Neuroharmony:  A new tool for harmonizing volumetric MRI data from unseen scanners

Rafael Garcia Dias; Cristina Scarpazza; Lea Baecker; Sandra Mendes Vieira; Walter Lopez Pinaya; Aiden Corvin; Alberto  Redolf; Barnaby  Nelson; Benedicto  Crespo-Facorro; Colm  McDonald; Diana Tordesillas-Gutiérrez; Dara Cannon; David  Mothersill; Dennis  Hernaus; Derek  Morris; Esther  Setien-Suero; Gary  Donohoe; Giovanni  Frisoniq; Giulia  Tronchin; Joao Sato; Machteld Marcelis; Matthew Kempton; Neeltje E. M. van Haren; Oliver Gruber; Patrick McGorry; Paul  Amminger; Philip McGuire; Qiyong Gong; René S.  Kahnz; Rosa Ayesa-Arriola; Therese van Amelsvoort; Victor Ortiz-Garcia de la Foz; Vince Calhoun; Wiepke Cahn; Andrea Mechelli

doi:10.1016/j.neuroimage.2020.117127

Neuroharmony: A new tool for harmonizing volumetric MRI data from unseen scanners

Rafael Garcia Dias, Cristina Scarpazza, Lea Baecker, Sandra Mendes Vieira, Walter Lopez Pinaya, Aiden Corvin, Alberto Redolf, Barnaby Nelson, Benedicto Crespo-Facorro, Colm McDonald, Diana Tordesillas-Gutiérrez, Dara Cannon, David Mothersill, Dennis Hernaus, Derek Morris, Esther Setien-Suero, Gary Donohoe, Giovanni Frisoniq, Giulia Tronchin, Joao SatoMachteld Marcelis, Matthew Kempton, Neeltje E. M. van Haren, Oliver Gruber, Patrick McGorry, Paul Amminger, Philip McGuire, Qiyong Gong, René S. Kahnz, Rosa Ayesa-Arriola, Therese van Amelsvoort, Victor Ortiz-Garcia de la Foz, Vince Calhoun, Wiepke Cahn, Andrea Mechelli

Research output: Contribution to journal › Article › peer-review

55 Citations (Scopus)

169 Downloads (Pure)

Abstract

The increasing availability of magnetic resonance imaging (MRI) datasets is boosting the interest in the application ofmachine learning in neuroimaging. A key challenge to the development of reliable machine learning models, and theirtranslational implementation in real-word clinical practice, is the integration of datasets collected using differentscanners. Current approaches for harmonizing multi-scanner data, such as the ComBat method, require a statisticallyrepresentative sample; therefore, these approaches are not suitable for machine learning models aimed at clinicaltranslation where the focus is on the assessment of individual scans from previously unseen scanners. To overcome thischallenge, we developed a tool (‘Neuroharmony’) that is capable of harmonizing single images from unseen/unknownscanners based on a set of image quality metrics, i.e. intrinsic characteristics which can be extracted from individual images without requiring a statistically representative sample. The tool was developed using a mega-dataset ofneuroanatomical data from 15,026 healthy subjects to train a machine learning model that captures the relationshipbetween image quality metrics and the relative volume corrections for each region of the brain prescribed by theComBat method. The tool resulted to be effective in reducing systematic scanner-related bias from new individualimages taken from unseen scanners without requiring any specifications about the image acquisition. Our approachrepresents a significant step forward in the quest to develop reliable imaging-based clinical tools. Neuroharmony andthe instructions on how to use it are available at https://github.com/garciadias/Neuroharmony.

Original language	English
Article number	171127
Journal	NeuroImage
Volume	220
Early online date	4 Jul 2020
DOIs	https://doi.org/10.1016/j.neuroimage.2020.117127
Publication status	Published - 15 Oct 2020

Access to Document

10.1016/j.neuroimage.2020.117127Licence: CC BY

Neuroharmony A new tool_GARCIA DIAS_Published online 4th July2020_GOLD VoR (CC BY)Final published version, 6.19 MBLicence: CC BY

Cite this

Garcia Dias, R., Scarpazza, C., Baecker, L., Mendes Vieira, S., Lopez Pinaya, W., Corvin, A., Redolf, A., Nelson, B., Crespo-Facorro, B., McDonald, C., Tordesillas-Gutiérrez, D., Cannon, D., Mothersill, D., Hernaus, D., Morris, D., Setien-Suero, E., Donohoe, G., Frisoniq, G., Tronchin, G., ... Mechelli, A. (2020). Neuroharmony: A new tool for harmonizing volumetric MRI data from unseen scanners. NeuroImage, 220, Article 171127. https://doi.org/10.1016/j.neuroimage.2020.117127

@article{ff064e40a8ad430ea1e71a9b79c3722d,

title = "Neuroharmony: A new tool for harmonizing volumetric MRI data from unseen scanners",

abstract = "The increasing availability of magnetic resonance imaging (MRI) datasets is boosting the interest in the application ofmachine learning in neuroimaging. A key challenge to the development of reliable machine learning models, and theirtranslational implementation in real-word clinical practice, is the integration of datasets collected using differentscanners. Current approaches for harmonizing multi-scanner data, such as the ComBat method, require a statisticallyrepresentative sample; therefore, these approaches are not suitable for machine learning models aimed at clinicaltranslation where the focus is on the assessment of individual scans from previously unseen scanners. To overcome thischallenge, we developed a tool ({\textquoteleft}Neuroharmony{\textquoteright}) that is capable of harmonizing single images from unseen/unknownscanners based on a set of image quality metrics, i.e. intrinsic characteristics which can be extracted from individual images without requiring a statistically representative sample. The tool was developed using a mega-dataset ofneuroanatomical data from 15,026 healthy subjects to train a machine learning model that captures the relationshipbetween image quality metrics and the relative volume corrections for each region of the brain prescribed by theComBat method. The tool resulted to be effective in reducing systematic scanner-related bias from new individualimages taken from unseen scanners without requiring any specifications about the image acquisition. Our approachrepresents a significant step forward in the quest to develop reliable imaging-based clinical tools. Neuroharmony andthe instructions on how to use it are available at https://github.com/garciadias/Neuroharmony.",

author = "{Garcia Dias}, Rafael and Cristina Scarpazza and Lea Baecker and {Mendes Vieira}, Sandra and {Lopez Pinaya}, Walter and Aiden Corvin and Alberto Redolf and Barnaby Nelson and Benedicto Crespo-Facorro and Colm McDonald and Diana Tordesillas-Guti{\'e}rrez and Dara Cannon and David Mothersill and Dennis Hernaus and Derek Morris and Esther Setien-Suero and Gary Donohoe and Giovanni Frisoniq and Giulia Tronchin and Joao Sato and Machteld Marcelis and Matthew Kempton and {van Haren}, {Neeltje E. M.} and Oliver Gruber and Patrick McGorry and Paul Amminger and Philip McGuire and Qiyong Gong and Kahnz, {Ren{\'e} S.} and Rosa Ayesa-Arriola and {van Amelsvoort}, Therese and {Ortiz-Garcia de la Foz}, Victor and Vince Calhoun and Wiepke Cahn and Andrea Mechelli",

year = "2020",

month = oct,

day = "15",

doi = "10.1016/j.neuroimage.2020.117127",

language = "English",

volume = "220",

journal = "NeuroImage",

issn = "1053-8119",

publisher = "ACADEMIC PRESS INC ELSEVIER SCIENCE",

}

Garcia Dias, R , Scarpazza, C , Baecker, L , Mendes Vieira, S , Lopez Pinaya, W, Corvin, A, Redolf, A, Nelson, B, Crespo-Facorro, B, McDonald, C, Tordesillas-Gutiérrez, D, Cannon, D, Mothersill, D, Hernaus, D, Morris, D, Setien-Suero, E, Donohoe, G, Frisoniq, G, Tronchin, G, Sato, J, Marcelis, M, Kempton, M, van Haren, NEM, Gruber, O, McGorry, P, Amminger, P, McGuire, P, Gong, Q, Kahnz, RS, Ayesa-Arriola, R, van Amelsvoort, T, Ortiz-Garcia de la Foz, V, Calhoun, V, Cahn, W & Mechelli, A 2020, 'Neuroharmony: A new tool for harmonizing volumetric MRI data from unseen scanners', NeuroImage, vol. 220, 171127. https://doi.org/10.1016/j.neuroimage.2020.117127

TY - JOUR

T1 - Neuroharmony

T2 - A new tool for harmonizing volumetric MRI data from unseen scanners

AU - Garcia Dias, Rafael

AU - Scarpazza, Cristina

AU - Baecker, Lea

AU - Mendes Vieira, Sandra

AU - Lopez Pinaya, Walter

AU - Corvin, Aiden

AU - Redolf, Alberto

AU - Nelson, Barnaby

AU - Crespo-Facorro, Benedicto

AU - McDonald, Colm

AU - Tordesillas-Gutiérrez, Diana

AU - Cannon, Dara

AU - Mothersill, David

AU - Hernaus, Dennis

AU - Morris, Derek

AU - Setien-Suero, Esther

AU - Donohoe, Gary

AU - Frisoniq, Giovanni

AU - Tronchin, Giulia

AU - Sato, Joao

AU - Marcelis, Machteld

AU - Kempton, Matthew

AU - van Haren, Neeltje E. M.

AU - Gruber, Oliver

AU - McGorry, Patrick

AU - Amminger, Paul

AU - McGuire, Philip

AU - Gong, Qiyong

AU - Kahnz, René S.

AU - Ayesa-Arriola, Rosa

AU - van Amelsvoort, Therese

AU - Ortiz-Garcia de la Foz, Victor

AU - Calhoun, Vince

AU - Cahn, Wiepke

AU - Mechelli, Andrea

PY - 2020/10/15

Y1 - 2020/10/15

N2 - The increasing availability of magnetic resonance imaging (MRI) datasets is boosting the interest in the application ofmachine learning in neuroimaging. A key challenge to the development of reliable machine learning models, and theirtranslational implementation in real-word clinical practice, is the integration of datasets collected using differentscanners. Current approaches for harmonizing multi-scanner data, such as the ComBat method, require a statisticallyrepresentative sample; therefore, these approaches are not suitable for machine learning models aimed at clinicaltranslation where the focus is on the assessment of individual scans from previously unseen scanners. To overcome thischallenge, we developed a tool (‘Neuroharmony’) that is capable of harmonizing single images from unseen/unknownscanners based on a set of image quality metrics, i.e. intrinsic characteristics which can be extracted from individual images without requiring a statistically representative sample. The tool was developed using a mega-dataset ofneuroanatomical data from 15,026 healthy subjects to train a machine learning model that captures the relationshipbetween image quality metrics and the relative volume corrections for each region of the brain prescribed by theComBat method. The tool resulted to be effective in reducing systematic scanner-related bias from new individualimages taken from unseen scanners without requiring any specifications about the image acquisition. Our approachrepresents a significant step forward in the quest to develop reliable imaging-based clinical tools. Neuroharmony andthe instructions on how to use it are available at https://github.com/garciadias/Neuroharmony.

AB - The increasing availability of magnetic resonance imaging (MRI) datasets is boosting the interest in the application ofmachine learning in neuroimaging. A key challenge to the development of reliable machine learning models, and theirtranslational implementation in real-word clinical practice, is the integration of datasets collected using differentscanners. Current approaches for harmonizing multi-scanner data, such as the ComBat method, require a statisticallyrepresentative sample; therefore, these approaches are not suitable for machine learning models aimed at clinicaltranslation where the focus is on the assessment of individual scans from previously unseen scanners. To overcome thischallenge, we developed a tool (‘Neuroharmony’) that is capable of harmonizing single images from unseen/unknownscanners based on a set of image quality metrics, i.e. intrinsic characteristics which can be extracted from individual images without requiring a statistically representative sample. The tool was developed using a mega-dataset ofneuroanatomical data from 15,026 healthy subjects to train a machine learning model that captures the relationshipbetween image quality metrics and the relative volume corrections for each region of the brain prescribed by theComBat method. The tool resulted to be effective in reducing systematic scanner-related bias from new individualimages taken from unseen scanners without requiring any specifications about the image acquisition. Our approachrepresents a significant step forward in the quest to develop reliable imaging-based clinical tools. Neuroharmony andthe instructions on how to use it are available at https://github.com/garciadias/Neuroharmony.

U2 - 10.1016/j.neuroimage.2020.117127

DO - 10.1016/j.neuroimage.2020.117127

M3 - Article

SN - 1053-8119

VL - 220

JO - NeuroImage

JF - NeuroImage

M1 - 171127

ER -

Neuroharmony: A new tool for harmonizing volumetric MRI data from unseen scanners

Abstract

Access to Document

Fingerprint

Cite this