Home > Research > Publications & Outputs > Cell-nuclear data reduction and prognostic mode...
View graph of relations

Cell-nuclear data reduction and prognostic model selection in bladder tumor recurrence

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Published

Standard

Cell-nuclear data reduction and prognostic model selection in bladder tumor recurrence. / Tasoulis, Dimitrios K; Spyridonos , P.; Pavlidis, Nicos et al.

In: Artificial Intelligence in Medicine, Vol. 38, No. 3, 11.2006, p. 291-303.

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Harvard

Tasoulis, DK, Spyridonos , P, Pavlidis, N, Cavouras , D, Ravazoula, P, Nikiforidis, G & Vrahatis, MN 2006, 'Cell-nuclear data reduction and prognostic model selection in bladder tumor recurrence', Artificial Intelligence in Medicine, vol. 38, no. 3, pp. 291-303. https://doi.org/10.1016/j.artmed.2006.07.008

APA

Tasoulis, D. K., Spyridonos , P., Pavlidis, N., Cavouras , D., Ravazoula, P., Nikiforidis, G., & Vrahatis, M. N. (2006). Cell-nuclear data reduction and prognostic model selection in bladder tumor recurrence. Artificial Intelligence in Medicine, 38(3), 291-303. https://doi.org/10.1016/j.artmed.2006.07.008

Vancouver

Tasoulis DK, Spyridonos P, Pavlidis N, Cavouras D, Ravazoula P, Nikiforidis G et al. Cell-nuclear data reduction and prognostic model selection in bladder tumor recurrence. Artificial Intelligence in Medicine. 2006 Nov;38(3):291-303. doi: 10.1016/j.artmed.2006.07.008

Author

Tasoulis, Dimitrios K ; Spyridonos , P. ; Pavlidis, Nicos et al. / Cell-nuclear data reduction and prognostic model selection in bladder tumor recurrence. In: Artificial Intelligence in Medicine. 2006 ; Vol. 38, No. 3. pp. 291-303.

Bibtex

@article{47863363c0c14110ba673c8791c86221,
title = "Cell-nuclear data reduction and prognostic model selection in bladder tumor recurrence",
abstract = "ObjectiveThe paper aims at improving the prediction of superficial bladder recurrence. To this end, feedforward neural networks (FNNs) and a feature selection method based on unsupervised clustering, were employed.Material and methodsA retrospective prognostic study of 127 patients diagnosed with superficial urinary bladder cancer was performed. Images from biopsies were digitized and cell nuclei features were extracted. To design FNN classifiers, different training methods and architectures were investigated. The unsupervised k-windows (UKW) and the fuzzy c-means clustering algorithms were applied on the feature set to identify the most informative feature subsets.ResultsUKW managed to reduce the dimensionality of the feature space significantly, and yielded prediction rates 87.95% and 91.41%, for non-recurrent and recurrent cases, respectively. The prediction rates achieved with the reduced feature set were marginally lower compared to the ones attained with the complete feature set. The training algorithm that exhibited the best performance in all cases was the adaptive on-line backpropagation algorithm.ConclusionsFNNs can contribute to the accurate prognosis of bladder cancer recurrence. The proposed feature selection method can remove redundant information without a significant loss in predictive accuracy, and thereby render the prognostic model less complex, more robust, and hence suitable for clinical use.",
keywords = "Prognosis of cancer recurrence, Neural networks , Unsupervised clustering , Feature selection",
author = "Tasoulis, {Dimitrios K} and P. Spyridonos and Nicos Pavlidis and D. Cavouras and P. Ravazoula and G. Nikiforidis and Vrahatis, {Michael N.}",
year = "2006",
month = nov,
doi = "10.1016/j.artmed.2006.07.008",
language = "English",
volume = "38",
pages = "291--303",
journal = "Artificial Intelligence in Medicine",
issn = "0933-3657",
publisher = "Elsevier",
number = "3",

}

RIS

TY - JOUR

T1 - Cell-nuclear data reduction and prognostic model selection in bladder tumor recurrence

AU - Tasoulis, Dimitrios K

AU - Spyridonos , P.

AU - Pavlidis, Nicos

AU - Cavouras , D.

AU - Ravazoula, P.

AU - Nikiforidis, G.

AU - Vrahatis, Michael N.

PY - 2006/11

Y1 - 2006/11

N2 - ObjectiveThe paper aims at improving the prediction of superficial bladder recurrence. To this end, feedforward neural networks (FNNs) and a feature selection method based on unsupervised clustering, were employed.Material and methodsA retrospective prognostic study of 127 patients diagnosed with superficial urinary bladder cancer was performed. Images from biopsies were digitized and cell nuclei features were extracted. To design FNN classifiers, different training methods and architectures were investigated. The unsupervised k-windows (UKW) and the fuzzy c-means clustering algorithms were applied on the feature set to identify the most informative feature subsets.ResultsUKW managed to reduce the dimensionality of the feature space significantly, and yielded prediction rates 87.95% and 91.41%, for non-recurrent and recurrent cases, respectively. The prediction rates achieved with the reduced feature set were marginally lower compared to the ones attained with the complete feature set. The training algorithm that exhibited the best performance in all cases was the adaptive on-line backpropagation algorithm.ConclusionsFNNs can contribute to the accurate prognosis of bladder cancer recurrence. The proposed feature selection method can remove redundant information without a significant loss in predictive accuracy, and thereby render the prognostic model less complex, more robust, and hence suitable for clinical use.

AB - ObjectiveThe paper aims at improving the prediction of superficial bladder recurrence. To this end, feedforward neural networks (FNNs) and a feature selection method based on unsupervised clustering, were employed.Material and methodsA retrospective prognostic study of 127 patients diagnosed with superficial urinary bladder cancer was performed. Images from biopsies were digitized and cell nuclei features were extracted. To design FNN classifiers, different training methods and architectures were investigated. The unsupervised k-windows (UKW) and the fuzzy c-means clustering algorithms were applied on the feature set to identify the most informative feature subsets.ResultsUKW managed to reduce the dimensionality of the feature space significantly, and yielded prediction rates 87.95% and 91.41%, for non-recurrent and recurrent cases, respectively. The prediction rates achieved with the reduced feature set were marginally lower compared to the ones attained with the complete feature set. The training algorithm that exhibited the best performance in all cases was the adaptive on-line backpropagation algorithm.ConclusionsFNNs can contribute to the accurate prognosis of bladder cancer recurrence. The proposed feature selection method can remove redundant information without a significant loss in predictive accuracy, and thereby render the prognostic model less complex, more robust, and hence suitable for clinical use.

KW - Prognosis of cancer recurrence

KW - Neural networks

KW - Unsupervised clustering

KW - Feature selection

U2 - 10.1016/j.artmed.2006.07.008

DO - 10.1016/j.artmed.2006.07.008

M3 - Journal article

VL - 38

SP - 291

EP - 303

JO - Artificial Intelligence in Medicine

JF - Artificial Intelligence in Medicine

SN - 0933-3657

IS - 3

ER -