Zapisane w:
| 1. autor: | |
|---|---|
| Format: | Recurso digital |
| Język: | angielski |
| Wydane: |
Zenodo
2025
|
| Hasła przedmiotowe: | |
| Dostęp online: | https://doi.org/10.5281/zenodo.15340552 |
| Etykiety: |
Dodaj etykietę
Nie ma etykietki, Dołącz pierwszą etykiete!
|
Spis treści:
- <p>In the manuscript "Assessing the credibility of quantitative information: a general framework" <span lang="EN-GB">we propose a general process called the 7S Framework, which can be used to assess the credibility of information, whether measured, inferred or predicted. The 7S Framework integrates and generalises the credibility assessment approaches used for measured information in metrology, inferred information in statistics, and predicted information in computational science and engineering. </span></p> <p><span lang="EN-GB">In this supplementary material, we apply the 7S framework to seven fairly different use cases to demonstrate that the proposed</span><span lang="EN-GB"> framework is effective, sufficiently general, and capable of capturing all the subtle differences that the concept of credibility implies for these different kinds of information. The seven use cases are:</span></p> <p>1. Measurement - Credibility of strain gauge measurements<br>2. Biophysical prediction - Credibility of CT-based prediction of human bone deformation<br>3. Biophysical prediction - Credibility of CT-based prediction of human bone strength<br>4. Biophysical prediction - Credibility of CT-based prediction of risk of hip fracture<br>5. Synthetic dataset – Virtual cohort for In Silico Trials of interventions to prevent hip fractures<br>6. In Silico Trial - Credibility of a model to predict interventions’ efficacy on a clinical cohort<br>7. Surrogate model - Credibility of an ML predictor to surrogate a biophysical model</p> <p><span lang="EN-GB">We propose the 7S Framework as a generalisation useful in the credibility assessments of complex <em>in silico</em> medicine scenarios such as in-silico augmented clinical trials, physics-informed machine learning predictors, or the use of synthetic datasets to overcome privacy limitations, train machine learning predictors, and run large-scale In Silico Trials.</span></p>