating **continual learning-based strategies** for training our computer vision models on ultrasound imaging data, specifically for ovarian tumor segmentation and classification. This endeavor represents a significant opportunity to contribute to a cutting-edge field, addressing challenges that are novel within the context of clinical applications.