MétaCan
Menu
Back to cohort
Record W7163753538 · doi:10.2196/86278

Intra-System Repeatability of S-Detect for Breast Ultrasound Classification on Identical Static Images: A Single-Center Retrospective Repeatability Study (Preprint)

2025· article· en· W7163753538 on OpenAlexvenueno aff
Liang Yongping, Ping Zhou, Yang Wang, Nan Zhang (46264), Qing Zhou, Xinghao Zhang, Haifeng Cai, Juan Zhang (48597)

Bibliographic record

VenueJMIR Medical Informatics · 2025
Typearticle
Languageen
FieldComputer Science
TopicAI in cancer detection
Canadian institutionsnot available
Fundersnot available
KeywordsRepeatabilityUltrasoundRetrospective cohort studyBreast MRIReproducibility

Abstract

fetched live from OpenAlex

Background: Computer-aided diagnostic systems such as S-Detect (Samsung Medison) are increasingly integrated into breast ultrasound workflows. Notwithstanding extensive past evaluation of S-Detect's diagnostic accuracy, its intrasystem repeatability at the software level with identical static images, a fundamental prerequisite for clinical reliability, has not been systematically investigated. Objective: This study aimed to evaluate the intrasystem repeatability of the S-Detect computer-aided diagnostic system in classifying breast nodules in identical static ultrasound images. Methods: This retrospective, registered, blinded repeatability study analyzed 398 breast nodules from 261 women (mean age 43.10, SD 12.57 years) who underwent surgery between February 2019 and March 2020 at a single institution. Identical stored static ultrasound images, acquired by a single experienced sonographer on 1 Samsung RS80A ultrasound system, were each analyzed twice using the same S-Detect workstation: immediately after acquisition (S-Detect 1) and again at least 4 weeks later under blinded conditions with manual cursor repositioning (S-Detect 2). Repeatability was assessed using concordance rate and Cohen κ. The diagnostic performance of each run was compared against surgical histopathology. Results: Of 398 nodules, 156 (39.2%) were initially classified as possibly benign, and 242 (60.8%) were initially classified as possibly malignant. On repeat analysis, 37.4% (149/398) and 62.6% (249/398) of the nodules were classified as possibly benign and malignant, respectively. A total of 4.5% (7/156) of the nodules initially classified as benign were reclassified as malignant, whereas no malignant-to-benign changes occurred. The overall concordance rate was 98.2% (391/398), with a Cohen κ of 0.95 (95% CI 0.94-0.99; P<.001). Diagnostic performance remained stable across runs (area under the curve=0.913 vs 0.902; P=0.702). Conclusions: Under controlled conditions with identical static images, S-Detect showed high intrasystem repeatability, underscoring strong software-level consistency, although its translation to real-world clinical reproducibility requires further validation.

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

Direct model labels (unvalidated)

Per-model category and study-design labels from the labeling rounds. They are machine output, unvalidated, and the disagreement between models ships as data. No study design here is MEDLINE-validated yet.

Model armCategoriesStudy designConfidence
gemmano category
Domain: not available · Genre: Empirical
About the Canadian research system: no · About a Canadian topic: no
Observationallow
gptno category
Domain: not available · Genre: Empirical
About the Canadian research system: no · About a Canadian topic: no
Observationallow
models agreeAgreement compares identical category sets and study designs across arms.

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)0.008
metaresearch head score (Gemma)0.021
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: Observational
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.008
Threshold uncertainty score0.040

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0080.021
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0010.001
Bibliometrics0.0010.001
Science and technology studies0.0010.001
Scholarly communication0.0010.001
Open science0.0010.001
Research integrity0.0010.000
Insufficient payload (model declined to judge)0.0010.001

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 head0.019
GPT teacher head0.305
Teacher spread0.286 · 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

Labeled directly by 2 models reading the full record.

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 abstractno

Explore more

Same venueJMIR Medical InformaticsSame topicAI in cancer detectionFrench-language works237,207