Ben Van Calster — statistical methods for clinical prediction, ovarian cancer diagnostics to decision-focused ML
2
papers here
1
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
86
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-17) 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 score yet: this profile has no validated publication census behind the axis, so productivity cannot be rated. It would measure cumulative peer-reviewed output and its growth over time; the raw feed does list an h-index of 86, pending platform verification.
Impact—
No impact score yet: citation distributions and field-normalized benchmarks are absent from this snapshot, so influence cannot be rated. It would measure citation impact against field norms; the raw feed lists an i10-index of 251, pending verification.
Focus100/100
Focus scores 100/100. The OpenAlex topic allocation is tightly concentrated: ovarian cancer diagnosis and treatment leads with 172 works, followed by statistical methods in epidemiology (142), artificial intelligence in healthcare (95), breast lesions and carcinomas (81) and reproductive health and contraception (70). The top two topics pair one clinical domain with its enabling methodology — a coherent statistics-meets-oncology agenda rather than scattered interests.
Collaboration46/100
Collaboration scores 46/100. The recent three-year window lists 99 works with 9 distinct co-authors, a concentrated network for that volume. Frequent partners anchor a European clinical-prediction-methods cluster: Richard D. Riley, Gary S. Collins, Maarten van Smeden, Ewout W. Steyerberg, Paula Dhiman, João Matos, Koen M. F. Gorgels, Lasai Barreñada and Wouter A. C. van Amsterdam. Both cataloged 2026 papers are middle-author, non-corresponding positions, consistent with collaborative rather than lone work.
Activity100/100
Activity scores 100/100. The counts-by-year series runs unbroken from 2000 (1 work) to 2026 (30 works), peaking at 49 works in 2023, with 33 in 2024, 36 in 2025 and 30 in 2026. The last three years total 99 works — 33 per year on average — indicating a sustained, currently high publication tempo rather than episodic bursts.
Reproducibility100/100
Reproducibility scores 100/100 (reproMean 100), grounded in the one Evolve-evaluated paper so far — the 2026 Smooth Net Benefit preprint, whose reproducibility readiness is rated high. The evidence base is a single paper (reproPapers: 1), so the perfect score reflects one strong data point rather than a track record; future reviews could move it.
Research summary
Ben Van Calster (Utrecht University, Netherlands; ORCID 0000-0003-1613-7450) is a methodologist in clinical prediction modelling, profiled here from public OpenAlex bibliographic data. The snapshot reports an h-index of 86 and an i10-index of 251, with an unbroken counts-by-year series from 2000 (1 work) to 2026 (30 works), peaking at 49 works in 2023 and totalling 99 works over the last three years. Topic allocation is tightly concentrated: ovarian cancer diagnosis and treatment (172 works) leads, followed by statistical methods in epidemiology (142), artificial intelligence in healthcare (95), breast lesions and carcinomas (81) and reproductive health and contraception (70) — a statistics-meets-oncology agenda. Radar dimensions score 100 for activity, focus and reproducibility, and 46 for collaboration (9 distinct co-authors in the recent window); output and impact are not yet scored. Recent cataloged work turns to how prediction models should be judged: a 2026 arXiv study introducing Smooth Net Benefit (σNB) as a training objective, which Evolve's review read as a largely negative result — only modest, context-dependent gains — with high reproducibility readiness; and a second 2026 arXiv paper on collapsibility of performance metrics in clinical predictive AI. Frequent co-authors — Riley, Collins, van Smeden, Steyerberg, Dhiman and others — place him within a European clinical-prediction-methods network, typically in middle authorship.
Collaboration pattern
Collaboration scores 46/100, with 9 distinct co-authors in the recent three-year window — a concentrated network for that output volume. Recurring partners form a European clinical-prediction-methods cluster: Richard D. Riley, Gary S. Collins, Maarten van Smeden, Ewout W. Steyerberg and Paula Dhiman, alongside João Matos, Koen M. F. Gorgels, Lasai Barreñada and Wouter A. C. van Amsterdam. Both 2026 cataloged papers list middle, non-corresponding authorship, consistent with contributing within larger methodological teams.
Research keywords
Highlights
Smooth Net Benefit: training on the decision, not the prediction
Collapsibility of performance metrics in clinical predictive AI
A second 2026 arXiv preprint (middle authorship, non-corresponding) examining how performance metrics behave in clinical predictive AI. It is cataloged without an Evolve verdict, keywords or reproducibility rating yet.
A statistics-meets-oncology portfolio with sustained tempo
OpenAlex topics pair methodology with clinical domains — ovarian cancer diagnosis and treatment (172 works), statistical methods in epidemiology (142), artificial intelligence in healthcare (95), breast lesions and carcinomas (81), reproductive health and contraception (70) — while annual output ran unbroken from 2000 to 2026, peaking at 49 works in 2023.
Timeline
- 2000Earliest year in the OpenAlex counts-by-year series, with 1 indexed work.
- 2008Annual output reached an early peak of 39 indexed works (27 in 2007), the highest single-year count before 2023.
- 2016Mid-period output of 36 indexed works, matching the 2012 level within an unbroken 2010s run of 16–36 works per year.
- 2023Highest single-year total in the record: 49 indexed works, up from 27 in 2022.
- 2026Two Evolve-cataloged arXiv preprints appear: Smooth Net Benefit as a training objective, and collapsibility of performance metrics in clinical predictive AI; both list middle, non-corresponding authorship.
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.
- Bibliometrics are drawn from an OpenAlex snapshot only (identity: openalex); output and impact dimension scores are not yet computed for this profile.
- h-index (86) and i10-index (251) are raw, unnormalized counts, and the profile is name-resolved from public bibliographic data; identity should be confirmed via ORCID 0000-0003-1613-7450.
- The reproducibility score of 100 rests on a single evaluated paper (reproPapers: 1), so the evidence base is thin.
- Only two 2026 arXiv preprints (middle authorship, non-corresponding) are in the Evolve catalog; one lacks a verdict, keywords and slug.