A novel way to use cross-validation to measure connectivity by machine learning allows epilepsy surgery outcome prediction
The rate of success of epilepsy surgery, ensuring seizure-freedom, is limited by the lack of epileptogenicity biomarkers. Previous evidence supports the critical role of functional connectivity during seizure generation to characterize the epileptogenic network (EN). However, EN dynamics is highly v...
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-02-01
|
| Series: | NeuroImage |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S1053811924004877 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832590945200111616 |
|---|---|
| author | Karla Ivankovic Alessandro Principe Justo Montoya-Gálvez Linus Manubens-Gil Riccardo Zucca Pablo Villoslada Mara Dierssen Rodrigo Rocamora |
| author_facet | Karla Ivankovic Alessandro Principe Justo Montoya-Gálvez Linus Manubens-Gil Riccardo Zucca Pablo Villoslada Mara Dierssen Rodrigo Rocamora |
| author_sort | Karla Ivankovic |
| collection | DOAJ |
| description | The rate of success of epilepsy surgery, ensuring seizure-freedom, is limited by the lack of epileptogenicity biomarkers. Previous evidence supports the critical role of functional connectivity during seizure generation to characterize the epileptogenic network (EN). However, EN dynamics is highly variable across patients, hindering the development of diagnostic biomarkers. Without relying on specific connectivity variables, we focused on a general hypothesis that the EN undergoes the greatest magnitude of connectivity change during seizure generation, compared to other brain networks. To test this hypothesis, we developed a novel method for quantifying connectivity change between network states and applied it to identify surgical resection areas.A network state was represented by random snapshots of connectivity within a defined time interval of an intracranial EEG recording. A binary classifier was applied to classify two network states. The classifier generalization performance estimated by cross-validation was employed as a continuous measure of connectivity change. The algorithm generated a network by iteratively adding nodes until the connectivity change magnitude decreased. The resulting network was compared to the surgical resection, and the overlap score was used to predict post-surgical outcomes. The framework was evaluated in a consecutive cohort of 21 patients with a post-surgical follow-up of minimum 3 years.The best overlap between connectivity change networks and resections was obtained at the transition from pre-seizure to seizure (surgical outcome prediction ROC-AUC=90.3 %). However, all patients except one were correctly classified when considering the most informative time intervals. Time intervals proportional to seizure length were more informative than the almost universally used fixed intervals.This study demonstrates that connectivity can be successfully classified with a machine learning analysis and provide information for distinguishing a separate epileptogenic functional network. In summary, the connectivity change analysis could accurately identify epileptogenic networks validated by surgery outcome classification. Connectivity change magnitude at seizure transition could potentially serve as an EN biomarker. The tool provided by this study may aid surgical decision-making. |
| format | Article |
| id | doaj-art-3e60a98642e041d782f39f51d8a6af67 |
| institution | Kabale University |
| issn | 1095-9572 |
| language | English |
| publishDate | 2025-02-01 |
| publisher | Elsevier |
| record_format | Article |
| series | NeuroImage |
| spelling | doaj-art-3e60a98642e041d782f39f51d8a6af672025-01-23T05:26:19ZengElsevierNeuroImage1095-95722025-02-01306120990A novel way to use cross-validation to measure connectivity by machine learning allows epilepsy surgery outcome predictionKarla Ivankovic0Alessandro Principe1Justo Montoya-Gálvez2Linus Manubens-Gil3Riccardo Zucca4Pablo Villoslada5Mara Dierssen6Rodrigo Rocamora7Hospital del Mar Research Institute, 08003 Barcelona, Spain; Universitat Pompeu Fabra, 08003 Barcelona, SpainHospital del Mar Research Institute, 08003 Barcelona, Spain; Universitat Pompeu Fabra, 08003 Barcelona, Spain; Epilepsy Unit - Neurology Dept. Hospital del Mar, 08003 Barcelona, Spain; Corresponding author at: Hospital del Mar, Passeig Marítim 25-29, 08003 Barcelona, Spain.Universitat Pompeu Fabra, 08003 Barcelona, SpainNew Cornerstone Science Laboratory, SEU-ALLEN Joint Center, State Key Laboratory of Digital Medical Engineering, Institute for Brain and Intelligence, Southeast University, Nanjing, Jiangsu 210096, ChinaHospital del Mar Research Institute, 08003 Barcelona, Spain; Radboud University, Nijmegen, the NetherlandsHospital del Mar Research Institute, 08003 Barcelona, Spain; Universitat Pompeu Fabra, 08003 Barcelona, SpainHospital del Mar Research Institute, 08003 Barcelona, Spain; Universitat Pompeu Fabra, 08003 Barcelona, Spain; Centre for Genomic Regulation (CRG), The Barcelona Institute for Science and Technology (BIST) 08003 Barcelona, Spain; Biomedical Research Networking Center on Rare Diseases (CIBERER), Barcelona, SpainHospital del Mar Research Institute, 08003 Barcelona, Spain; Universitat Pompeu Fabra, 08003 Barcelona, Spain; Epilepsy Unit - Neurology Dept. Hospital del Mar, 08003 Barcelona, SpainThe rate of success of epilepsy surgery, ensuring seizure-freedom, is limited by the lack of epileptogenicity biomarkers. Previous evidence supports the critical role of functional connectivity during seizure generation to characterize the epileptogenic network (EN). However, EN dynamics is highly variable across patients, hindering the development of diagnostic biomarkers. Without relying on specific connectivity variables, we focused on a general hypothesis that the EN undergoes the greatest magnitude of connectivity change during seizure generation, compared to other brain networks. To test this hypothesis, we developed a novel method for quantifying connectivity change between network states and applied it to identify surgical resection areas.A network state was represented by random snapshots of connectivity within a defined time interval of an intracranial EEG recording. A binary classifier was applied to classify two network states. The classifier generalization performance estimated by cross-validation was employed as a continuous measure of connectivity change. The algorithm generated a network by iteratively adding nodes until the connectivity change magnitude decreased. The resulting network was compared to the surgical resection, and the overlap score was used to predict post-surgical outcomes. The framework was evaluated in a consecutive cohort of 21 patients with a post-surgical follow-up of minimum 3 years.The best overlap between connectivity change networks and resections was obtained at the transition from pre-seizure to seizure (surgical outcome prediction ROC-AUC=90.3 %). However, all patients except one were correctly classified when considering the most informative time intervals. Time intervals proportional to seizure length were more informative than the almost universally used fixed intervals.This study demonstrates that connectivity can be successfully classified with a machine learning analysis and provide information for distinguishing a separate epileptogenic functional network. In summary, the connectivity change analysis could accurately identify epileptogenic networks validated by surgery outcome classification. Connectivity change magnitude at seizure transition could potentially serve as an EN biomarker. The tool provided by this study may aid surgical decision-making.http://www.sciencedirect.com/science/article/pii/S1053811924004877Machine learningFunctional connectivitySeizure generationEpileptogenic networkEpilepsy surgeryOutcome prediction |
| spellingShingle | Karla Ivankovic Alessandro Principe Justo Montoya-Gálvez Linus Manubens-Gil Riccardo Zucca Pablo Villoslada Mara Dierssen Rodrigo Rocamora A novel way to use cross-validation to measure connectivity by machine learning allows epilepsy surgery outcome prediction NeuroImage Machine learning Functional connectivity Seizure generation Epileptogenic network Epilepsy surgery Outcome prediction |
| title | A novel way to use cross-validation to measure connectivity by machine learning allows epilepsy surgery outcome prediction |
| title_full | A novel way to use cross-validation to measure connectivity by machine learning allows epilepsy surgery outcome prediction |
| title_fullStr | A novel way to use cross-validation to measure connectivity by machine learning allows epilepsy surgery outcome prediction |
| title_full_unstemmed | A novel way to use cross-validation to measure connectivity by machine learning allows epilepsy surgery outcome prediction |
| title_short | A novel way to use cross-validation to measure connectivity by machine learning allows epilepsy surgery outcome prediction |
| title_sort | novel way to use cross validation to measure connectivity by machine learning allows epilepsy surgery outcome prediction |
| topic | Machine learning Functional connectivity Seizure generation Epileptogenic network Epilepsy surgery Outcome prediction |
| url | http://www.sciencedirect.com/science/article/pii/S1053811924004877 |
| work_keys_str_mv | AT karlaivankovic anovelwaytousecrossvalidationtomeasureconnectivitybymachinelearningallowsepilepsysurgeryoutcomeprediction AT alessandroprincipe anovelwaytousecrossvalidationtomeasureconnectivitybymachinelearningallowsepilepsysurgeryoutcomeprediction AT justomontoyagalvez anovelwaytousecrossvalidationtomeasureconnectivitybymachinelearningallowsepilepsysurgeryoutcomeprediction AT linusmanubensgil anovelwaytousecrossvalidationtomeasureconnectivitybymachinelearningallowsepilepsysurgeryoutcomeprediction AT riccardozucca anovelwaytousecrossvalidationtomeasureconnectivitybymachinelearningallowsepilepsysurgeryoutcomeprediction AT pablovilloslada anovelwaytousecrossvalidationtomeasureconnectivitybymachinelearningallowsepilepsysurgeryoutcomeprediction AT maradierssen anovelwaytousecrossvalidationtomeasureconnectivitybymachinelearningallowsepilepsysurgeryoutcomeprediction AT rodrigorocamora anovelwaytousecrossvalidationtomeasureconnectivitybymachinelearningallowsepilepsysurgeryoutcomeprediction AT karlaivankovic novelwaytousecrossvalidationtomeasureconnectivitybymachinelearningallowsepilepsysurgeryoutcomeprediction AT alessandroprincipe novelwaytousecrossvalidationtomeasureconnectivitybymachinelearningallowsepilepsysurgeryoutcomeprediction AT justomontoyagalvez novelwaytousecrossvalidationtomeasureconnectivitybymachinelearningallowsepilepsysurgeryoutcomeprediction AT linusmanubensgil novelwaytousecrossvalidationtomeasureconnectivitybymachinelearningallowsepilepsysurgeryoutcomeprediction AT riccardozucca novelwaytousecrossvalidationtomeasureconnectivitybymachinelearningallowsepilepsysurgeryoutcomeprediction AT pablovilloslada novelwaytousecrossvalidationtomeasureconnectivitybymachinelearningallowsepilepsysurgeryoutcomeprediction AT maradierssen novelwaytousecrossvalidationtomeasureconnectivitybymachinelearningallowsepilepsysurgeryoutcomeprediction AT rodrigorocamora novelwaytousecrossvalidationtomeasureconnectivitybymachinelearningallowsepilepsysurgeryoutcomeprediction |