MétaCan
Menu
Back to cohort
Record W4413978744 · doi:10.1093/bjd/ljaf345

Multiharmonic imaging-based automated recognition of cutaneous T-cell lymphoma

2025· article· en· W4413978744 on OpenAlexfundno aff
Shayantani Ghosh, Olesya Pavlova, Alexandra Latshaw, Doyoung Kim, Christoph Iselin, Pauline Bernard, Pacôme Prompsy, Yun‐Tsan Chang, Davide Staedler, Yi-Chien Tsai, Luigi Bonacina, Emmanuella Guenova

Bibliographic record

VenueBritish Journal of Dermatology · 2025
Typearticle
Languageen
FieldMedicine
TopicCutaneous lymphoproliferative disorders research
Canadian institutionsnot available
FundersHorizon 2020 Framework ProgrammeSwiss Cancer Research FoundationCentre Hospitalier Universitaire VaudoisSchweizerischer Nationalfonds zur Förderung der Wissenschaftlichen ForschungMallinckrodt PharmaceuticalsH2020 LEIT Information and Communication TechnologiesUniversité de LausanneUniversité de GenèveHelsinnH2020 Leadership in Enabling and Industrial TechnologiesNational Science Foundation
KeywordsMycosis fungoidesBiopsyPathologyMedicineConvolutional neural networkSkin biopsyLymphomaAtypical LymphocyteCutaneous T-cell lymphomaHaematoxylinStainingArtificial intelligenceComputer science

Abstract

fetched live from OpenAlex

BACKGROUND: Cutaneous T-cell lymphomas (CTCLs) are a heterogeneous group of non-Hodgkin lymphomas, with mycosis fungoides (MF) being the most common type, accounting for approximately 60% of all lymphomas arising primarily in the skin. Diagnosis of MF is challenging, especially in its early stages when the number of atypical T lymphocytes is small, and clinical and histopathological changes are often nonspecific. This leads to significant delays of 3-5 years in diagnosis and treatment. Thus, novel diagnostic methods are needed to adjust the diagnostic and therapeutic strategies of CTCL. Nonlinear optical microscopy (NLOM) is promising for its sensitivity to specific tissue structures through harmonic generation and its ability to image in three dimensions. OBJECTIVES: To image haematoxylin and eosin-stained skin samples with NLOM and detect atypical epidermotropism and dermal cells in MF skin samples using an artificial intelligence (AI) model. METHODS: We used brightfield microscopy and NLOM to analyse haematoxylin and eosin-stained biopsy samples from MF skin lesions. Expert clinicians labelled the images, which were used to train a convolutional neural network to recognize skin lymphocytes. The model was applied to independent testing datasets obtained from both imaging modalities to assess its performance in detecting characteristic features of skin T lymphocytes. Additionally, NLOM was performed on fresh, unstained biopsy samples to highlight its potential for in vivo skin imaging. RESULTS: NLOM successfully imaged epidermal and dermal structures in haematoxylin and eosin-stained MF tissue sections with subcellular resolution. The trained AI model detected lymphocyte epidermotropism and dermal infiltration in the images. Moreover, NLOM imaged fresh, unstained biopsies up to 400 µm deep through the epidermis to the dermis. CONCLUSIONS: We demonstrate that NLOM, combined with AI, can detect lymphocyte epidermotropism and dermal infiltration in haematoxylin and eosin-stained MF skin tissue. This approach offers dermatologists a powerful tool to improve the diagnosis and prognosis of MF, paving the way for more timely and precise therapeutic strategies. An author video to accompany this article is available online.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame machine prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.

metaresearch head score (Codex)0.000
metaresearch head score (Gemma)0.001
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Observational · Consensus signal: none
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.001
Threshold uncertainty score0.003

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0000.001
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0010.000
Science and technology studies0.0000.000
Scholarly communication0.0000.000
Open science0.0000.000
Research integrity0.0010.000
Insufficient payload (model declined to judge)0.0010.000

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.011
GPT teacher head0.291
Teacher spread0.280 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designObservational
Domainnot available
GenreEmpirical

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations0
Published2025
Admission routes1
Has abstractyes

Explore more

Same venueBritish Journal of DermatologySame topicCutaneous lymphoproliferative disorders researchFrench-language works237,207