Research

Overview

The identification of neural circuit abnormalities associated with psychiatric symptoms is key for developing neural markers for early risk detection and ultimately for developing rationally-guided therapeutics. Traditional categorical approaches for psychiatric diagnoses, such as the DSM-5 criteria, often fail to align with individual clinical profiles and lack consideration of the neural mechanisms or circuits underlying specific symptoms, complicating treatment planning.

We use precision functional and anatomical neuroimaging measurements in humans, including fMRI, EEG and PET, to map neural circuit abnormalities underlying symptoms of mental illness. We integrate longitudinal assessments and time-series analyses to examine how neural signals relate to symptoms that manifest over various timescales. Using machine learning (ML) techniques (dimensionality reduction, multivariate linear models and AI-driven models, etc), we conduct multimodal neuroimaging studies combined with longitudinal phenotyping, computational modelling and neuroinformatics to establish reproducible brain-behavior associations.

The overarching goal of our group is to develop quantitative methods for testing hypotheses about the neurobiological mechanisms underlying specific cognitive and psychiatric symptoms. Despite recent advances in neuroimaging, actionable neural markers for specific symptoms remain scarce. Key challenges include overfitting, which limits the generalizability of neural-to-symptom models, and the high dimensionality of neuroimaging data. We aim to address these issues by developing robust feature reduction techniques to select features from neural and behavioral data, which can in turn facilitate the application of ML-based approaches in precision psychiatry.

By leveraging large datasets (e.g. Human Connectome Project-Early Psychosis, Biopolar-Schizophrenia Network on Intermediate Phenotypes [B-SNIP], Philadelphia Neurodevelopmental Cohort [PNC], and Adolescent Brain Cognitive Development [ABCD]) and acquiring new data for cross-sectional and longitudinal assessments, we will answer big-picture questions about the neurobiological mechanisms of psychiatric symptoms. Ultimately, our work will contribute to identify neural circuit abnormalities underlying psychiatric and cognitive symptoms, develop multimodal neural markers of psychopathology, and develop symptom-specific pharmacological treatment strategies.

Ongoing Projects

• Longitudinal Imaging of Schizophrenia in Adults (LISA)
• Neural dynamics modelling
• Brain states in fMRI and EEG
• Open-source BrainCAP software development

Research Interests

Innovations in Functional Neuroimaging Analysis

We develop methods and tools to study how brain functions are mapped across regions.

1. Lee K, Tak S, Ye JC. A data-driven sparse GLM for fMRI analysis using sparse dictionary learning with MDL criterion. IEEE Trans Med Imaging. 2011 May;30(5):1076-89. doi: 10.1109/TMI.2010.2097275. Epub 2010 Dec 6. PMID: 21138799.

2. Lee K, Lina JM, Gotman J, Grova C. SPARK: Sparsity-based analysis of reliable k-hubness and overlapping network structure in brain functional connectivity. Neuroimage. 2016 Jul 1;134:434-449. doi: 10.1016/j.neuroimage.2016.03.049. Epub 2016 Apr 2. PMID: 27046111.

3. Lee K, Khoo HM, Fourcade C, Gotman J, Grova C. Automatic classification and removal of structured physiological noise for resting state functional connectivity MRI analysis. Magn Reson Imaging. 2019 May;58:97-107. doi: 10.1016/j.mri.2019.01.019. Epub 2019 Jan 26. PMID: 30695721.

4. Lee YB, Lee J, Tak S, Lee K, Na DL, Seo SW, Jeong Y, Ye JC; Alzheimer’s Disease Neuroimaging Initiative. Sparse SPM: Group Sparse-dictionary learning in SPM framework for resting-state functional connectivity MRI analysis. Neuroimage. 2016 Jan 15;125:1032-1045. doi: 10.1016/j.neuroimage.2015.10.081. Epub 2015 Oct 31. PMID: 26524138.

Developing Neural Markers of Psychiatric Symptoms to Inform Personalized Treatments

We develop methods to identify neural circuit abnormalities, generate multimodal markers, and inform symptom-specific pharmacological strategies, focusing on psychosis and depression.

1. Lee K, Ji JL, Helmer M, Murray JD, Krystal JH, Anticevic A. A framework for advancing mechanistic neuro-behavioral biomarkers in psychiatry. Biol Psychiatry. 2025 Oct 8:S0006-3223(25)01488-X. doi: 10.1016/j.biopsych.2025.09.013. Epub ahead of print. PMID: 41072638.

2. Lee K, Ji JL, Fonteneau C, Berkovitch L, Rahmati M, Pan L, Repovš G, Krystal JH, Murray JD, Anticevic A. Human brain state dynamics are highly reproducible and associated with neural and behavioral features. PLoS Biol. 2024 Sep 24;22(9):e3002808. doi: 10.1371/journal.pbio.3002808. PMID: 39316635; PMCID: PMC11421804.

3. Berkovitch L, Lee K, Ji J, Helmer M, Rahmati M, Demsar J, Kraljic A, Matkovic A, Tamayo Z, Murray J, Repovs G, Krystal J, Martin W, Fonteneau C, Anticevic A. A common symptom geometry of mood improvement under sertraline and placebo associated with distinct neural patterns. Psychol Med. 2025 Jul 4;55:e185. doi: 10.1017/S0033291725100962. PMID: 40611472; PMCID: PMC12270277.

4. Rahmati M, Moujaes F, Suljič NP, Ji JL, Berkovitch L, Lee K, Fonteneau C, Schleifer CH, Adkinson B, Savič A, Santamauro N, Tamayo Z, Diehl C, Kolobaric A, Flynn M, Camarro T, Curtis CE, Repovš G, Fineberg SK, Morgan P, Preller KH, Krystal JH, Murray JD, Cho YT, Anticevic A, Ketamine alters tuning of neural and behavioral spatial working memory precision. bioRxiv, 2025. doi: 10.1101/2025.02.10.637233. Submitted.

Developing Neural Markers of Epilepsy, Sleep and Cognition

We use multimodal neuroimaging to identify neural mechanisms and develop markers of cognitive impairments.

1. Lee K, Khoo HM, Lina JM, Dubeau F, Gotman J, Grova C. Disruption, emergence and lateralization of brain network hubs in mesial temporal lobe epilepsy. Neuroimage Clin. 2018 Jun 30;20:71-84. doi: 10.1016/j.nicl.2018.06.029. PMID: 30094158; PMCID: PMC6070692.

2. Cross NE, Pomares FB, Nguyen A, Perrault AA, Jegou A, Uji M, Lee K, Razavipour F, Ali OBK, Aydin U, Benali H, Grova C, Dang-Vu TT. An altered balance of integrated and segregated brain activity is a marker of cognitive deficits following sleep deprivation. PLoS Biol. 2021 Nov 4;19(11):e3001232. doi: 10.1371/journal.pbio.3001232. PMID: 34735431; PMCID: PMC8568176.

3. Lee K*, Wang Y*, Cross NE, Jegou A, Razavipour F, Pomares FB, Perrault AA, Nguyen A, Aydin Ü, Uji M, Abdallah C, Anticevic A, Frauscher B, Benali H, Dang-Vu TT, Grova C. NREM sleep brain networks modulate cognitive recovery from sleep deprivation. bioRxiv [Preprint]. 2024 Jul 2:2024.06.28.601285. doi: 10.1101/2024.06.28.601285. PMID: 39005401; PMCID: PMC11244911. (*: co-first).

4.Dansereau CL, Bellec P, Lee K, Pittau F, Gotman J, Grova C. Detection of abnormal resting-state networks in individual patients suffering from focal epilepsy: an initial step toward individual connectivity assessment. Front Neurosci. 2014 Dec 23;8:419. doi: 10.3389/fnins.2014.00419. PMID: 25565949; PMCID: PMC4274904.

Metabolic and Physiological Signatures of Human Brain Networks

We integrate neuroimaging with physiological and behavioral recordings to study brain state and metabolism.

1. Lee K, Horien C, O’Connor D, Garand-Sheridan B, Tokoglu F, Scheinost D, Lake EMR, Constable RT. Arousal impacts distributed hubs modulating the integration of brain functional connectivity. Neuroimage. 2022 Sep;258:119364. doi: 10.1016/j.neuroimage.2022.119364. Epub 2022 Jun 9. PMID: 35690257; PMCID: PMC9341222.

2. Tak S, Jang J, Lee K, Ye JC. Quantification of CMRO(2) without hypercapnia using simultaneous near-infrared spectroscopy and fMRI measurements. Phys Med Biol. 2010 Jun 7;55(11):3249-69. doi: 10.1088/0031-9155/55/11/017. Epub 2010 May 17. PMID: 20479515.

3. Razavipour SF, Ali OBK, Lee K, Grimault S, Blinder S, Soucy J-P, Benali H, Gauthier CJ, Grova. Multiresolution metabolic profile of functional hubness in the resting human brain. 2023. Submitted.

Reproducible Mental Health Research with Open Science and Artificial Intelligence

We enhance reproducibility through open science and AI integration in neuroimaging.

1. Lee K*, Borghesani V*, de Moraes F, Olsen R, Kam JWY, Tzovara A, Badhwar A. Brain Mappers of Tomorrow: An international multilingual initiative for science dissemination. Aperture Neuro. 2024;4. doi: 10.52294/001c.123400. (*: co-first)

2. Kam JWY*, Badhwar A*, Borghesani V, Lee K, Noble S, Raamana PR, Ratnanather JT, Tan DGH, Oestreich LKL, Lee HW, Marzetti L, Nakua H, Rippon G, Olsen R, Pozzobon A, Uddin LQ, Yanes JA, Tzovara A. Creating diverse and inclusive scientific practices for research datasets and dissemination. Imaging Neurosci (Camb). 2024 Jul 12;2:imag-2-00216. doi: 10.1162/imag_a_00216. PMID: 40800266; PMCID: PMC12272200. (*: co-first)

3. Horien C, Noble S, Greene AS, Lee K, Barron DS, Gao S, O’Connor D, Salehi M, Dadashkarimi J, Shen X, Lake EMR, Constable RT, Scheinost D. A hitchhiker’s guide to working with large, open-source neuroimaging datasets. Nat Hum Behav. 2021 Feb;5(2):185-193. doi: 10.1038/s41562-020-01005-4. Epub 2020 Dec 7. PMID: 33288916; PMCID: PMC7992920.

4. Luo X, Rechardt A, Sun G, Nejad KK, Yáñez F, Yilmaz B, Lee K, Cohen AO, Borghesani V, Pashkov A, Marinazzo D, Nicholas J, Salatiello A, Sucholutsky I, Minervini P, Razavi S, Rocca R, Yusifov E, Okalova T, Gu N, Ferianc M, Khona M, Patil KR, Lee PS, Mata R, Myers NE, Bizley JK, Musslick S, Bilgin IP, Niso G, Ales JM, Gaebler M, Ratan Murty NA, Loued-Khenissi L, Behler A, Hall CM, Dafflon J, Bao SD, Love BC. Large language models surpass human experts in predicting neuroscience results. Nat Hum Behav. 2025 Feb;9(2):305-315. doi: 10.1038/s41562-024-02046-9. Epub 2024 Nov 27. PMID: 39604572; PMCID: PMC11860209.