Multi‐level temporal abstraction for medical scenario construction

Silvent, Anne-Sophie; Dojat, Michel; Garbay, Catherine

doi:10.1002/acs.855

Cited by 20 publications

(13 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In attempting to use propositional models for this task, her group needs a discrete representation of the real-valued metrological data, and they noted in a recent paper "We are currently using (Lin et al's SAX approach to creating discrete data from continuous data" (McGovern et al 2006). In Duchene and Garbay (2005), Duchene et al (2004), Silvent et al (2003Silvent et al ( , 2004, the authors use SAX and random projection to discover motifs in telemedicine time series. In Chen et al (2005) the authors convert palmprint to time series, then to SAX, then they do biometric recognition.…”

Section: Applications Of Saxmentioning

confidence: 99%

Experiencing SAX: a novel symbolic representation of time series

Lin

Keogh

et al. 2007

Data Min Knowl Disc

1,335

851

View full text Add to dashboard Cite

Many high level representations of time series have been proposed for data mining, including Fourier transforms, wavelets, eigenwaves, piecewise polynomial models, etc. Many researchers have also considered symbolic representations of time series, noting that such representations would potentiality allow researchers to avail of the wealth of data structures and algorithms from the text processing and bioinformatics communities. While many symbolic representations of time series have been introduced over the past decades, they all suffer from two fatal flaws. First, the dimensionality of the symbolic representation is the same as the original data, and virtually all data mining algorithms scale poorly with dimensionality. Second, although distance measures can be defined on the symbolic approaches, these distance measures have little correlation with distance measures defined on the original time series.In this work we formulate a new symbolic representation of time series. Our representation is unique in that it allows dimensionality/numerosity reduction, and it also allows distance measures to be defined on the symbolic approach that lower bound corresponding distance measures defined on the original series. As we shall demonstrate, this latter feature is particularly exciting because it allows one to run certain data mining algorithms on the efficiently manipulated symbolic representation, while producing identical results to the algorithms that operate on the original data. In particular, we will demonstrate the utility of our representation on various data mining tasks of clustering, classification, query by content, anomaly detection, motif discovery, and visualization.

show abstract

Section: Applications Of Saxmentioning

confidence: 99%

Experiencing SAX: a novel symbolic representation of time series

Lin

Keogh

et al. 2007

Data Min Knowl Disc

1,335

851

View full text Add to dashboard Cite

show abstract

“…In Step 1, clinical algorithms are developed based on retrospective clinical data streams with known outcomes. TA is performed on the data, transforming the data streams into a time stamped interval-based representation by extracting the most relevant data features [4], such as states, trends and temporal relationships; DM is also possible as a preprocessing step to TA [2,13]. The temporally abstracted data is then processed using predictive DM approaches with the goal of identifying early indicators for complications of interest.…”

Section: 2mentioning

confidence: 99%

Extending CRISP-DM to incorporate temporal data mining of multidimensional medical data streams: A neonatal intensive care unit case study

Catley

Smith

McGregor

et al. 2009

2009 22nd IEEE International Symposium on Computer-Based Medical Systems

View full text Add to dashboard Cite

Using a Neonatal Intensive Care Unit (NICU) case study, this work investigates the current CRoss Industry Standard Process for Data Mining (CRISP-DM) approach for modeling Intelligent Data Analysis (IDA)-based systems that perform temporal data mining (TDM). The case study highlights the need for an extended CRISP-DM approach when modeling clinical systems applying Data Mining (DM) and Temporal Abstraction (TA). As the number of such integrated TA/DM systems continues to grow, this limitation becomes significant and motivated our proposal of an extended CRISP-DM methodology to support TDM, known as CRISP-TDM. This approach supports clinical investigations on multi-dimensional time series data. This research paper has three key objectives: 1) Present a summary of the extended CRISP-TDM methodology; 2) Demonstrate the applicability of the proposed model to the NICU data, focusing on the challenges associated with multi-dimensional time series data; and 3) Describe the proposed IDA architecture for applying integrated TDM.

show abstract

“…Previous works [1,2] clearly establish that it is a difficult task due to the ambiguity of the data and the impossibility to directly interpret them. To solve this problem, they propose to take into account some contextual knowledge based on the idea that a person physiology is usually influenced by the environmental conditions and its activities.…”

Section: Introductionmentioning

confidence: 99%

“…In [2,1], authors relies on simple data model such as interval or linear models to infer the person's situations. More recently [4] proposes to extract some features from physiological data and to classify them using a risk criterion.…”

Section: Introductionmentioning

confidence: 99%

Probabilistic model definition for physiological state monitoring

Amate

Forbes

Fontecave-Jallon

et al. 2011

2011 IEEE Statistical Signal Processing Workshop (SSP)

Self Cite

View full text Add to dashboard Cite

Assessing the global situation of a person from physiological data is a well-known difficult problem. In previous work, we propose a system that does not produce a diagnosis but instead follows a set of hypotheses and decides of an alarming situation with this information. In this paper we focus on data processing part of the system taking into account the complexity and the ambiguity of the data. We propose a statistical approach with a global model based on Hidden Markov Model and we present data models that rely on classical physiological parameters and expert's knowledge. We then learn a model that depends on the person and its environment, and we define and compute confidence values to assess the plausibility of hypotheses.

show abstract

Multi‐level temporal abstraction for medical scenario construction

Cited by 20 publications

References 21 publications

Experiencing SAX: a novel symbolic representation of time series

Experiencing SAX: a novel symbolic representation of time series

Extending CRISP-DM to incorporate temporal data mining of multidimensional medical data streams: A neonatal intensive care unit case study

Probabilistic model definition for physiological state monitoring

Contact Info

Product

Resources

About