Postdoctoral Researcher

Heriot Watt University

Project Overview

The AI Pathologist project is an interdisciplinary initiative aimed at developing an advanced AI-driven diagnostic system using 3D medical imaging modalities (CT, MRI, PET) and explainable AI (XAI) techniques to support medical professionals in improving diagnostic accuracy, reducing workload, and enhancing patient outcomes. The Postdoc Researcher will play a key role in implementing system prototypes, managing data pipelines, and supporting the development and validation of AI models in collaboration with clinicians and engineers.

Key Responsibilities

1. AI Model Development & Testing

• Assist in developing machine learning and deep learning models for medical imaging analysis.
• Implement and fine-tune models using PyTorch, TensorFlow, or related frameworks.
• Run experiments, perform hyperparameter tuning, and maintain organized experiment logs.

2. Medical Imaging Data Management

• Support data preprocessing, annotation preparation, and dataset organization.
• Handle CT, MRI, PET, and histopathology images, ensuring proper anonymization and compliance with ethical guidelines.
• Collaborate with clinicians to understand imaging structures and diagnostic targets.

3. Explainable AI Implementation

• Integrate XAI methods (e.g., Grad-CAM, Guided Backprop, SHAP) into the system.
• Generate and interpret visual explanations for model outputs.

4. System Prototyping & Software Development

• Contribute to the development of the AI Pathologist prototype, including UI/UX, backend integration, and workflow automation.
• Support deployment of models using Docker, APIs, or cloud environments.

5. Documentation & Reporting

• Prepare technical documentation, coding standards, and user guides.
• Assist in preparing project reports, research papers, and meeting presentations.

6. Collaboration

• Work closely with radiologists, pathologists, and clinicians to understand requirements and validate outputs.
• Attend project meetings and provide weekly progress updates.

Minimum Requirements (Essential)

Education

• Master degree or PhD in Computer Science, Artificial Intelligence, Data Science, Computer Engineering, Biomedical Engineering, or a related field.

Technical Skills

• Experience with Python and machine learning libraries (PyTorch, TensorFlow, scikit-learn).
• Understanding of deep learning, CNNs, Capsule Networks, and image processing fundamentals.
• Basic knowledge of handling medical images (e.g., DICOM).
• Familiarity with version control (Git/GitHub).

Soft Skills

• Strong analytical and problem-solving skills.
• Good communication and teamwork ability.
• Ability to document work clearly and follow research procedures.

Preferred Requirements (Highly Desirable)

Technical Skills

• Experience with:
  • 3D imaging models (3D CNNs, UNet variants, VNet)
  • Explainable AI tools (Grad-CAM, LIME, SHAP)
  • Medical imaging workflows, PACS systems
  • Docker, APIs, and model deployment
  • Cloud platforms (AWS, Azure, GCP)
• Experience working on healthcare-related AI projects.

Clinical Collaboration

• Previous collaboration with hospitals or medical researchers.
• Understanding of radiology/pathology imaging interpretation.

Other Advantages

• Publications or strong project portfolio in computer vision/medical AI.
• Experience with data annotation tools.
• Knowledge of cybersecurity and data governance in healthcare.