Fusing CT radiomics with foundation models for renal cancer classification at University College Dublin
2
papers here
2
reviewed by Evolve
2
first-author
0
corresponding
—
works · OpenAlex
—
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-02) 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 publication-count or venue data is available yet, so this axis is unscored. It would measure total research output — peer-reviewed papers, preprints, and their spread across venues and years.
Impact—
Citation-based metrics — citations, h-index, and citing works — are not present in this snapshot, so impact is unscored. This axis would capture how widely the two renal-classification preprints are read, reused, and cited by other groups.
Focus0/100
The radar scores focus at 0, driven by the raw field distribution: the two reviewed papers split evenly, 0.5 each, between medical image analysis (computational radiology) and medical image analysis (radiomics + foundation models). In substance the spread is narrow — both sit inside CT-based renal tumour analysis — but the classifier treats the two labels as fully dispersed, so topical concentration registers at the floor.
Collaboration22/100
Collaboration scores 22 on the radar, matching the raw count of just 2 distinct coauthors across the reviewed record. Both 2026 preprints are first-author, non-corresponding works; the frequent coauthors are Sourav Bhattacharjee and Abraham Campbell, suggesting a small, stable University College Dublin team rather than a broad external network.
Activity—
Publication cadence cannot be assessed: the snapshot provides no yearly output series, so activity is unscored. It would measure publishing rhythm — papers per year and the gaps between them — currently visible only as two preprints both dated 2026.
Reproducibility55/100
Reproducibility scores 55, the mean across 2 scored papers, each rated medium readiness. Evolve's verdicts flag the same gap in both: headline results — 82.7% AUC on KiTS23 and AUC 0.804/0.854 with RenalCLIP — rest on small test sets with untested margins and no significance testing, so released code, data splits, and baseline details would decide actual reproducibility.
Research summary
Yuan Liang is a medical image analysis researcher at University College Dublin whose reviewed work centres on renal tumour characterisation from CT. The Evolve catalog holds two 2026 arXiv first-author preprints, both fusing handcrafted radiomic features with foundation-model representations for renal cell carcinoma (RCC) classification. One compares 2D and 3D CT encodings and reports a gated 3D fusion of radiomics with MedVAE latents reaching 82.7% AUC on KiTS23 RCC, above either feature family alone; Evolve judges the claim plausible but notes it rests on a 60-case test set without significance testing. The other applies radiomics-conditioned FiLM modulation to RenalCLIP features for clear cell RCC, reporting AUCs of 0.804 and 0.854, with Evolve cautioning that margins over simpler fusion strategies and a strong SVC baseline are small and untested. The radar records reproducibility at 55 across two medium-readiness papers, collaboration at 22 with two distinct coauthors — Sourav Bhattacharjee and Abraham Campbell — and focus at 0, reflecting an even 0.5/0.5 split between computational radiology and radiomics-plus-foundation-models, both within renal imaging. Output, impact, and activity metrics are unavailable, so this remains an early-stage, preprint-centric profile.
Collaboration pattern
The reviewed record shows a compact collaboration footprint: 2 distinct coauthors and a radar collaboration score of 22. Sourav Bhattacharjee and Abraham Campbell appear as the frequent coauthors, consistent with a University College Dublin medical-imaging group. Both papers are first-author and non-corresponding, so senior collaborators likely supervise; no external or multi-institution partnerships are visible in this snapshot.
Research keywords
Highlights
Gated radiomics–MedVAE fusion tops either feature set alone on KiTS23
Radiomics-conditioned FiLM modulation of RenalCLIP features
A consistent thesis: radiomics and foundation representations are complementary
Across both 2026 preprints the same argument recurs — handcrafted radiomics and foundation-model encodings carry complementary information for renal tumour classification — making fusion design (gated fusion, FiLM conditioning) the unifying methodological signature of this early record.
Timeline
- 2026Publishes a comparative study of 2D and 3D CT encodings on arXiv, reporting radiomics + MedVAE gated fusion at 82.7% AUC on KiTS23 renal cell carcinoma.
- 2026Publishes radiomics-conditioned FiLM modulation of RenalCLIP features for clear cell renal cell carcinoma classification, reporting AUC 0.804/0.854.
- 2026Both first-author arXiv preprints appear at University College Dublin, establishing a radiomics-plus-foundation-model line in renal imaging.
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 are computed from a limited snapshot (OpenAlex-style data); output, impact, and activity axes lack underlying metrics and are unscored.
- Only two 2026 arXiv preprints are reviewed; no peer-reviewed venue, citation, or funding information is available.
- 'Yuan Liang' is a common name and no ORCID is provided; this profile is name-resolved from public bibliographic data and identity should be confirmed before reuse.
- Country is not specified in the source data; the institution is listed as University College Dublin.