CRISP-TDM<inf>0</inf> for standardized knowledge discovery from physiological data streams: Retinopathy of prematurity and blood oxygen saturation case study
- Publication Type:
- Conference Proceeding
- Citation:
- 2017 IEEE Life Sciences Conference, LSC 2017, 2018, 2018-January pp. 226 - 229
- Issue Date:
- 2018-01-23
Closed Access
| Filename | Description | Size | |||
|---|---|---|---|---|---|
| 08268184.pdf | Published version | 956.73 kB |
Copyright Clearance Process
- Recently Added
- In Progress
- Closed Access
This item is closed access and not available.
© 2017 IEEE. The CRoss Industry Standard Process for Temporal Data Mining (CRISP-TDM) that supports physiological stream temporal data mining and CRISP-DM0 that supports null hypothesis driven confirmatory data mining in combination was proposed by prior research. This combined CRISP-TDM0 is utilised as the standardised approach to managing, reporting and performing retrospective clinical research and is designed to solve the limitation in knowledge discovery amongst physiological data streams [1]. The temporal abstractions (TA) of high fidelity blood oxygenation saturation (SpO2) levels of nine premature neonates are analysed using data collected by the Artemis Platform that complies with the Big Data concept [2] and correlated with Retinopathy of Prematurity (ROP) data. The hourly SpO2, TA pattern visualisation manifested three clusters and this is further supported by mathematical review of time percentage spent in target, below and over oxygenation. Clustering based on ROP stage and gestational age identified probable association within these three clusters. However known risk factors showed no association with ROP.
Please use this identifier to cite or link to this item: