3D X-Ray Microscopy PhD: Ptychography and Artificial Intelligence

3D X-Ray Microscopy PhD: Ptychography and Artificial Intelligence

Quick Decision

This Sheffield PhD proposes to improve computational X-ray microscopy using ptychography and AI, then extend the method toward three-dimensional imaging in collaboration with Diamond Light Source. It is a particularly relevant opportunity for applicants interested in inverse problems, computational imaging, image processing or scientific software who also want to work with experimental beamline data. The project description names low-carbon cement and large-scale mapping of neuronal connections as possible application areas; these are examples of future trials, not confirmed thesis work packages.

The key practical advantage is a close topic match between the advertised work, the supervisors' imaging expertise and Diamond's operational ptychography/tomography facility. The main technical constraint is scale: Diamond's own rough estimates put 3D acquisition at more than an hour for a 10 µm field of view and more than 24 hours for a 200 µm field at around 100 nm resolution, depending on sample and desired resolution. Funding is through a competitive EPSRC Doctoral Landscape Award (DLA). The deadline is 17:00 UK time on Wednesday 13 January 2027, with a required start date of 1 October 2027. International candidates face a Home-first competition and a 30% cap on awards; the exact stipend and duration for this place are not published in the materials reviewed. Applicants should contact the supervisors early and confirm the award terms and research scope before committing time to a full application. (DLA application guidance; Diamond I13-1 imaging modalities)

Research Problem and Doctoral Work

What Is Unsolved and Why It Matters

Ptychography reconstructs sample information computationally from diffraction measurements collected across overlapping illumination positions. The opportunity description frames its challenge as improving the resolution and contrast available from X-ray microscopy, then developing methods that can recover three-dimensional rather than only two-dimensional information. A central research problem is to make richer reconstructions useful at realistic sample sizes: 3D imaging adds a much larger data and computation burden, and long acquisition times can limit throughput.

Diamond's I13-1 beamline confirms that ptychography is a principal imaging modality and that tomography and complementary measurements are available. For roughly 100 nm resolution, its indicative acquisition table estimates more than 1 hour for a 10 µm 3D field, more than 6 hours for 100 µm and more than 24 hours for 200 µm. These are facility estimates rather than a schedule for this PhD, and actual times vary with sample composition and target resolution. They make acquisition efficiency, data handling, reconstruction runtime, robustness and the resolution/volume trade-off useful questions to discuss with the supervisors. (I13-1 imaging modalities)

Work, Methods, and Potential Outputs

The advertised work is computational: develop, study and implement new X-ray imaging approaches with Diamond beamline scientists; use ptychography and AI-supported algorithms to improve resolution or contrast; and pursue a 2D-to-3D extension. The project description also anticipates trials on examples ranging from low-CO₂ cement to large-scale neuronal-connectivity mapping. It does not name a specific model architecture, software stack, sample set, benchmark, target volume or expected publication. The final balance between algorithm design, X-ray physics, beamline experiments and application-led analysis therefore remains to be agreed.

Recent research in the wider ptychography toolkit shows how algorithmic choices affect robustness and scale. Maiden, Mei and Li's 2024 WASP paper introduces a weighted average of sequential projections and compares it with alternatives using simulations and real measurements; the authors report fast early convergence, robustness to noise and initial conditions, and a small memory footprint. These are relevant dimensions for a project with growing 3D datasets, but the reported behavior is tied to the paper's tests and is not a promise of performance on this PhD's X-ray volumes. (WASP paper)

Hu, Zhang and Maiden's 2024 near-field multi-slice study combines computational optical sectioning with ptychography to image optically thick samples in 3D and maintain axial information beyond conventional depth-of-field limits in the tested setup. It demonstrates a methodological route for depth-aware reconstruction, while remaining an optical rather than X-ray study. Song and colleagues' 2024 ptycho-endoscopy paper, which includes Maiden, adapts ptychographic reconstruction to a lensless fiber-bundle geometry. It illustrates the team's broader computational-imaging activity, but is not evidence for the performance of synchrotron X-ray tomography. (Near-field multi-slice paper; ptycho-endoscopy paper)

Allars, Maiden, Batey and Allen's 2025 multi-slice electron-ptychography study uses simulated data to compare depth-of-field behavior across beam-energy and convergence-angle settings. It reports regimes where multi-slice ptychography has an advantage while also showing that signal and detector conditions still matter. This is useful evidence that imaging performance depends on the acquisition setup, but it concerns electron microscopy and simulations, not X-ray experiments. A separate 2024 synchrotron paper on cement pastes demonstrates that near-field X-ray ptychographic tomography can address complex materials microstructure; it makes the project's cement motivation plausible as a field application, without confirming that the PhD will use those samples or produce a low-carbon cement formulation. (2025 electron ptychography paper; 2024 cement tomography paper)

Plausible outputs include improved reconstruction or acquisition methods, validated 3D X-ray imaging workflows and application-specific image analyses. These are research possibilities implied by the topic, not advertised deliverables. Ask which performance criteria, sample areas and algorithmic comparisons will define progress.

School, Programme and Field Context

The University of Sheffield is the named university and the School of Electrical and Electronic Engineering is the stated academic home. Diamond Light Source is the collaborating UK national synchrotron facility; the sources reviewed do not describe it as a degree-awarding partner or advertise a joint degree. Sheffield EEE links its PhD community to the Engineering Doctoral School and the University-wide Doctoral Development Programme, which provides general PGR training and development. The University's research-degree code gives most full-time PhDs a normal registration period of 3–4 years and a four-year maximum, subject to funder and training requirements; the exact period is set in the offer, and this DLA project's duration remains unconfirmed. (EEE PhD study; Doctoral Development Programme; research degree code of practice)

Sheffield is ranked joint 82nd in the overall QS World University Rankings 2027. The University reports that 92% of its research was placed in the highest two categories in REF 2021, an institution-wide measure. Neither establishes a subject ranking for EEE or a rating of this research group. The University address confirms Sheffield, United Kingdom, as the degree-awarding institution's location; Diamond's I13-1 facility is in Oxfordshire. (QS profile; Sheffield facts and figures; University contact details; I13-1 beamline)

The official DLA page sets the 1 October 2027 start date, but the project-specific award length, study mode, progression checkpoints and training requirements beyond the general PGR offer are not stated in the reviewed project information. Confirm them in the offer and with the School. (DLA application guidance; registration policy)

The adjacent career market includes scientific software and data analysis, beamline or facility science, imaging instrumentation, photonics research and computational imaging roles in materials or technology. The UK Photonics Leadership Group's 2025 sector profile reports 84,800 UK photonics employees and 8.2% employment growth from 2022 to 2024; Diamond also lists science, engineering, software/computing and software/data-analysis career families. This supports the existence of related employment settings, not a vacancy forecast for X-ray ptychography or a job-placement promise. (UK photonics sector data; Diamond careers)

Supervisor and Lab Assessment

Supervisor Research Directions

Prof Andrew Maiden's EEE profile identifies research in computational imaging and ptychography, closely matching the core method. Dr Charith Abhayaratne's work covers image and video analysis, computer vision, machine learning and signal processing. His group lists current doctoral work in image super-resolution and biological image analysis. The combination is a credible basis for algorithmic image reconstruction, but the sources do not specify who will lead which part of this project or how AI will be applied. (Andrew Maiden profile; Charith Abhayaratne profile; research team)

Recent Important Publications

• Maiden, Mei and Li (2024), WASP. Adds and evaluates an iterative ptychographic phase-retrieval approach, including noise and initialization comparisons. Its emphasis on convergence, memory and parallel computation is relevant to the project's scaling questions; its exact performance is test-dependent. (Paper record)
• Hu, Zhang and Maiden (2024), near-field multi-slice ptychography. Demonstrates computational optical sectioning and 3D phase imaging in thick samples. It supports the methodological direction of depth-resolved reconstruction, but is not an X-ray result. (Paper record)
• Song et al. (2024), ptycho-endoscopy. Demonstrates a computational ptychographic strategy in a constrained optical fiber-bundle geometry. It evidences breadth in image reconstruction, not a direct route to the project's synchrotron experiments. (Paper record)
• Allars, Maiden, Batey and Allen (2025), multi-slice electron ptychography. Uses simulated electron data to examine depth of field and acquisition-parameter limits. It is useful for understanding how system settings shape reconstruction, while modality transfer to X-rays must be tested. (Paper)

Lab, Resources, and Collaboration Environment

Diamond I13-1 is operational and supports coherent X-ray modalities including ptychography, tomography and Bragg coherent diffraction imaging. Its documentation describes dedicated acquisition, processing and analysis software for ptychographic tomography, a beamtime data service and the possibility of follow-up processing on a beamline workstation. These are relevant infrastructure signals, not a guarantee of this student's allocation, local compute, storage, software support or travel. The project description says the student would work with beamline scientists; ask how often that interaction occurs and what sample/data access is planned. (I13-1 overview; computing and software)

Funding, Application, and Living Cost

The position is described as a competitive, fully funded EPSRC DLA opportunity, within Sheffield's 2027 competition of funded science, engineering, health and maths projects. (DLA scholarship overview) Sheffield's current instructions require the unique DLA form; standard PhD applications do not qualify for this competition. The form allows up to two relevant project choices and strongly encourages discussing fit with supervisors. It requires a 1 October 2027 start and closes at 17:00 UK time on 13 January 2027. The form also asks applicants to disclose and explain any AI use; read the linked DLA Use of AI in Recruitment Statement and follow the current instructions. (DLA application guidance; current DLA opportunities)

International applicants should treat the award as highly competitive. EPSRC limits international awardees to 30% of total places; Sheffield considers Home applicants first. One saved record anticipates up to ten awards without specifying the scope, while another refers to up to ten international awards; confirm the quota with Sheffield. The University says it will fund the Home/International fee difference for international candidates who win a place, while UKRI's own fee contribution is capped at the Home rate. Sheffield's 2027/28 EEE fee lookup lists Overseas PhD tuition at £31,870–£36,870; the 2027/28 Home fee is still to be confirmed, and research-support enhancements may apply. UKRI lists a minimum full-time stipend of £21,805 from 1 October 2026, but that is not a confirmed 2027/28 stipend for this project. The specific offer should state stipend, duration, fees and research costs. (DLA application guidance; UKRI student support; Sheffield fee lookup)

Sheffield's 2026/27 postgraduate accommodation examples include shared-bathroom rooms at £155.68–£173.32 per week on 42- or 51-week contracts; University rent includes utilities, contents insurance and internet. The University's monthly cost guide also lists survey-based figures for groceries (£155), transport (£73), going out (£60), takeaways/eating out (£45), clothing/shopping (£33), mobile (£15) and subscriptions (£14). These figures reflect different evidence sources and spending patterns, not a guaranteed monthly total; private housing bills may be extra, and amounts can change before October 2027. (Sheffield living-cost guidance; 2026/27 accommodation rents)

Applicant Fit and Preparation

Required Background and Skills

The reviewed project information does not specify a required degree subject or classification, programming language, prior imaging experience or beamline experience. The DLA application page gives competition-wide instructions but does not establish project-specific academic prerequisites. Applicants should ask the supervisors or DLA team for the formal requirements rather than treating topic methods as mandatory admissions criteria.

Preferred Experience and Suggested Preparation

No preferred background is formally stated. Based on the advertised work, relevant preparation may include image or signal processing, inverse problems, numerical optimization, computational imaging, machine learning, optics/X-ray physics or scientific programming. A portfolio or research example showing how an applicant worked with noisy measurements, tested reconstruction choices or connected a model to real data could help explain methodological readiness. Prior access to a synchrotron is not stated as a prerequisite. Before applying, discuss whether the thesis is weighted toward computational method development, experimental imaging or one of the named applications. (Andrew Maiden profile; Charith Abhayaratne profile; DLA application guidance)

Next Actions and Questions to Verify

Before Applying

1. Confirm that project code C3.5-ELE-Maiden is listed in the live DLA form and that the deadline remains 17:00 UK time on 13 January 2027.
2. Ask the supervisors to clarify the expected balance between algorithms, X-ray acquisition, AI and application work; ask whether cement and neuronal-connectivity examples are committed collaborations or illustrative possibilities.
3. Confirm academic/English eligibility, Home or Overseas fee status, and the exact stipend, tuition coverage, award duration and research-cost provision. If choosing a second project, follow the University's instruction to select only a genuinely relevant option.
4. Read the DLA AI-use statement and complete the form's disclosure question as required.

At Interview

• What volume, resolution and reconstruction-runtime benchmarks will define project progress?
• What beamtime, compute, data storage, software and training will be available, and how are beamline scientists involved in supervision?
• How are the two supervisors' responsibilities divided, and what sample/application work is confirmed?

Before Accepting an Offer

• Check the written offer for registration length, stipend schedule, full tuition coverage including any fee enhancement, and the treatment of visa, travel and research costs.
• Confirm data/sample permissions, partner responsibilities, publication conditions and the planned milestones for the award.

Evidence and Sources

University, Programme, and Funding

• Current EPSRC DLA opportunities
• How to apply for the EPSRC DLA
• Fully funded EPSRC DLA projects
• EEE PhD study
• Engineering Doctoral School
• Doctoral Development Programme
• Research degree code of practice
• University contact details and location
• University facts and figures
• QS World University Rankings 2027 profile
• UKRI support for funded students
• Sheffield postgraduate research fees

Supervisors, Papers, and Facility

• Professor Andrew Maiden
• Dr Charith Abhayaratne
• Charith Abhayaratne research team
• Diamond I13-1 X-ray Coherence beamline
• I13-1 imaging modalities and acquisition-time estimates
• I13-1 computing and software
• WASP ptychographic phase retrieval
• Near-field multi-slice ptychography
• Ptycho-endoscopy on a lensless fiber bundle
• Depth of field in multi-slice electron ptychography
• X-ray ptychographic nanoimaging of cement pastes

Living Costs and Career Context

• Sheffield living-cost guidance
• 2026/27 postgraduate accommodation rents
• UK photonics sector data
• Diamond careers