A probabilistic SVM based decision system for pain diagnosis

Yang, Jinglin; Li, Han-Xiong; Hu, Yong

doi:10.1016/j.eswa.2011.01.106

Cited by 25 publications

(5 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…From the prior section, and from the literature, it is seen that the kernel approach shows a good compatibility with newborn EEG, thus for this exercise a probabilistic SVM (PSVM) learner was used to assign probability values to the samples. These were classified in terms of having seizures and followed by splitting the probability values into two groups, which reflected the severity of the seizure; from here a classification model was used to design the automatic recognition of these cases [43] .…”

Section: Probabilistic Seizure Predictionmentioning

confidence: 99%

“…The implemented PSVM model works in a similar way to the discrete SVM classifier, but also utilizes the Platt scaling, which serves as a conversion mechanism to transform classifier scores into a form of probability distribution where, as mentioned, each sample assigned to a class is accompanied by a probability score [43] . The Platt scaling, which is an integral part of this process, works by transforming output scores into probabilistic representa-tions by utilizing the logistic regression model, which can be defined as Equation ( 5) [44] :…”

Section: Probabilistic Seizure Predictionmentioning

confidence: 99%

See 1 more Smart Citation

Artificial Intelligence-assisted Care for Human Newborns with Neurological Impairments

Nsugbe¹

2022

digit. tech. res. and appl.

View full text Add to dashboard Cite

Seizures are a widespread condition affecting 50~65 million people in the world, and newborns are also susceptible to them. EEG is used to monitor the brain activity of newborns with suspected brain injuries, followed by a qualitative waveform interpretation by a group of clinical experts, where the means towards detection of seizures include a set of distinct characteristics in the waveform. This means of seizure detection has been critiqued, particularly due to subjectivity where, at times, waveform reviewing clinicians fail to reach a consensus on the presence of seizure activity in the brain of a newborn. As a means towards dealing with this problem, the author investigated the use of Artificial Intelligence-driven prediction machines capable of an automated diagnosis of seizure, based on a newborn’s EEG waveform. This approach used a reduced selection of EEG electrodes, the Linear Series Decomposition Learner (LSDL), an ensemble of a group of features, and performance comparison across multiple classification models. Secondary work was also carried out, which leveraged the patient information available alongside the EEG dataset. This involved the use of EEG towards predicting the level of asphyxia within the neonatal brain. The results from the seizure prediction exercise showed an increment in prediction performance of the seizures when preprocessed with the LSDL. The results spanned a range of figures (depending on the classification model), with the highest accuracy of 88.1%, while a probabilistic approach towards predicting the extent of seizures provided a maximum accuracy of 93.5%. The results from the secondary analysis showed a maximum accuracy for asphyxia prediction of 89.1%. The obtained results have helped to demonstrate that a reduced selection of electrode segments, alongside the selected algorithms, can serve towards the prediction of seizures for newborns within a neonatal intensive care unit.

show abstract

Section: Probabilistic Seizure Predictionmentioning

confidence: 99%

Section: Probabilistic Seizure Predictionmentioning

confidence: 99%

Artificial Intelligence-assisted Care for Human Newborns with Neurological Impairments

Nsugbe¹

2022

digit. tech. res. and appl.

View full text Add to dashboard Cite

show abstract

“…Finally, as proposed by [40,49], the Support Vector Machine (SVM) [94] aims to map the training data to a higher dimensional space and separate the different classes of data, by constructing the optimal separating hyper-plane. This model has good generalisation ability and a robustness for high dimensional data [61,64].…”

Section: Statistical Learning Algorithmsmentioning

confidence: 99%

“…The SVM is more suited to training and performs better compared to ANN [69]. However it is very sensitive to uncertainties [49,61], and a too high dimensional space can lead to overfitting of the data [69,95] and so slow the speed of the training [64,96].…”

Section: Statistical Learning Algorithmsmentioning

confidence: 99%

See 1 more Smart Citation

Applied computer technologies in clinical decision support systems for pain management: A systematic review

Pombo

Araújo

Viana

2014

Journal of Intelligent &Amp; Fuzzy Systems

View full text Add to dashboard Cite

Millions of people around the world suffer from pain, acute or chronic and this raises the importance of its screening, assessment and treatment. Pain, is highly subjective and the use of clinical decision support systems (CDSSs) can play an important part in improving the accuracy of pain assessment, and lead to better clinical practices. This review examines CDSSs, in relation to computer technologies and was conducted with the following electronic databases: CiteSeer x , IEEE Xplore, ISI Web of Knowledge, Mendeley, Microsoft Academic Search, PubMed, Science Accelerator, Science.gov, ScienceDirect, SpringerLink, and The Cochrane Library. The studies referenced were compiled with several criteria in mind. Firstly, that they constituted a decision support system. Secondly, that study data included pain values or results based on the detection of pain. Thirdly, that they were published in English, between 1992 and 2011, and finally that they focused on patients with acute or chronic pain. In total, thirty-nine studies highlighted the following topics: rule based algorithms, artificial neural networks, rough and fuzzy sets, statistical learning algorithms, terminologies, questionnaires and scores. The median accuracy ranged from 53% to 87.5%. The lack of integration with mobile devices, the limited use of web-based interfaces and the scarcity of systems that allow for data to be inserted by patients were all limitations that were detected.

show abstract

A New Method for Classifying Random Variables Based on Support Vector Machine

Abaszade

Effati

2018

J Classif

View full text Add to dashboard Cite

A probabilistic SVM based decision system for pain diagnosis

Cited by 25 publications

References 22 publications

Artificial Intelligence-assisted Care for Human Newborns with Neurological Impairments

Artificial Intelligence-assisted Care for Human Newborns with Neurological Impairments

Applied computer technologies in clinical decision support systems for pain management: A systematic review

A New Method for Classifying Random Variables Based on Support Vector Machine

Contact Info

Product

Resources

About