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

As the area of application of deep neural networks expands to areas requiringexpertise, e.g., in medicine and law, more exquisite annotation processes forexpert knowledge training are required. In particular, it is difficult toguarantee generalization performance in the clinical field in the case ofexpert knowledge training where opinions may differ even among experts onannotations. To raise the issue of the annotation generation process forexpertise training of CNNs, we verified the annotations for surgical phaserecognition of laparoscopic cholecystectomy and subtotal gastrectomy forgastric cancer. We produce calibrated annotations for the seven phases ofcholecystectomy by analyzing the discrepancies of previously annotated labelsand by discussing the criteria of surgical phases. For gastrectomy for gastriccancer has more complex twenty-one surgical phases, we generate consensusannotation by the revision process with five specialists. By training theCNN-based surgical phase recognition networks with revised annotations, weachieved improved generalization performance over models trained with originalannotation under the same cross-validation settings. We showed that theexpertise data annotation pipeline for deep neural networks should be morerigorous based on the type of problem to apply clinical field.

Rethinking Generalization Performance of Surgical Phase Recognition with Expert-Generated Annotations