About Dotplot
Dotplot is developing a breast health monitoring tool that empowers women to perform accurate monthly breast self‑checks at home. Our mobile app works alongside a handheld ultrasound device to guide users through the scan in real time using a 3D model of their chest. We simultaneously capture location specific images for analysis.
Our mission is to improve breast health awareness and support earlier detection of breast cancer through intuitive, user-friendly technology.
The role
We are looking for a Machine Learning Research Lead to own the Novel Inference Layer (NIL): Dotplot's personalised change-tracking system for identifying meaningful session-to-session changes in breast tissue.
This is a standalone, Python-led research role spanning dataset readiness, novelty characterisation, model prototyping, benchmarking, and scientific review. Direct ultrasound experience is welcome but not essential. Strong adjacent medical-imaging experience in areas such as CT, MRI, or sensor fusion is also relevant.
What you will own:
- Establish NIL research governance, secure data access, experiment tracking, dataset versioning, leakage prevention and reproducible analysis processes for Dotplot's annotated user-generated ultrasound data.
- Review the state of the art in longitudinal personalised modelling, session-to-session alignment, change and anomaly detection, and uncertainty calibration under variable acquisition conditions.
- Formally characterise NIL's novelty for personalised longitudinal baselines
- Develop and evaluate reproducible Python prototypes for alignment, personalised baseline modelling, change detection and confidence estimation using limited, imperfect real-world data.
- Define baseline comparators, validation metrics, statistical safeguards and performance thresholds for longitudinal alignment and uncertainty calibration.
- Prepare structured review materials, document evidence, and decisions.
What we are looking for:
- Advanced degree (Master’s or PhD) or equivalent research experience, in Computer Science, Machine Learning, Computer Vision, Medical Imaging, Biomedical Engineering or a closely related field.
- Strong practical experience developing and evaluating machine-learning systems in Python using a framework such as PyTorch.
- Research or applied experience in medical imaging or a closely adjacent imaging or sensing field, including CT, MRI, ultrasound, computer vision or sensor fusion.
- Strong understanding of longitudinal modelling, registration or alignment, distribution shift, uncertainty quantification and robust experimental design.
- Experience benchmarking academic methods and working with limited, imperfect or non-standardised real-world datasets.
- Ability to turn a broad research concept into precise claims, hypotheses, experiments, comparators and acceptance criteria, and communicate them clearly to scientific and non-technical audiences.
- Interest in women’s health, early cancer detection, and building accessible, user-focused technology solutions.
Helpful but not essential experience:
- Ultrasound imaging, lay-user acquisition, or objective image-quality assessment.
- Longitudinal imaging, personalised modelling, conformal prediction, Bayesian uncertainty, or calibration metrics.
- Microsoft Azure, controlled research environments, and machine-learning experiment tracking.
- Evidence generation for patents, grants or medical device development.
Practical details:
Location: London-based, hybrid, with 2-3 days each week in the office.
Start Date: Latest end of October.
Engagement: The role can begin as a contract, with the longer-term model agreed as the programme develops.
Pay: £54,000.00 per year
Work Location: Hybrid remote in Battersea SW11
