Search

Your search keyword '"Tapani, Karoliina"' showing total 17 results
Start Over Author "Tapani, Karoliina"
17 results on '"Tapani, Karoliina"'

1. Validating an SVM-based neonatal seizure detection algorithm for generalizability, non-inferiority and clinical efficacy

Author

Tapani, Karoliina T., Nevalainen, Päivi, Vanhatalo, Sampsa, and Stevenson, Nathan J.

Subjects
Computer Science - Machine Learning
Abstract

Neonatal seizure detection algorithms (SDA) are approaching the benchmark of human expert annotation. Measures of algorithm generalizability and non-inferiority as well as measures of clinical efficacy are needed to assess the full scope of neonatal SDA performance. We validated our neonatal SDA on an independent data set of 28 neonates. Generalizability was tested by comparing the performance of the original training set (cross-validation) to its performance on the validation set. Non-inferiority was tested by assessing inter-observer agreement between combinations of SDA and two human expert annotations. Clinical efficacy was tested by comparing how the SDA and human experts quantified seizure burden and identified clinically significant periods of seizure activity in the EEG. Algorithm performance was consistent between training and validation sets with no significant worsening in AUC (p>0.05, n =28). SDA output was inferior to the annotation of the human expert, however, re-training with an increased diversity of data resulted in non-inferior performance ($\Delta\kappa$=0.077, 95% CI: -0.002-0.232, n=18). The SDA assessment of seizure burden had an accuracy ranging from 89-93%, and 87% for identifying periods of clinical interest. The proposed SDA is approaching human equivalence and provides a clinically relevant interpretation of the EEG.

Published
2022

3. Corrigendum: Building an open source classifier for the neonatal EEG background: a systematic feature-based approach from expert scoring to clinical visualization

Author

Montazeri, Saeed, primary, Pinchefsky, Elana, additional, Tse, Ilse, additional, Marchi, Viviana, additional, Kohonen, Jukka, additional, Kauppila, Minna, additional, Airaksinen, Manu, additional, Tapani, Karoliina, additional, Nevalainen, Päivi, additional, Hahn, Cecil, additional, Tam, Emily W. Y., additional, Stevenson, Nathan J., additional, and Vanhatalo, Sampsa, additional

Published
2024
Full Text
View/download PDF

4. Incorporating spike correlations into an SVM-based neonatal seizure detector

Author

Tapani, Karoliina T., Vanhatalo, Sampsa, Stevenson, Nathan J., Magjarevic, Ratko, Editor-in-chief, Ładyżyński, Piotr, Series editor, Ibrahim, Fatimah, Series editor, Lacković, Igor, Series editor, Rock, Emilio Sacristan, Series editor, Eskola, Hannu, editor, Väisänen, Outi, editor, Viik, Jari, editor, and Hyttinen, Jari, editor

Published
2018
Full Text
View/download PDF

5. Building an Open Source Classifier for the Neonatal EEG Background: A Systematic Feature-Based Approach From Expert Scoring to Clinical Visualization.

Author

Montazeri, Saeed, Pinchefsky, Elana, Tse, Ilse, Marchi, Viviana, Kohonen, Jukka, Kauppila, Minna, Airaksinen, Manu, Tapani, Karoliina, Nevalainen, Päivi, Hahn, Cecil, Tam, Emily W. Y., Stevenson, Nathan J., and Vanhatalo, Sampsa

Subjects
NEONATAL intensive care units, ELECTROENCEPHALOGRAPHY, ARTIFICIAL neural networks, DATA visualization, COMPUTATIONAL neuroscience
Abstract

Neonatal brain monitoring in the neonatal intensive care units (NICU) requires a continuous review of the spontaneous cortical activity, i.e., the electroencephalograph (EEG) background activity. This needs development of bedside methods for an automated assessment of the EEG background activity. In this paper, we present development of the key components of a neonatal EEG background classifier, starting from the visual background scoring to classifier design, and finally to possible bedside visualization of the classifier results. A dataset with 13,200 5-minute EEG epochs (8-- 16 channels) from 27 infants with birth asphyxia was used for classifier training after scoring by two independent experts. We tested three classifier designs based on 98 computational features, and their performance was assessed with respect to scoring system, pre- and post-processing of labels and outputs, choice of channels, and visualization in monitor displays. The optimal solution achieved an overall classification accuracy of 97% with a range across subjects of 81--100%. We identified a set of 23 features that make the classifier highly robust to the choice of channels and missing data due to artefact rejection. Our results showed that an automated bedside classifier of EEG background is achievable, and we publish the full classifier algorithm to allow further clinical replication and validation studies. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

10. Building an Open Source Classifier for the Neonatal EEG Background

Author

Moghadam, Saeed Montazeri, Pinchefsky, Elana, Tse, Ilse, Marchi, Viviana, Kohonen, Jukka, Kauppila, Minna, Airaksinen, Manu, Tapani, Karoliina, Nevalainen, Päivi, Hahn, Cecil, Tam, Emily W.Y., Stevenson, Nathan J., Vanhatalo, Sampsa, University of Helsinki, University of Montreal, Department of Computer Science, Dept Signal Process and Acoust, University of Toronto, Queensland Institute of Medical Research, Aalto-yliopisto, and Aalto University

Subjects
EEG monitoring, support vector machine, neonatal EEG, background classifier, neonatal intensive care unit, artificial neural network, EEG trend
Abstract

openaire: EC/H2020/813483/EU//INFANS Neonatal brain monitoring in the neonatal intensive care units (NICU) requires a continuous review of the spontaneous cortical activity, i.e., the electroencephalograph (EEG) background activity. This needs development of bedside methods for an automated assessment of the EEG background activity. In this paper, we present development of the key components of a neonatal EEG background classifier, starting from the visual background scoring to classifier design, and finally to possible bedside visualization of the classifier results. A dataset with 13,200 5-minute EEG epochs (8–16 channels) from 27 infants with birth asphyxia was used for classifier training after scoring by two independent experts. We tested three classifier designs based on 98 computational features, and their performance was assessed with respect to scoring system, pre- and post-processing of labels and outputs, choice of channels, and visualization in monitor displays. The optimal solution achieved an overall classification accuracy of 97% with a range across subjects of 81–100%. We identified a set of 23 features that make the classifier highly robust to the choice of channels and missing data due to artefact rejection. Our results showed that an automated bedside classifier of EEG background is achievable, and we publish the full classifier algorithm to allow further clinical replication and validation studies.

Published
2021

11. Building an Open Source Classifier for the Neonatal EEG Background: A Systematic Feature-Based Approach From Expert Scoring to Clinical Visualization

Author

Montazeri Moghadam, Saeed, Pinchefsky, Elana, Tse, Ilse, Marchi, Viviana, Kohonen, Jukka, Kauppila, Minna, Airaksinen, Manu, Tapani, Karoliina, Nevalainen, Päivi, Hahn, Cecil, W. Y. Tam, Emily, Stevenson, Nathan J., Vanhatalo, Sampsa, Department of Neurosciences, Department of Computer Science, HUSLAB, HUS Medical Imaging Center, BioMag Laboratory, Kliinisen neurofysiologian yksikkö, and Clinicum

Subjects
EEG monitoring, ALGORITHMS, 217 Medical engineering, neonatal EEG, TERM, neonatal intensive care unit, PROGNOSTIC VALUE, ELECTROENCEPHALOGRAPHY, Neonatal intensive care unit (NICU), support vector machine, background classifier, INTERRATER AGREEMENT, artificial neural network, Neuroscience, Original Research, EEG trend
Abstract

Neonatal brain monitoring in the neonatal intensive care units (NICU) requires a continuous review of the spontaneous cortical activity, i.e., the electroencephalograph (EEG) background activity. This needs development of bedside methods for an automated assessment of the EEG background activity. In this paper, we present development of the key components of a neonatal EEG background classifier, starting from the visual background scoring to classifier design, and finally to possible bedside visualization of the classifier results. A dataset with 13,200 5-minute EEG epochs (8–16 channels) from 27 infants with birth asphyxia was used for classifier training after scoring by two independent experts. We tested three classifier designs based on 98 computational features, and their performance was assessed with respect to scoring system, pre- and post-processing of labels and outputs, choice of channels, and visualization in monitor displays. The optimal solution achieved an overall classification accuracy of 97% with a range across subjects of 81–100%. We identified a set of 23 features that make the classifier highly robust to the choice of channels and missing data due to artefact rejection. Our results showed that an automated bedside classifier of EEG background is achievable, and we publish the full classifier algorithm to allow further clinical replication and validation studies.

Published
2021

12. Building an Open Source Classifier for the Neonatal EEG Background: A Systematic Feature-Based Approach From Expert Scoring to Clinical Visualization

Author

Moghadam, Saeed Montazeri, primary, Pinchefsky, Elana, additional, Tse, Ilse, additional, Marchi, Viviana, additional, Kohonen, Jukka, additional, Kauppila, Minna, additional, Airaksinen, Manu, additional, Tapani, Karoliina, additional, Nevalainen, Päivi, additional, Hahn, Cecil, additional, Tam, Emily W. Y., additional, Stevenson, Nathan J., additional, and Vanhatalo, Sampsa, additional

Published
2021
Full Text
View/download PDF

16. Optimization of registration pipeline for head MR and CT images

Author

Salli, Eero, Sähkötekniikan korkeakoulu, Parkkonen, Lauri, Tapani, Karoliina, Salli, Eero, Sähkötekniikan korkeakoulu, Parkkonen, Lauri, and Tapani, Karoliina

Abstract

In this thesis work, the aim was to find a robust, optimal rigid registration process to accurately and automatically align computed tomography (CT) and magnetic resonance (MR) images of the brain. For patients undergoing, for example, stereoelectroencephalography (epilepsy patients) or implantation of stimulating electrodes in the brain (Parkinson’s patients), it is crucial to be able to combine information from low-dose CT and MR with great precision. Registration was performed with SimpleITK interface to the image registration framework of the United States National Library of Medicine Insight Segmentation and Registration Toolkit (ITK). In the optimization process an existing SimpleITK example was used as a basis for the registration algorithm, which was then optimized one block at a time beginning with the initial alignment. Registration accuracy was determined by comparing the automatic transform of our registration algorithm to the transform of a semiautomatic registration performed with a semiautomatic ITK based software, ipcWorkstation, which is used and developed in HUS Medical Imaging Center. As a result, a robust rigid registration algorithm was developed. The maximum registration errors with the final algorithm were less than 2 mm for 7 out of 15 and less than 4 mm for 12 out of 15 patients. The algorithm performs registration up to initial rotations of 45 degrees. The fast development of automated registration algorithm presented in this thesis appears promising to be used for other applications as well. This kind of block-wise optimization pattern could be used to optimize the registration either for images of other parts of the body or for other imaging modalities such as positron emission tomography (PET) and MR., Tämän diplomityön tarkoituksena oli löytää optimaalinen ja automaattinen tietokonetomografia- ja magneettikuvien kohdennusmenetelmä. Kohdennus suoritettiin käyttäen hyväksi SimpleITK-ohjelmakirjastoa, joka perustuu ITK kuvakohdennus ohjelmakirjastoon (engl. the United States National Library of Medicine Insight Segmentation and Registration Toolkit). Optimointi aloitettiin SimpleITKesimerkin pohjalta, jonka parametreja optimoitiin osa kerrallaan lähtien liikkeelle kohdennuksen alustuksesta. Kohdennustarkkuus määritettiin vertaamalla optimoidulla kohdennusohjelmalla saatua automaattista muunnosmatriisia puoliautomaattisella menetelmällä saatuun muunnosmatriisiin. Puoliautomaattinen muunnos tehtiin HUS-Kuvantamisessa kehitetyllä ipcWorkstation-ohjelmalla, joka myös perustuu ITK-ohjelmakirjastoon. Työn tuloksena saatiin luotettavasti toimiva jäykän kuvakohdennuksen suorittava algoritmi, joka pohjautuu SimpleITK:n Python-kirjastoon. Seitsemällä 15 potilaasta suurin kohdennusvirhe oli alle 2 mm ja 12:lla 15 potilaasta alle 4 mm. Kohdennus onnistuu jopa 45 asteen lähtökohtaisilla kulmaeroilla. Työssä käytettyä nopeaa algoritmikehitystekniikkaa voitaisiin käyttää optimointiin muillekin sovelluksille. Tulevaisuudessa algoritmioptimointia osa kerrallaan voisi hyödyntää kohdennusparametrien optimointiin jonkin muun vartalon alueen rakenteellisten kuvien kohdennukseen tai eri kuvamodaliteettien kohdennukseen kuten positroniemissiotomografia- ja magneettikuvien kohdennukseen.

Published
2016

17. Hybrid neonatal EEG seizure detection algorithms achieve the benchmark of visual interpretation of the human expert .

Author

Stevenson N, Tapani K, and Vanhatalo S

Subjects
Humans, Infant, Newborn, Algorithms, Electroencephalography, Epilepsy diagnosis, Seizures diagnosis
Abstract

Neonatal EEG seizure detection algorithms (NSDAs) have an upper bound of performance related to the agreement between visual interpretation of human experts. No published algorithms have reported performance that has reached this upper bound. In this paper, we combined two recently developed NSDAs in order to improve detection performance. An outlier detection stage was also added to improve robustness in the presence of unseen data. A large database of EEG from 79 term infants labeled by three independent human experts was used to develop and test the sufficiency of the hybrid NSDA. The inter-observer agreement (IOA) between experts was high (κ = 0.757, 95%CI: 0.665-0.836, n=79). The area under the receiver operator characteristic of the NSDA compared to the consensus annotation of the human experts was 0.952 (95%CI: 0.0927-0.971). The IOA of seizure detection between the NSDA and human experts was not significantly less than the IOA among human experts (Δκ = 0.022, 95%CI: -0.20 to 0.072) and was further improved by increasing the minimum seizure duration from 10s to 30s (Δκ = -0.002, 95%CI: -0.073 to 0.055). Automated methods of neonatal EEG seizure detection have sufficient accuracy to replace human interpretation, potentially, providing reliable interpretations for clinicians in the neonatal intensive care unit.

Published
2019
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results