The Temple University Hospital EEG corpus

Harati, Amir; Choi, Soon-Wook; Tabrizi, M.H.N.; Obeid, Iyad; Jacobson, Mercedes P.

doi:10.1109/globalsip.2013.6736803

Cited by 19 publications

(11 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…We measured the performance of L-PCR system for neurology-focused cohorts by considering EEG reports from the TUH EEG Corpus 28 using the relevance judgments described in Datasets and Experimental Settings . Table 3(a) presents the performance of the L-PCR system compared to (1) a BM25 baseline and (2) a text-only variant of MERCuRY 4 (a multi-modal retrieval system incorporating polarity information and the BM25 ranking function) using cross validation at the cohort description level (e.g.…”

Section: Resultsmentioning

confidence: 99%

“…For the neurology-specific setting, we relied on the publicly-available collection of EEG reports from the Temple University Hospital (TUH) EEG Corpus. 28 , 29 It contains EEG reports collected over 25 000 sessions for 15 000 patients over 12 years. While the TUH EEG corpus contains EEG signal information as well as EEG reports, we considered only the EEG reports.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Learning relevance models for patient cohort retrieval

Goodwin

Harabagiu

2018

JAMIA Open

View full text Add to dashboard Cite

ObjectiveWe explored how judgements provided by physicians can be used to learn relevance models that enhance the quality of patient cohorts retrieved from Electronic Health Records (EHRs) collections.MethodsA very large number of features were extracted from patient cohort descriptions as well as EHR collections. The features were used to investigate retrieving (1) neurology-specific patient cohorts from the de-identified Temple University Hospital electroencephalography (EEG) Corpus as well as (2) the more general cohorts evaluated in the TREC Medical Records Track (TRECMed) from the de-identified hospital records provided by the University of Pittsburgh Medical Center. The features informed a learning relevance model (LRM) that took advantage of relevance judgements provided by physicians. The LRM implements a pairwise learning-to-rank framework, which enables our learning patient cohort retrieval (L-PCR) system to learn from physicians’ feedback.Results and DiscussionWe evaluated the L-PCR system against state-of-the-art traditional patient cohort retrieval systems, and observed a 27% improvement when operating on EEGs and a 53% improvement when operating on TRECMed EHRs, showing the promise of the L-PCR system. We also performed extensive feature analyses to reveal the most effective strategies for representing cohort descriptions as queries, encoding EHRs, and measuring cohort relevance.ConclusionThe L-PCR system has significant promise for reliably retrieving patient cohorts from EHRs in multiple settings when trained with relevance judgments. When provided with additional cohort descriptions, the L-PCR system will continue to learn, thus offering a potential solution to the performance barriers of current cohort retrieval systems.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Learning relevance models for patient cohort retrieval

Goodwin

Harabagiu

2018

JAMIA Open

View full text Add to dashboard Cite

show abstract

“…Current state of the art systems do not operate at this level of accuracy due to a lack of adequate machine learning resources and hence are not in widespread use. Therefore, NEDC has been developing a large database of clinical EEGs [2] collected at Temple University Hospital that we believe is crucial for the application of more powerful machine learning approaches. Figure 1.…”

Section: Introductionmentioning

confidence: 99%

The TUH EEG CORPUS: A big data resource for automated EEG interpretation

Harati

Lopez

Obeid

et al. 2014

2014 IEEE Signal Processing in Medicine and Biology Symposium (SPMB)

Self Cite

View full text Add to dashboard Cite

The Neural Engineering Data Consortium (NEDC) is releasing its first major big data corpus -the Temple University Hospital EEG Corpus. This corpus consists of over 25,000 EEG studies, and includes a neurologist's interpretation of the test, a brief patient medical history and demographic information about the patient such as gender and age. For the first time, there is a sufficient amount of data to support the application of state of the art machine learning algorithms. In this paper, we present pilot results of experiments on the prediction of some basic attributes of an EEG from the raw EEG signal data using a 3,762 session subset of the corpus. Standard machine learning approaches are shown to be capable of predicting commonly occurring events from simple features with high accuracy on closed-loop testing, and can deliver error rates below 50% on a 6-way open set classification problem. This is very promising performance since commercial technology fails on this data.

show abstract

“…La base de datos EPILEPSIAE (una colección europea de información sobre epilepsia) contiene los registros EEG de 275 pacientes de los centros de epilepsia del Hospital Universitario de Friburgo, Alemania, del Hospital de la Universidad de Coimbra, Portugal, y del Hospital de la Pitier-Salpêtrière París, Francia [7].Finalmente, el corpus HUT-EEG liberó recientemente 14 años de registros EEG clínicos recogidos en el Hospital de la Universidad de Temple. La colección consta de 16,986 sesiones de 10,874 sujetosúnicos; los registros han sido organizados y se combinan con informes clínicos que describen los pacientes y las exploraciones [6].…”

Section: Introductionunclassified

“…Relativamente poca de esta información está disponible públicamente a la comunidad de investigación en una forma que seaútil para el aprendizaje automático [6], y la que se encuentra disponible está dirigida a estudios médicos. Hasta donde tenemos conocimiento, en la literatura no se cuenta con un corpus disponible públicamente de señales EEG con etiquetas de emociones.…”

Section: Introductionunclassified

SEEG-BUAP: Un corpus de señales electro-encefalográficas con emociones inducidas

Morales¹,

Vilariño²,

Aguilar³

et al. 2016

RCS

View full text Add to dashboard Cite

http://www.lke.buap.mx/ Resumen. En este artículo se presenta un corpus de señales electroencefalo-gráficas etiquetado manualmente con emociones inducidas a traves de videos y encabezados de noticias. El corpus, al que se le ha denominado SEEG-BUAP, fue construido en base a una población de 40 personas a quienes se le mostró una serie de videos que detonaban algún tipo de emoción del estilo "alegría", "tristeza", etc. Se espera que la colección de datos construida y puesta a disposición pública sirva como base para el desarrollo de modelos de clasificación automática basadas en emociones básicas detectadas a través de señales electro-encefalográficas.

show abstract

The Temple University Hospital EEG corpus

Cited by 19 publications

References 6 publications

Learning relevance models for patient cohort retrieval

Learning relevance models for patient cohort retrieval

The TUH EEG CORPUS: A big data resource for automated EEG interpretation

SEEG-BUAP: Un corpus de señales electro-encefalográficas con emociones inducidas

Contact Info

Product

Resources

About