English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

A medical AI system's generalizability describes the continuity of itsperformance acquired from varying geographic, historical, and methodologicsettings. Previous literature on this topic has mostly focused on "how" toachieve high generalizability with limited success. Instead, we aim tounderstand "when" the generalizability is achieved: Our study presents amedical AI system that could estimate its generalizability status for unseendata on-the-fly. We introduce a latent space mapping (LSM) approach utilizingFrechet distance loss to force the underlying training data distribution into amultivariate normal distribution. During the deployment, a given test data'sLSM distribution is processed to detect its deviation from the forceddistribution; hence, the AI system could predict its generalizability statusfor any previously unseen data set. If low model generalizability is detected,then the user is informed by a warning message. While the approach isapplicable for most classification deep neural networks, we demonstrate itsapplication to a brain metastases (BM) detector for T1-weightedcontrast-enhanced (T1c) 3D MRI. The BM detection model was trained using 175T1c studies acquired internally, and tested using (1) 42 internally and (2) 72externally acquired exams from the publicly distributed Brain Mets datasetprovided by the Stanford University School of Medicine. Generalizabilityscores, false positive (FP) rates, and sensitivities of the BM detector werecomputed for the test datasets. The model predicted its generalizability to below for 31% of the testing data, where it produced (1) ~13.5 FPs at 76.1% BMdetection sensitivity for the low and (2) ~10.5 FPs at 89.2% BM detectionsensitivity for the high generalizability groups respectively. The resultssuggest that the proposed formulation enables a model to predict itsgeneralizability for unseen data.

Prediction of Model Generalizability for Unseen Data: Methodology and  Case Study in Brain Metastases Detection in T1-Weighted Contrast-Enhanced 3D  MRI