ArticleNPJ digital medicine2024
Shortcut learning in medical AI hinders generalization: method for estimating AI model generalization without external data.
Article in NPJ digital medicine, 2024. The graph could read no effect estimate from its abstract, so it casts no vote on the map. Cited by 50 papers.
What it found
Each row is one number read from the abstract, on the scale the paper reported it, with its interval. Left of the dashed line favours the treatment, right favours the comparator. Under each row is the sentence it came from. New to these charts? A ten-minute tutorial.
The abstract states no effect estimate the extractor could read, or names no intervention and outcome on the map, so this paper lights no cell and moves no belief. It is still indexed, cited and linked below.
The trial behind it
Trials whose registry record cites this paper, or whose number appears in the abstract. A trial that started after this paper was published is citing it as background, not reporting it.
Neither the registry nor the abstract names a trial number. If this is a trial report, that itself is worth knowing.
Who cites it
50 citing papers in PubMed.
- Comment on "Single-slice CT body-weight prediction with image and metadata fusion: a multi-model comparison".Radiological physics and technology · 2026Article
- Artificial intelligence and radiomics in bladder cancer MRI: a scoping review of applications, performance, and barriers to clinical translation.Abdominal radiology (New York) · 2026Article
- Multicohort Evaluation of Acquisition Confounding in Single-Site Lung-CT Malignancy Classification with Frozen Foundation-Model Features.Journal of imaging informatics in medicine · 2026Article
- Deep learning for developmental dysplasia of the hip: comparative classification strategies for ultrasound imaging with heterogeneous image quality.Physical and engineering sciences in medicine · 2026Article
- The cult of the external validation set: undertraining is only half of the radiomics validation crisis.European radiology · 2026Article
- Fully automated deep learning for the identification and classification of dental implant brands from multimodal radiographic images.Clinical oral investigations · 2026Article
- Large language models for lumbar spondylolisthesis detection: a multi-center pilot comparative radiographic accuracy study.European journal of orthopaedic surgery & traumatology : orthopedie traumatologie · 2026Article
- PatchChestCT: A patch-level spatial annotation dataset for nine abnormalities in chest CT.Scientific data · 2026Article
- Wearable Flexible Sensors for Cardiovascular Disease Monitoring.Advanced materials (Deerfield Beach, Fla.) · 2026Review
- Toward universal representations of the living: Physiological invariance for transportable medical AI.PLOS digital health · 2026Review
- Zero-shot detection of apical periodontitis on periapical radiographs using multimodal large language models: diagnostic accuracy and error patterns.BMC oral health · 2026Article
- Self-supervised representation learning reveals explainable physiological structure in high-dimensional magnetocardiography.NPJ digital medicine · 2026Article
- Automated diagnosis of usual interstitial pneumonia on chest CT via the mean curvature of isophotes.European journal of radiology open · 2026Article
- Detecting dataset bias in medical AI using a generalized and modality agnostic auditing approach.NPJ digital medicine · 2026Article
- External validation improves generalizability, replicability and reproducibility in predictive models for neuroimaging.Nature methods · 2026Review
- Artificial intelligence across the cardiovascular diagnostic pathway: a case-based narrative review.Cardiovascular diagnosis and therapy · 2026Review
- Deep Learning and Cardiovascular Diseases: An Updated Narrative Review.Journal of clinical medicine · 2026Review
- Precision oncology in the age of AI: lessons from AI-driven drug discovery and clinical translation.BJC reports · 2026Review
- A foundation generative model for breast ultrasound image analysis.Nature biomedical engineering · 2026Article
- RoentMod: a synthetic chest X-ray modification model to identify and correct image interpretation model shortcuts.NPJ digital medicine · 2026Article
Corrections and comments
PubMed lists nothing against this paper. Absence here is not a guarantee, only a check that was made.
Authors and funding
11 authors.
Funding
No grant is acknowledged in the PubMed record.
Abstract
Healthcare datasets are becoming larger and more complex, necessitating the development of accurate and generalizable AI models for medical applications. Unstructured datasets, including medical imaging, electrocardiograms, and natural language data, are gaining attention with advancements in deep convolutional neural networks and large language models. However, estimating the generalizability of these models to new healthcare settings without extensive validation on external data remains challenging. In experiments across 13 datasets including X-rays, CTs, ECGs, clinical discharge summaries, and lung auscultation data, our results demonstrate that model performance is frequently overestimated by up to 20% on average due to shortcut learning of hidden data acquisition biases (DAB). Shortcut learning refers to a phenomenon in which an AI model learns to solve a task based on spurious correlations present in the data as opposed to features directly related to the task itself. We propose an open source, bias-corrected external accuracy estimate, P
Identifiers
What OpenQuestion holds
Registered trials
Read under generation 80e0d062 · epoch 390. Bibliography from PubMed, PubMed Central and OpenAlex; grants from NIH RePORTER; trial links from ClinicalTrials.gov; estimates, votes and beliefs from the OpenQuestion graph.