Paper · 42823116

Diagnostic accuracy of a DenseNet-121 deep learning algorithm for chest radiograph triage in health assessment applicants: a prospective shadow-mode validation study in Nepal.

Shrestha L and 2 others. BMJ Open, 2026.

11 sourcespapers-2026-11-12-0.1.0

Cite this record

ConductScience Library. Diagnostic accuracy of a DenseNet-121 deep learning algorithm for chest radiograph triage in health assessment applicants: a prospective shadow-mode validation study in Nepal. Papers, record 42823116. Release papers-2026-11-12-0.1.0, 12 Nov 2026. https://conductscience.org/library/papers/42823116

On this page

Paper

PubMed's record of the paper, with iCite's citation figures. No abstract is shown.

Title: Diagnostic accuracy of a DenseNet-121 deep learning algorithm for chest radiograph triage in health assessment applicants: a prospective shadow-mode validation study in Nepal.
Journal: BMJ Open
Year: 2026
Publication types: Journal Article, Validation Study
PubMed: https://pubmed.ncbi.nlm.nih.gov/42823116/
DOI: https://doi.org/10.1136/bmjopen-2026-124868
PMC: https://pmc.ncbi.nlm.nih.gov/articles/PMC13630079/
The journal: https://conductscience.org/library/journal-index/101552874
Revised in PubMed: 2026-10-03

Source PubMed

Authors3

The authors in the order PubMed's record lists them, with the ORCID iDs it states.

Order	Author	Forename	ORCID iD
1	Shrestha L	Lochan	https://orcid.org/0000-0003-0814-3332
2	Maharjan D	Dinesh	https://orcid.org/0000-0001-6177-7233
3	Bista U	Uttam	https://orcid.org/0009-0001-4898-5748

Source PubMed

Topics14

NLM's MeSH indexing of the paper, its major topics first.

MeSH term	Major
Triage	Major topic
Deep Learning	Major topic
Radiography, Thoracic	Major topic
Humans
Prospective Studies
Nepal
Algorithms
Convolutional Neural Networks
Sensitivity and Specificity
Male
Female
ROC Curve
Reproducibility of Results
Adult

Source PubMed

Grants0

Each grant PubMed's record writes, matched to the NIH project NIH's grant and paper links name for this paper, and each of those links the record does not write.

None in this release.

Trials0

The trials whose ClinicalTrials.gov record lists the paper, and the trial numbers the paper's PubMed record names.

None in this release.

Comments and corrections0

PubMed's comments and corrections, NLM's own indexing: retractions first, then expressions of concern, errata and updates, and comments last.

None in this release.

References8

The papers the reference list in PubMed's record names, those in our copy of PubMed, in the list's order.

Order	PMID	Title	Journal	Year	In Papers	Source
1	2,348,207	High agreement but low kappa: I. The problems of two paradoxes.	J Clin Epidemiol	1990	https://conductscience.org/library/papers/2348207	PubMed
2	33,526,932	DECIDE-AI: new reporting guidelines to bridge the development-to-implementation gap in clinical artificial intelligence.	Nat Med	2021	https://conductscience.org/library/papers/33526932	PubMed
3	42,823,116	Diagnostic accuracy of a DenseNet-121 deep learning algorithm for chest radiograph triage in health assessment applicants: a prospective shadow-mode validation study in Nepal.	BMJ Open	2026	https://conductscience.org/library/papers/42823116	PubMed
4	3,203,132	Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach.	Biometrics	1988	https://conductscience.org/library/papers/3203132	PubMed
5	30,399,157	Variable generalization performance of a deep learning model to detect pneumonia in chest radiographs: A cross-sectional study.	PLoS Med	2018	https://conductscience.org/library/papers/30399157	PubMed
6	8,870,764	Statistical methodology: I. Incorporating the prevalence of disease into the sample size calculation for sensitivity and specificity.	Acad Emerg Med	1996	https://conductscience.org/library/papers/8870764	PubMed
7	34,183,345	Developing a reporting guideline for artificial intelligence-centred diagnostic test accuracy studies: the STARD-AI protocol.	BMJ Open	2021	https://conductscience.org/library/papers/34183345	PubMed
8	26,511,519	STARD 2015: an updated list of essential items for reporting diagnostic accuracy studies.	BMJ	2015	https://conductscience.org/library/papers/26511519	PubMed

Cited by1

The papers in our copy of PubMed whose reference list names this one, newest first.

PMID	Title	Journal	Year	In Papers
42,823,116	Diagnostic accuracy of a DenseNet-121 deep learning algorithm for chest radiograph triage in health assessment applicants: a prospective shadow-mode validation study in Nepal.	BMJ Open	2026	https://conductscience.org/library/papers/42823116

Source PubMed

Genes0

The genes the Gene search release ties to the paper: Swiss-Prot's references first, then PubTator's text mining.

None in this release.

Drugs0

The drugs whose trials, in the Pipelines release, list the paper.

None in this release.

Data0

The data bank entries PubMed's record names. Trial numbers are under Trials.

None in this release.

Data dates

The date of each source this release read.

PubMed: 2026-11-06; PubMed
iCite: 2026-11-05; iCite
PMC's ID file: 2026-11-05; PMC's ID file
NIH RePORTER's grant and paper links: 2026-11-01; NIH RePORTER's grant and paper links
Trials: 2026-11-12
Gene search: 2026-11-11
Pipelines: 2026-09-18

Sources 11

ChEMBL, through Open TargetsCC BY-SA 3.0
ClinicalTrials.govNo license stated
HGNCCC0
iCiteCC BY 4.0
NCBI GenePublic domain (US government)
NIH RePORTER's grant and paper linksPublic domain
Open TargetsCC0 1.0
PMC's ID filePublic domain (US government)
PubMedNLM's terms
PubTator 3Public domain
UniProtCC BY 4.0

Each source's terms

ChEMBL, through Open TargetsCC BY-SA 3.0, for the parts from ChEMBL. Read on 8 Nov 2026 at https://platform-docs.opentargets.org/licence
ClinicalTrials.govnone stated
HGNCCC0: https://www.genenames.org/about/license/
iCiteCC BY 4.0
NCBI GeneNone named for the file. NCBI says information created by or for the US government on its site is in the public domain, and asks that NLM be acknowledged: https://www.ncbi.nlm.nih.gov/home/about/policies/
NIH RePORTER's grant and paper linksPublic domain (NIH RePORTER)
Open TargetsCC0 1.0. Open Targets asks that users cite its latest paper: Buniello et al. 2025, Open Targets Platform: facilitating therapeutic hypotheses building in drug discovery, Nucleic Acids Research 53(D1):D1467, PMID 39657122. Read on 8 Nov 2026 at https://platform-docs.opentargets.org/licence
PMC's ID fileNone named for the file. NCBI says information created by or for the US government on its site is in the public domain, and asks that NLM be acknowledged: https://www.ncbi.nlm.nih.gov/home/about/policies/
PubMedNLM's terms
PubTator 3Public domain
UniProtCC BY 4.0. UniProt's licence file: "All databases and documents in the UniProt FTP directory may be copied and redistributed freely, without advance permission, provided that this copyright statement is reproduced with each copy." UniProt asks that users cite its latest paper: The UniProt Consortium, UniProt: the Universal Protein Knowledgebase in 2025, Nucleic Acids Research 53:D609-D617, https://doi.org/10.1093/nar/gkae1010. Read on 9 Nov 2026 at https://www.uniprot.org/help/license

ClinicalTrials.gov's data carry an international copyright outside the United States: https://clinicaltrials.gov/about-site/terms-conditions.

For software

Read this paper from software.

The record and each of its lists are JSON, twenty rows a page. No key.

GET https://conductscience.org/api/papers/v1/42823116

GET https://conductscience.org/api/papers/v1/42823116/authors

API reference For AI assistants