nabu.science
MethodologyAboutContact

 

Reliability concern: This paper was retracted on 2023-06-28. See notice: 10.1155/2023/9867187.

  • Snapshot
  • Key Claims
  • Strengths & Limitations
  • Commentary
  • From the Authors
  • Author Note / Field Feedback
  • Dimension breakdown
  • Impact Profile
Free · About 5 minutes
Report #NBU-81B36E9C·Framework v5.0·Pipeline v0.4·Evaluated 01 Jun 2026·OECD: 3.2 Clinical medicine·Subfield: Pulmonary and Respiratory Medicine·Basis: Full text·Type: Cross-sectional study

The Diagnosis of Early Gastric Cancer Based on Medical Imaging Technology and Mathematical Modeling.

Xiu et al.Xiu, Ma, Ding et al.·2022·Computational and mathematical methods in medicine·10.1155/2022/8721654 ↗

In brief

A clinically relevant BLI-versus-NBI comparison plus a small ML add-on, constrained by unresolved internal reporting inconsistencies that block verification of key accuracy claims.

ReliabilityConcerns4 of 4 checks completeRead the 3 entries ↓
Abstract

The key to reducing the mortality of gastric cancer is early detection, early diagnosis, and early treatment of gastric cancer. Early diagnosis of gastric cancer is the key to early detection and diagnosis of gastric cancer. Early diagnosis and treatment of gastric cancer is of great significance for improving the curative effect and reducing mortality of gastric cancer. The purpose of this paper is to study the diagnosis of early gastric cancer based on medical imaging techniques and mathematical modeling. The effect of W-DeepLab network-assisted diagnosis of images under white light was analyzed, and the value of Narrow Band Imaging and Blue Laser Imaging in the diagnosis of early gastric cancer was compared. Because Blue Laser Imaging endoscopy can clearly observe the demarcation line and microvascular morphology; but when using Narrow Band Imaging observation, part of the demarcation line and microvascular morphology is not observed. The results show that Blue Laser Imaging is brighter than Narrow Band Imaging's endoscopic images, and it is easier to observe the microstructure of lesions under endoscopy, so as to accurately determine the nature of lesions.

Strengths · Limitations

Strengths

#methodological rigour

Large clinically sourced cohort

The clinical comparison reports 4,040 patients undergoing endoscopic assessment with pathology used as the reference standard, which is a potentially informative real-world dataset if fully and consistently reported.

↳ Methods §3.1; Methods §3.2

#contribution

Direct comparison of BLI vs NBI

The study directly contrasts BLI and NBI diagnostic performance and reports a statistically significant difference in sensitivity, which targets a concrete diagnostic objective.

↳ Results §4.3

#positioning

Limitations are explicitly acknowledged

The conclusions section notes single-center retrospective design and variability in physician experience, which are central considerations for interpreting diagnostic accuracy results.

↳ Conclusions §5 (limitations paragraph)

Limitations

#reporting

Baseline demographics reported inconsistently

The control-group gender counts sum to 100 despite a stated n=1864, undermining confidence in the reported baseline characteristics and related comparability claims.

↳ Results §4.1

#methodological rigour

Diagnostic counts not auditable

The NBI/BLI early gastric cancer counts in text are internally unclear, preventing reconstruction of 2×2 tables needed to verify sensitivity and specificity comparisons.

↳ Results §4.3 (Table 4/Figure 4 referenced)

#contribution

ML target mismatched to main claim

W-DeepLab is presented as diagnosing intestinal metaplasia on only 100 raw white-light images, but the paper’s main framing is early gastric cancer diagnosis and the linkage between these components is not established.

↳ Methods §3.3; Results §4.2; Abstract

Does this match your read? Add your assessment ↓

Withdrawn — shown for reference only

Nabu Commentary

The paper’s main clinical result hinges on diagnostic accuracy comparisons between BLI and NBI, but the Results section contains inconsistencies that block straightforward verification, including an impossible gender breakdown in §4.1 and unclear EGC diagnostic counts in §4.3. The methods describe quasi-allocation by appointment-order odd/even numbering without establishing baseline equivalence beyond gender or addressing operator effects, which increases the risk that modality differences reflect confounding rather than imaging performance (§3.1; §5). The ML component reports strong metrics for intestinal metaplasia on a dataset stated to have only 100 raw images before augmentation, with limited reporting of evaluation safeguards, and it is not integrated into the EGC comparison narrative (§3.3–§4.2). These features support moderate-to-low quality and translation scores even though the clinical problem itself is high-need.

Evaluate another paper

Free · About 5 minutes · Research preview

Author Note / Field Feedback

Nabu’s assessment, alongside the field’s view.

Are you an author of this paper?

Dimension breakdown

Evidence Score

Out of 5
012345
2.0±0.2

Contribution
2.2

The paper addresses an important diagnostic problem and reports a large single-center cohort for BLI vs NBI, but the ML component targets intestinal metaplasia rather than early gastric cancer and is not integrated with the clinical comparison. Over-broad conclusions about BLI superiority exceed what is supported by the reported sensitivity/specificity contrast.

Abstract; Methods §3.1 and §3.3; Results §4.2–§4.3; Conclusions §5

"This paper proposes the W-DeepLab network for computer-aided diagnosis of intestinal metaplasia"

Methodological Rigour
1.8

Group assignment by appointment-order odd/even numbering is not true randomization and is not paired with adequate baseline comparability or confounder control beyond gender. The diagnostic-accuracy reporting in §4.3 is not reconstructable from the text, and the ML evaluation is based on only 100 raw images with limited split/reporting detail.

Methods §3.1–§3.3; Results §4.1–§4.3

"have only 100 valid raw data before augmentation"

Reporting
2.3

While the paper follows a standard IMRAD-like structure, key result statements are contradictory or ambiguous, particularly in demographics and diagnostic counts. The narrative also shifts between early gastric cancer diagnosis and intestinal metaplasia segmentation without clearly stating how the parts relate.

Results §4.1–§4.3; Abstract; Methods §3.3; Results §4.2

"There were 38 men and 62 women"

Positioning
2.0

Background citations motivate early detection and describe BLI/NBI and VS classification, but the discussion does not clearly position the BLI-vs-NBI finding against closely comparable prior evidence. Limitations are listed but not used to qualify the strength or generalisability of the diagnostic-accuracy claims.

Introduction §1; Background §2.1; Conclusions §5

Reliability

Concerns4 of 4 checks


Internal CoherenceConcern

Multiple numerical and scope inconsistencies make key results difficult to reconcile, including baseline demographics and diagnostic-count reporting for the NBI/BLI comparison, plus a mismatch between early gastric cancer claims and intestinal metaplasia ML outcomes.

Research ConductNoted

The paper describes consent and a retrospective single-center study but does not provide verifiable ethics/IRB details, and its data-availability statement does not correspond to accessible underlying data.

Flags: 2 declared / 5 total

Reference IntegrityClean

All 18 cited references resolved to real publications via PubMed, Crossref, or OpenAlex.

References not verified automatically are resolved manually within 24 hours.

Post-Publication RecordConcern

This paper was retracted on 2023-06-28. See notice: 10.1155/2023/9867187.

Last checked: 17 Oct 2026

Sources: Crossref (incl. Retraction Watch) ✓

Impact Profile

Impact areaHealth · Cancer and neoplasms · Stomach cancer
Type of advanceEmpirical
Route to useSuggested
AdvanceExtends
BeneficiariesLarge
CitationsFWCI 0.3
Policy · patentNot yet traced
ReachedNot reachedNot assessable or unverified
ReadinessPiloted
Foundational
Controlled
Piloted
Validated
Implementable
FoundationalImplementable

The comparison of NBI and BLI was shown working once in a single-center retrospective study of 4040 patients, and the W-DeepLab network was tested only on an internal dataset, with the authors themselves noting further multi-case verification is required.

From the paper
“in order to provide some theoretical basis for the diagnosis of early gastric cancer”

The paper names a real-world use, improving endoscopic diagnosis of early gastric cancer, but does not specify who would act on the findings or in what decision setting.

Pathway

Gastroenterologists and endoscopy services could adopt BLI endoscopy and W-DeepLab computer-aided image analysis to improve early gastric cancer detection during diagnostic gastroscopy.

Barrier

The single-center retrospective design, small training image dataset, and lack of multicenter or prospective validation stand between these findings and routine clinical adoption.

Evaluated 13 Nov 2026 · GBD 2023 · glm_v1_3_3 · impact_profile_v1_3_3

Report an issue →

AI-generated, human-governed. Something look off? Contact us to request a review.

Keep reading

Related evaluations

Other papers Nabu has evaluated blind, starting with the same field.

  1. Implications of Cranial Arterial Stenosis and Dolichoectasia for Cerebral Small-Vessel Disease Etiopathogenesis: Findings From a Prospective Mild Stroke Cohort

    Han, Clancy, Arteaga-Reyes, et al. · 2026 · Clinical medicine

    Primary claim

    These findings support a nonatheromatous, intrinsic microvascular pathology, particularly segmental arteriolar disorganization, as the principal mechanism of lacunar stroke and cSVD.

    Tentative
    Evidence Score
    3.4±0.1
    Reliability
    Caveats
  2. A generative AI-discovered TNIK inhibitor for idiopathic pulmonary fibrosis: a randomized phase 2a trial

    Xu, Ren, Wang, et al. · 2025 · Basic medicine

    Primary claim

    We observed increased forced vital capacity at the highest dosage with a mean change of +98.4 ml (95% confidence interval 10.9 to 185.9) for patients in the 60 mg rentosertib QD group

    Tentative
    Evidence Score
    3.4±0.2
    Reliability
    Concerns
  3. C9ORF72 repeat expansion causes vulnerability of motor neurons to Ca2+-permeable AMPA receptor-mediated excitotoxicity

    Selvaraj, Livesey, Zhao, et al. · 2018 · Basic medicine

    Primary claim

    increased expression of GluA1 AMPA receptor (AMPAR) subunit occurs in MNs with C9ORF72 mutations that leads to increased Ca2+-permeable AMPAR expression and results in enhanced selective MN vulnerability to excitotoxicity.

    Tentative
    Evidence Score
    3.1±0.2
    Reliability
    Caveats

How the rubric works →