Health AI Central
Share:
Peer-Reviewed Publication
Comput Biol Med2024;180108906.September 1, 2024Journal Article

Non-relevant segment recognition via hard example mining under sparsely distributed events.

Bogyu Park1, Hyeongyu Chi2, Jihyun Lee3, Bokyung Park4, Jiwon Lee5, Soyeon Shin6, Woo Jin Hyung7, Min-Kook Choi8
1AI Research, Hutom, Dokmak-ro 279, Mapo-gu, 04151, Seoul, Republic of Korea. Electronic address: bgpark@hutom.io.
2AI Research, Hutom, Dokmak-ro 279, Mapo-gu, 04151, Seoul, Republic of Korea. Electronic address: hyeongyuc96@hutom.io.
3AI Research, Hutom, Dokmak-ro 279, Mapo-gu, 04151, Seoul, Republic of Korea. Electronic address: jihyun.lee@hutom.io.
4AI Research, Hutom, Dokmak-ro 279, Mapo-gu, 04151, Seoul, Republic of Korea. Electronic address: bokyung@hutom.io.
5AI Research, Hutom, Dokmak-ro 279, Mapo-gu, 04151, Seoul, Republic of Korea. Electronic address: jiwon@hutom.io.
6AI Research, Hutom, Dokmak-ro 279, Mapo-gu, 04151, Seoul, Republic of Korea. Electronic address: soyeon.shin@hutom.io.
7AI Research, Hutom, Dokmak-ro 279, Mapo-gu, 04151, Seoul, Republic of Korea; Yonsei University College of Medicine, Yonsei-ro 50, Seodaemun-gu, 03722, Seoul, Republic of Korea. Electronic address: wjhyung@hutom.io.
8AI Research, Hutom, Dokmak-ro 279, Mapo-gu, 04151, Seoul, Republic of Korea. Electronic address: mkchoi@hutom.io.
PubMed DOI

Abstract

We propose on/offline hard example mining (HEM) techniques to alleviate the degradation of the generalization performance in the sparse distribution of events in non-relevant segment (NRS) recognition and to examine their utility for long-duration surgery. Through on/offline HEM, higher recognition performance can be achieved by extracting hard examples that help train NRS events, for a given trai…

Create a free account to keep reading

Free members get 10 full research views every month across publications, clinical trials, FDA clearances, adverse events, and NIH grants. No credit card required.

Create Free AccountSign In

Want unlimited research access? See Pro plans

Data Accuracy Notice: Research intelligence on Health AI Central is aggregated from public sources (PubMed, ClinicalTrials.gov, FDA, NIH, CMS, and others) and refreshed nightly. Classifications and derived metrics are produced by automated methods described in our Methodology. We recommend verifying critical data points against the primary sources before making decisions.