kth.sePublications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Fast three-dimensional image generation for healthy brain aging using diffeomorphic registration
KTH, School of Engineering Sciences in Chemistry, Biotechnology and Health (CBH), Biomedical Engineering and Health Systems, Medical Imaging.ORCID iD: 0000-0003-4175-395X
Division of Radiology, Department for Clinical Science, Intervention and Technology (CLINTEC), Karolinska Institutet, Stockholm, Sweden; Medical Radiation Physics and Nuclear Medicine, Functional Unit of Nuclear Medicine, Karolinska University Hospital, Huddinge, Stockholm, Sweden, Stockholm.
Department of Psychology, Faculty of Health Sciences, University Fernando Pessoa Canarias, Las Palmas, Spain.
Division of Clinical Geriatrics, Centre for Alzheimer Research, Department of Neurobiology, Care Sciences, and Society (NVS), Karolinska Institutet, Stockholm, Sweden; Department of Neuroimaging, Centre for Neuroimaging Sciences, Institute of Psychiatry, Psychology and Neuroscience, King's College London, London, United Kingdom.
Show others and affiliations
2023 (English)In: Human Brain Mapping, ISSN 1065-9471, E-ISSN 1097-0193, Vol. 44, no 4, p. 1289-1308Article in journal (Refereed) Published
Abstract [en]

Predicting brain aging can help in the early detection and prognosis of neurodegenerative diseases. Longitudinal cohorts of healthy subjects scanned through magnetic resonance imaging (MRI) have been essential to understand the structural brain changes due to aging. However, these cohorts suffer from missing data due to logistic issues in the recruitment of subjects. This paper proposes a methodology for filling up missing data in longitudinal cohorts with anatomically plausible images that capture the subject-specific aging process. The proposed methodology is developed within the framework of diffeomorphic registration. First, two novel modules are introduced within Synthmorph, a fast, state-of-the-art deep learning-based diffeomorphic registration method, to simulate the aging process between the first and last available MRI scan for each subject in three-dimensional (3D). The use of image registration also makes the generated images plausible by construction. Second, we used six image similarity measurements to rearrange the generated images to the specific age range. Finally, we estimated the age of every generated image by using the assumption of linear brain decay in healthy subjects. The methodology was evaluated on 2662 T1-weighted MRI scans from 796 healthy participants from 3 different longitudinal cohorts: Alzheimer's Disease Neuroimaging Initiative, Open Access Series of Imaging Studies-3, and Group of Neuropsychological Studies of the Canary Islands (GENIC). In total, we generated 7548 images to simulate the access of a scan per subject every 6 months in these cohorts. We evaluated the quality of the synthetic images using six quantitative measurements and a qualitative assessment by an experienced neuroradiologist with state-of-the-art results. The assumption of linear brain decay was accurate in these cohorts (R2 ∈ [.924,.940]). The experimental results show that the proposed methodology can produce anatomically plausible aging predictions that can be used to enhance longitudinal datasets. Compared to deep learning-based generative methods, diffeomorphic registration is more likely to preserve the anatomy of the different structures of the brain, which makes it more appropriate for its use in clinical applications. The proposed methodology is able to efficiently simulate anatomically plausible 3D MRI scans of brain aging of healthy subjects from two images scanned at two different time points.

Place, publisher, year, edition, pages
Wiley , 2023. Vol. 44, no 4, p. 1289-1308
Keywords [en]
brain aging, diffeomorphic registration, medical image generation, synthetic brain aging
National Category
Radiology, Nuclear Medicine and Medical Imaging Medical Imaging
Identifiers
URN: urn:nbn:se:kth:diva-328720DOI: 10.1002/hbm.26165ISI: 000897602500001PubMedID: 36468536Scopus ID: 2-s2.0-85144023223OAI: oai:DiVA.org:kth-328720DiVA, id: diva2:1765436
Note

QC 20230610

Available from: 2023-06-10 Created: 2023-06-10 Last updated: 2025-02-09Bibliographically approved
In thesis
1. Methods for brain MRI image synthesis and analysis
Open this publication in new window or tab >>Methods for brain MRI image synthesis and analysis
2024 (English)Doctoral thesis, comprehensive summary (Other academic)
Abstract [en]

Magnetic resonance imaging (MRI) technology has made it possible to observe the inside of the brain in vivo, providing a window for researchers to better understand the processes of human brain development and aging, as well as neurological diseases and their underlying mechanisms. With the development of human brain imaging technology and the accumulation of neuroimaging data, the demand for automated computational algorithms has also increased. The success of machine learning technology in the field of computer vision has promoted its application and development in brain imaging. However, the field of medical imaging faces a series of unique challenges that have not been systematically addressed by the broad deep learning research community. In this thesis, the challenges faced in the context of three specific neuroimaging research problems are tackled. For each problem, the specific challenges are first explained, followed by the proposed solutions.

First, I focus on the specific challenges of using neuroimaging for research in the field of Alzheimer's disease (AD). In the study of AD, one difficulty is to distinguish the effects of normal aging and disease on the anatomical structure of the brain to achieve more accurate and earlier identification. Longitudinal data are an important source for understanding how normal aging affects the brain. However, the cost of collecting longitudinal data can be prohibitive, and current public longitudinal datasets still suffer from inconsistent time intervals and missing scans due to logistical constraints. To address this, I propose in the first study a diffeomorphic registration-based approach to fill in missing time points within current longitudinal datasets. To help distinguish between normal aging and disease-induced morphological changes, I propose in the second study a generative framework that can predict future brain MRI changes along healthy or pathological trajectories based on any given MRI scan. This framework is designed with diffeomorphic constraints to ensure the topological consistency of predicted morphological changes. Finally, in the third study I develop and validate a new deformation-based morphometry (DBM) framework that leverages registration-derived deformation fields for comparisons between healthy and AD groups. This new framework explicitly decomposes deformation into components attributable to normal aging and disease, providing a more nuanced understanding of their interaction.

Secondly, I focus on the pressing challenges in pediatric brain tumor segmentation. Due to the rarity of pediatric brain tumors and the high level of expertise required for tumor labeling—especially when multi-modality imaging is necessary—creating accurate automated segmentation algorithms is challenging. In response, I propose in the fourth study an unsupervised domain adaptation (UDA) framework for segmentation, utilizing relatively abundant adult brain tumor data and labels to assist in pediatric tumor segmentation.

Finally, in this thesis, I tackle the challenge of accurately registering longitudinal neuroimaging. Registration is the most fundamental and core task in the neuroimaging analysis pipelines, and its accuracy often has a serious impact on an array of subsequent analysis steps and downstream tasks. In longitudinal neuroimaging analysis, rigid registration is commonly employed to align scans obtained from the same individual. However, the deep learning-based registration community has largely focused on cross-sectional registration applications, with insufficient accuracy observed in longitudinal data. For this reason, I optimize a state-of-the-art model in the fifth study to make it more accurate for longitudinal registration.

Abstract [sv]

Magnetresonanstomografi (MRT) har gjort det möjligt att observera hjärnans inre strukturer i realtid och erbjuder forskare en inblick i hur den mänskliga hjärnan utvecklas och åldras, samt de mekanismer som ligger till grund för neurologiska sjukdomar. Med den tekniska utvecklingen inom hjärnavbildning och ökningen av neuroimagingdata har efterfrågan på automatiserade algoritmer också vuxit. Den snabba utvecklingen inom djupinlärning och datorseende har lett till framsteg även vid dess tillämpning på hjärnavbildning. Medicinsk avbildning har dock sina egna specifika utmaningar, vilka hittills inte systematiskt har behandlats (removed duplicated "har") av generell forskning inom djupinlärning. I denna avhandling är målet att bemöta utmaningarna som uppstår inom tre specifika forskningsproblem inom neuroimaging. För varje problem kommer de specifika utmaningarna först att förklaras och därefter presenteras föreslagna lösningar.

De första studierna behandlar de specifika utmaningar som uppstår vid neuroimaging av Alzheimers patienter. Ett stort problem inom AD-forskning är att kunna särskilja effekterna av normalt åldrande och sjukdomsrelaterade förändringar på hjärnans anatomi för tidigare sjukdomsidentifiering. Longitudinell data är en viktig grund för att förstå hur normalt åldrande påverkar hjärnan. Insamlingen av sådana data är dock ofta mycket kostsam, och många öppna dataset har oregelbundna tidsintervall och avhopp från försökspersoner. I den första studien utvecklades därför en metod baserad på diffeomorfisk registrering för att fylla i saknade tidpunkter i longitudinella dataset. För att ytterligare kunna skilja mellan åldersrelaterade och sjukdomsrelaterade morfologiska förändringar presenterar den andra studien en generativ modell som förutsäger hjärnförändringar baserat på en enda MRI-resultat, enligt antingen frisk eller patologisk utveckling. Modellen är utformad med diffeomorfiska begränsningar för att garantera att de predikterade morfologiska förändringarna överensstämmer med hjärnans topologi. I den tredje studien introduceras och valideras istället ett nytt deformationsbaserat morfometri-framework som använder deformationsfält från registrering för att utföra jämförelser mellan olika grupper. Deformationen delas, i detta framework, upp i komponenter som kan tillskrivas normalt åldrande respektive patologiska förändringar vilket möjliggör en mer nyanserad förståelse av hur de interagerar.

Det andra problemområdet som avhandlingen behandlar är pediatrisk hjärntumörssegmentering. Annoterade dataset för pediatriska hjärntumörer är generellt små eftersom annoteringen kräver hög precision och kompetens vilket leder till att utvecklingen av övervakade segmenteringsalgoritmer är utmanande. I den fjärde studien föreslår vi istället en modell för domänanpassning med annoterad data för vuxna hjärntumörspatienter till den pediatriska kontexten. Modellen tränas utan annoterade pediatriska tumörer. 

Det tredje problemet som avhandlingen behandlar är registrering av longintudinell neuroimagingdata. Registrering är en grundläggande och central uppgift i neuroimaginganalys, och dess noggrannhet har ofta stor påverkan på efterföljande analyser. Vid analys av longitudinella neuroimagingdata kan rigid registrering ofta erhålla högre spatiell precision eftersom datan härrör från samma individ. Tidigare djupinlärningsmetoder för registrering har fokuserat på att registrera grupper av individer till ett gemensamt koordinatrum, vilket har visat sig ge otillräcklig noggrannhet vid longitudinell data. Den femte studien utvidgar därför en befintlig modell för att öka dess noggrannhet vid longitudinell registrering.

Place, publisher, year, edition, pages
Stockholm, Sweden: KTH Royal Institute of Technology, 2024. p. 71
Series
TRITA-CBH-FOU ; 2024:53
Keywords
Brain MRI, Image Registration, Machine Learning, Deep Learning, Aging, Alzheimer’s Disease, Unsupervised Domain Adaptation, Image Segmentation, Hjärn-MRT, Bildregistrering, Maskininlärning, Djupinlärning, Åldrande, Alzheimers sjukdom, Oövervakad domänanpassning, Bildsegmentering
National Category
Medical Imaging
Research subject
Technology and Health
Identifiers
urn:nbn:se:kth:diva-357014 (URN)978-91-8106-139-0 (ISBN)
Public defence
2025-01-10, T1, Hälsovägen 11C, via Zoom: https://kth-se.zoom.us/j/68696837704, Huddinge, 13:00 (English)
Opponent
Supervisors
Note

QC 2024-12-02

Available from: 2024-12-02 Created: 2024-12-01 Last updated: 2025-02-09Bibliographically approved

Open Access in DiVA

No full text in DiVA

Other links

Publisher's full textPubMedScopus

Authority records

Fu, JingruMoreno, Rodrigo

Search in DiVA

By author/editor
Fu, JingruMoreno, Rodrigo
By organisation
Medical Imaging
In the same journal
Human Brain Mapping
Radiology, Nuclear Medicine and Medical ImagingMedical Imaging

Search outside of DiVA

GoogleGoogle Scholar

doi
pubmed
urn-nbn

Altmetric score

doi
pubmed
urn-nbn
Total: 151 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf