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

Successful deployment of machine learning algorithms in healthcare requirescareful assessments of their performance and safety. To date, the FDA approveslocked algorithms prior to marketing and requires future updates to undergoseparate premarket reviews. However, this negates a key feature of machinelearning--the ability to learn from a growing dataset and improve over time.This paper frames the design of an approval policy, which we refer to as anautomatic algorithmic change protocol (aACP), as an online hypothesis testingproblem. As this process has obvious analogy with noninferiority testing of newdrugs, we investigate how repeated testing and adoption of modifications mightlead to gradual deterioration in prediction accuracy, also known as``biocreep'' in the drug development literature. We consider simple policiesthat one might consider but do not necessarily offer any error-rate guarantees,as well as policies that do provide error-rate control. For the latter, wedefine two online error-rates appropriate for this context: Bad Approval Count(BAC) and Bad Approval and Benchmark Ratios (BABR). We control these rates inthe simple setting of a constant population and data source using policiesaACP-BAC and aACP-BABR, which combine alpha-investing, group-sequential, andgate-keeping methods. In simulation studies, bio-creep regularly occurred whenusing policies with no error-rate guarantees, whereas aACP-BAC and -BABRcontrolled the rate of bio-creep without substantially impacting our ability toapprove beneficial modifications.

Approval policies for modifications to Machine Learning-Based Software as a Medical Device: A study of bio-creep