Dimitris K. Iakovidis

University of Thessaly · Greece

Unclaimed · AI-structured profileORCID OpenAlex

University of Thessaly ML researcher spanning medical image analysis and explainable AI for engineering

papers here

reviewed by Evolve

first-author

corresponding

—

works · OpenAlex

h-index · OpenAlex

Bibliometric figures come from OpenAlex (2026-10-13) 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 consolidated output score is available yet. This axis would measure total publication volume and productivity over time; the snapshot lists yearly work counts since 2001, but Evolve has not computed a scored output dimension.

Impact—

No impact score is available yet. This axis would measure citation-based influence; raw h-index 38 and i10-index 105 are provided, but a computed impact dimension could not be derived from the current data.

Focus0/100

The focus score is 0, reflecting an evenly split field distribution: the three cataloged 2026 preprints each occupy about one third — predictive maintenance / applied machine learning, computational mechanics / scientific machine learning, and machine learning with explainable AI. The longer OpenAlex topic history is likewise broad, spanning image retrieval and classification (53), medical image segmentation (40), AI cancer detection (35), gastrointestinal bleeding (34) and radiomics (25).

Collaboration39/100

Collaboration scores 39 from 6 distinct co-authors detected in the snapshot — a compact, small-team pattern. The recurring partners on recent preprints are Dionisis Kalogeropoulos, Georgia Sovatzidi, Panagiotis G. Kalozoumis and Georgios Triantafyllou. On all three cataloged 2026 papers Iakovidis appears as last author (non-corresponding), consistent with a senior supervisory role in a stable group rather than large multi-institution consortia.

Activity100/100

Activity is the strongest axis at 100. The OpenAlex record spans 26 consecutive years (2001–2026) without a gap, with 41 works in the last three years. Annual counts have run at double digits almost continuously since 2019 (11, 12, 12, 14, 12, 13, 10), and 2026 already shows 18 works — matching the 2008 historical peak. All three Evolve-cataloged preprints date from 2026.

Reproducibility20/100

Reproducibility scores 20, the mean readiness across the 3 Evolve-reviewed papers, each rated low. Reviewers cite concrete gaps: DeepFEAv2's comparison against DeepFEA is confounded with no ablations; the maritime neuro-fuzzy paper reports no variance in its statistics and overstates SOTA margins against a tied SVM (0.99); the naval-propulsion paper's 99.24% accuracy claim is weakened by baseline-comparison and data-leakage questions.

Research summary

Dimitris K. Iakovidis is a computer scientist at the University of Thessaly (Greece) whose OpenAlex-resolved record stretches from 2001 to 2026 across two broad strands: medical image analysis — led by image retrieval and classification (53 works), medical image segmentation (40), AI-based cancer detection (35), gastrointestinal bleeding diagnosis (34) and radiomics (25) — and, more recently, applied machine learning for engineering. His citation profile is substantial (h-index 38, i10-index 105 in the provided snapshot), and activity is the standout axis (score 100): 41 works in the last three years, with 18 recorded in 2026, matching the 2008 peak. Recent arXiv output concentrates on explainable AI — neuro-fuzzy models for maritime decision-making, fuzzy-logic condition-based maintenance of naval propulsion with a reported 99.24% accuracy on public CODLAG data, and DeepFEAv2, a deep-learning surrogate for transient finite element analysis on unstructured meshes. On all three cataloged papers he appears as last, non-corresponding author. Evolve's reviews flag reproducibility concerns (mean readiness 20/100 across 3 papers, all rated low) and note that headline comparisons are sometimes confounded, lack ablations, or report no variance. Collaboration is comparatively compact (score 39; 6 distinct co-authors detected), with recurring partners including Dionisis Kalogeropoulos, Georgia Sovatzidi, Panagiotis G. Kalozoumis and Georgios Triantafyllou.

Collaboration pattern

Collaboration is compact: the snapshot detects 6 distinct co-authors (score 39). Recent preprints rotate within one small core — Dionisis Kalogeropoulos, Georgia Sovatzidi, Panagiotis G. Kalozoumis and Georgios Triantafyllou. Iakovidis sits in the last-author position on all three cataloged 2026 papers (none corresponding), consistent with supervising a small applied-ML group rather than joining large consortia.

Research keywords

medical image analysisimage retrieval and classificationmedical image segmentationAI-based cancer detectiongastrointestinal bleeding diagnosisradiomicsexplainable AIneuro-fuzzy systemspredictive maintenancecondition-based maintenancefinite element analysis surrogatesdeep learning

Highlights

DeepFEAv2: transient FEA surrogates beyond structured meshes
A 2026 arXiv preprint extends deep-learning surrogates to transient finite element analysis on unstructured meshes using ConvLSTM and autoencoder components, with large reported gains over DeepFEA. Evolve's review calls it promising but notes the comparison is confounded and no ablations are provided; reproducibility readiness is rated low.
Explainable neuro-fuzzy models for trustworthy maritime decisions
The paper pairs a neuro-fuzzy network with saliency maps and fuzzy-rule explanations for decision support in maritime settings. Evolve's review finds the explainability approach promising but flags that the 'outperforms SOTA' claim overstates ties in Table I (an SVM also reaches 0.99) and that the statistics lack variance; reproducibility readiness is rated low.
Explainable condition-based maintenance of naval propulsion systems
A 2026 preprint distills a residual neural network with a fuzzy head into local fuzzy decision trees for explainable predictive maintenance of naval propulsion, reporting 99.24% accuracy on public CODLAG data. Evolve's review notes that baseline-comparison and leakage concerns temper the advantage claims; reproducibility readiness is rated low.

Timeline

2001Earliest works appear in the OpenAlex record (3 works).
2008Annual output reaches an early peak of 18 works.
2019Output climbs to 11 works, beginning a sustained high-activity run.
2022One of the strongest pre-2026 years, with 14 works.
202618 works recorded; three arXiv preprints enter the Evolve catalog (DeepFEAv2, maritime neuro-fuzzy, naval-propulsion maintenance), each with Iakovidis as last author.

Papers in the Evolve catalog

Frequent co-authors

Dionisis Kalogeropoulos×2 Georgia Sovatzidi×2 Dimitris K. Iakovidis×2 Panagiotis G. Kalozoumis×2 Panagiotis G. Kalozoumis×1 Georgios Triantafyllou×1

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 dimensions are computed from OpenAlex records only; the output and impact axes lacked sufficient data to score.
The h-index (38) and i10-index (105) come from the provided snapshot; per-paper citation counts were not available for verification.
'Dimitris K. Iakovidis' is a common Greek name: this profile is name-resolved from public bibliographic data (ORCID 0000-0002-5027-5323), and the frequent-coauthor list contained duplicates plus the author's own name — an apparent resolution artifact.
Quality assessments rest on only 3 Evolve-reviewed papers (all 2026 arXiv preprints); the long medical-imaging output is not individually reviewed.