Télécom Paris medical-imaging researcher: multimodal 3D registration benchmarking and cortical sulci deep learning
2
papers here
2
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
—
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-15) and cover the author's entire public record, not just the papers listed here. Reproducibility signal comes from Evolve's own reviews.
Research profile
Output—
No output metric is available yet, so this axis is unscored. It would measure publication volume — paper counts, authorship positions and venue spread over time — which a snapshot of two 2026 arXiv papers cannot support.
Impact—
No citation evidence underpins this axis, so it stays unscored. Impact would measure citations and field-weighted influence of the work; both cataloged papers are 2026 arXiv preprints, too recent for accumulated citation data.
Focus0/100
The radar records a focus score of 0, and the underlying field distribution splits evenly at 50/50 between two adjacent sub-fields: medical image analysis via computer vision and computational neuroimaging. Both cataloged papers stay inside medical image analysis — a multimodal 3D registration benchmark on the vision side, cortical sulci labeling on the neuroimaging side — so the profile reads as narrow in discipline yet evenly forked between method benchmarking and neuroimaging application.
Collaboration48/100
Collaboration scores 48/100, built on 10 distinct coauthors in the snapshot. The two cataloged papers are multi-author arXiv efforts in which Gori holds middle and last positions rather than first author. His frequent-coauthor list spans a clinical-imaging group — Matteo Barbieri, Giammarco La Barbera, Juan Pablo De La Plata, Sabine Sarnacki, Saeb Tounsi, Joël Chavas, Isabelle Bloch — and a neuroimaging group — Vincent Frouin, Denis Rivière, Jean-François Mangin — mirroring his two sub-fields.
Activity—
No longitudinal signal is available, so this axis is unscored. Activity would track publishing cadence over time — papers per year and gaps between outputs — but the snapshot contains only two 2026 papers.
Reproducibility55/100
Reproducibility scores 55/100, the mean over 2 assessed papers, both rated medium readiness. The registration benchmark compares 9 methods on 3 datasets yet lacks statistical testing, and the sulci-labeling study reports concrete Dice gains (0.77 vs 0.72) on a small N=22 test set; code or data availability is not verified here, so the score reflects partially replicable, cautiously positive evidence.
Research summary
Pietro Gori is a researcher at Télécom Paris (ORCID 0000-0003-3456-8529) whose cataloged work sits squarely in medical image analysis, split evenly at 50/50 between a computer-vision strand and a computational-neuroimaging strand. The Evolve catalog holds two 2026 arXiv papers. As last author on 'Benchmarking Intra-Patient 3D Deformable Multimodal Image Registration', he contributes to an evaluation of nine multimodal 3D registration methods across three datasets that finds no robust winner and substantial metric disagreement — credible and useful, in Evolve's judgment, though lacking statistical testing. As a middle author on 'Geometric-to-Semantic Spherical Transfer Learning for Cortical Sulci Labeling', he is part of a team showing that self-supervised pretraining on around 30,000 UK Biobank surfaces with soft-initialized TPI lifts Dice from 0.72 to 0.77, on a small 22-subject test set. Radar scores report collaboration at 48/100 with 10 distinct coauthors and reproducibility at 55/100 (mean over 2 papers, both medium readiness); output, impact and activity remain unscored for insufficient data. Frequent coauthors span clinical-imaging collaborators and a neuroimaging group including Jean-François Mangin, Vincent Frouin and Denis Rivière. This full-tier profile is an early snapshot: evidence is genuine but thin, and all quoted figures come from the two cataloged papers.
Collaboration pattern
The snapshot records 10 distinct coauthors and a collaboration score of 48/100. Both cataloged papers are multi-author arXiv efforts in which Gori takes middle and last authorship positions rather than first author. His frequent-coauthor list spans a clinical-imaging cluster (Matteo Barbieri, Giammarco La Barbera, Juan Pablo De La Plata, Sabine Sarnacki, Saeb Tounsi, Joël Chavas, Isabelle Bloch) and a neuroimaging cluster (Vincent Frouin, Denis Rivière, Jean-François Mangin), matching his two sub-fields.
Research keywords
Highlights
No robust winner in 3D multimodal registration benchmark
Spherical transfer learning lifts sulci-labeling Dice to 0.77
Medium reproducibility readiness across both 2026 outputs
Both cataloged papers carry medium reproducibility readiness: the registration benchmark shares 9-method comparisons over 3 datasets but lacks statistical testing, while the sulci study reports concrete Dice gains on only 22 test subjects — together yielding the 55/100 reproducibility mean over 2 papers.
Timeline
- 2026Last-author position on the arXiv benchmark of intra-patient 3D deformable multimodal image registration: 9 methods, 3 datasets, no robust winner.
- 2026Middle-author contribution to geometric-to-semantic spherical transfer learning for cortical sulci labeling: Dice 0.77 vs 0.72 with UK Biobank self-supervised pretraining.
- 2026Both cataloged papers are assessed at medium reproducibility readiness, producing a reproducibility mean of 55 over 2 papers.
Papers in the Evolve catalog
Frequent co-authors
About this profile
This profile is AI-structured by Evolve.Pub from public bibliographic data (OpenAlex) and Evolve's own paper reviews. It is not written or endorsed by the researcher unless the profile is claimed. Radar dimensions are computed from real bibliometrics with fixed formulas — raw figures are shown alongside; missing dimensions mean insufficient data, never zero. Errors? Claim the profile to correct or hide it.
- Bibliometric scores derive from public bibliographic data (OpenAlex-type sources) only; the output, impact and activity axes are unscored for insufficient data.
- Only two 2026 arXiv papers from the Evolve catalog inform this profile; journal articles and earlier career history are not represented.
- The reproducibility score of 55 rests on just 2 assessed papers, both at medium readiness — a small evidence base.
- Identity is resolved via ORCID 0000-0003-3456-8529 and public bibliographic records; Evolve verdicts reflect the cataloged papers only.