AI Science Tracker
EN

Peking Union Medical College Hospital

2 discoveries from 2025 to 2026

Discoveries

Prognostic prediction of pulmonary neuroendocrine carcinoma using a transformer-based artificial intelligence model: a two-center retrospective cohort study.

Source check: Source checked. Not independent scientific confirmation.

Finding: A two-center retrospective cohort compared a transformer-based model with conventional approaches for overall and disease-free survival prediction. The best-performing model differed by endpoint, so performance should be described by task rather than as one universal winner.

Technology: Transformer-based survival prediction model

Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.

AI contribution
Transformer-based survival prediction model
Test and result
Two-center retrospective cohort evaluation reported model-specific discrimination for overall and disease-free survival.
Limitation / still unknown
Retrospective cohorts do not establish prospective clinical benefit; the strongest model varied by endpoint.
Evidence sources and annotation context

Evidence context (descriptions follow the cited sources; test type is a catalog classification):

Evidence sources: https://doi.org/10.3389/fonc.2026.1948262

Source checked · editorial details

The source page was checked against this entry. Independent confirmation is not assessed by this label.

Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.

Review date: 22 Nov 2026. Review scope: Primary result-bearing material for the stated method, result and evaluation, with bibliographic version information..

No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.

Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.

Radiomic models distinguish pathology-defined pancreatic cyst risk in surgical cohorts

Source check: Source checked. Not independent scientific confirmation.

Finding: Learned CT-radiomic scores combined with a conventional imaging feature distinguished source-defined benign and malignant pancreatic cyst groups in surgical cohorts. Models developed using 262 patients were evaluated in two 50-patient test sets and a separately recruited 34-patient cohort. The result is retrospective pathology-reference diagnostic evaluation, not evidence that the models improve screening or decisions about surgery.

Technology: LASSO feature selection; multiple learned radiomic classifiers; clinical-radiomic regression

Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.

AI contribution
Learned classifiers map selected CT image features to pathology-defined cyst labels; resulting radiomic scores add information to a clinical imaging-feature model evaluated in separate patient datasets.
Test and result
Eleven radiomic classifiers were trained in 262 patients, assessed in 50 temporal internal and 50 external cases, and combined with solid-component information. A34-patient prospectively recruited cohort was subsequently analyzed with those models.
Limitation / still unknown
All included patients underwent surgery, restricting the disease spectrum and reference availability. Prospective recruitment does not establish that a frozen model was evaluated before outcomes; validation is classified as retrospective. Full preprocessing/feature-selection nesting is not completely specified. The source-defined grouping combines different lesion types/grades and is not a universal management threshold. Separate fluid-omics mechanism claims are outside this record.
Evidence sources and annotation context

Evidence context (descriptions follow the cited sources; test type is a catalog classification):

Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC12120750/

Source checked · editorial details

The source page was checked against this entry. Independent confirmation is not assessed by this label.

Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.

Review date: 21 Nov 2026. Review scope: Exact primary XML front DOI/title/PMCID/date; clinical cohorts, pathology-label definitions, radiomic/statistical model and validation Methods/Results. Raw biological assay protocols and mechanistic omics sections not processed. Same-day primary bibliographic/notice snapshot reused..

No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.

Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.

Open the interactive profile for the timeline, or browse all organizations and model families.