About

Paul A. Yushkevich, Ph.D., is a Professor of Radiology at the University of Pennsylvania's Perelman School of Medicine. His research portfolio combines theoretical work on statistical shape characterization using a symmetry-based representation of shape with applied and translational research in biomedical image analysis. He is particularly interested in analysis techniques tailored to specific anatomical structures, with key work involving automatic segmentation and morphometry of the hippocampal formation in magnetic resonance imaging (MRI). The hippocampal formation plays a central role in memory function and is a site of early neurodegeneration in Alzheimer’s disease. Dr. Yushkevich has developed a detailed computational atlas of the hippocampal formation from postmortem MRI microscopy, which is integrated with histology for analysis of the hippocampal subfields in in vivo MRI. His team has pioneered techniques for automatic segmentation of hippocampal subfields in high-resolution, hippocampal-focused T2-weighted in vivo MRI, demonstrating excellent agreement with manual segmentation. His research also includes strategies to improve segmentation accuracy using machine learning, surface-based tract-specific analysis for diffusion MRI, and automatic cardiac MRI segmentation with explicit priors. His work aims to develop imaging-based biomarkers for neurodegenerative diseases, particularly Alzheimer’s disease, and he has contributed to advancing the understanding of neurodegeneration through innovative imaging analysis methods.

Research topics

• Medicine
• Psychology
• Pathology
• Neuroscience
• Artificial Intelligence
• Biology
• Computer Science
• Internal medicine
• Oncology
• Radiology
• Genetics
• Nuclear medicine
• Chemistry
• Audiology
• Database

Selected publications

• Automatic Segmentation of Medial Temporal Lobe Subregions in Multi‐Scanner, Multi‐Modality Magnetic Resonance Imaging of Variable Quality

  Hippocampus · 2025-10-07

  articleOpen accessSenior author

  Volumetry of subregions in the medial temporal lobe (MTL) computed from automatic segmentation in MRI can track neurodegeneration in Alzheimer's disease. However, poor quality MR images can lead to unreliable segmentation of MTL subregions. Considering that different MRI contrast mechanisms and field strengths (jointly referred to as "modalities" here) offer distinct advantages in imaging different parts of the MTL, we developed a multi-modality segmentation model using both 7T and 3T structural MRI to obtain robust segmentation in poor-quality images. MRI modalities including 3T T1-weighted, 3T T2-weighted, 7T T1-weighted and 7T T2-weighted (7T-T2w) of 197 participants were collected from a longitudinal aging study at the Penn Alzheimer's Disease Research Center. Among them, 7T-T2w was used as the primary modality, and all other modalities were rigidly registered to the 7T-T2w. A model derived from nnU-Net took these registered modalities as input and outputted subregion segmentation in 7T-T2w space. 7T-T2w images most of which had high quality from 25 selected training participants were manually segmented to train the multi-modality model. Modality augmentation, which randomly replaced certain modalities with Gaussian noise, was applied during training to guide the model to extract information from all modalities. The multi-modality model delivered good performance regardless of 7T-T2w quality, while the single-modality model under-segmented subregions in poor-quality images. The multi-modality model generally demonstrated stronger discrimination of A + MCI versus A-CU. Intra-class correlation and Bland-Altman plots demonstrate that the multi-modality model had higher longitudinal segmentation consistency in all subregions while the single-modality model had low consistency in poor-quality images. The multi-modality MRI segmentation model provides an improved biomarker for neurodegeneration in the MTL that is robust to image quality. It also provides a framework for other studies which may benefit from multimodal imaging.

  PublisherDOI

• Temporal Modeling of Amyloid and Tau Trajectories in Alzheimer’s Disease using PET and Plasma Biomarkers

  medRxiv · 2025-09-07 · 1 citations

  preprintOpen access

  ABSTRACT Objective To compare PET and plasma-based temporal modeling of amyloid and tau biomarkers in Alzheimer’s disease Methods Longitudinal amyloid PET, 18 F-flortaucipir tau-PET, and Fujirebio Lumipulse plasma p-tau 217 from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) and University of Pennsylvania Alzheimer’s Disease Research Center (Penn ADRC) were used to generate biomarker trajectory models using Sampled Iterative Local Approximation (SILA). SILA models using plasma p-tau 217 were compared to amyloid and tau PET-based models to estimate tau onset age (ETOA) and estimate amyloid onset age (EAOA), and factors influencing ETOA and time from ETOA to dementia were evaluated for PET and plasma-based models. Results Plasma-based models generated similar results to PET for EAOA and ETOA, with stronger model agreement for ETOA than EAOA. Accuracy of estimated onset age compared to actual onset age was high within modality with slightly greater error when comparing across modalities (i.e. plasma to PET). For both plasma and PET models, earlier ETOA was associated with younger EAOA, female sex, and ≥1 ApoE ε4 allele. Earlier dementia onset after ETOA was associated with later ETOA for both plasma and PET models, while male sex was associated with shorter tau to dementia gap in plasma models. Interpretation Temporal modeling of plasma biomarkers provides comparable information to PET-based models, particularly for tau onset age. Plasma-based temporal modeling can serve as a widely accessible tool for clinical assessment of biological disease duration that places the patient on the disease timeline, which may allow for improved discussion of prognosis and treatment decisions.

  PublisherOA PDFDOI

• Harmonized Protocol for Subfield Segmentation in the Hippocampal Body on High-Resolution <i>in vivo</i> MRI from the Hippocampal Subfields Group (HSG)

  bioRxiv (Cold Spring Harbor Laboratory) · 2025-05-04 · 5 citations

  preprintOpen access

  Abstract Hippocampal subfields differentially develop and age, and they vary in vulnerability to neurodegenerative diseases. Innovation in high-resolution imaging has accelerated clinical research on human hippocampal subfields, but substantial differences in segmentation protocols impede comparisons of results across laboratories. The Hippocampal Subfields Group (HSG) is an international organization seeking to address this issue by developing a histologically-valid, reliable, and freely available segmentation protocol for high-resolution T 2 -weighted 3 tesla MRI ( http://www.hippocampalsubfields.com ). Here, we report the first portion of the protocol focused on subfields in the hippocampal body; protocols for the head and tail are in development. The body protocol includes definitions of the internal boundaries between subiculum, Cornu Ammonis (CA) 1-3 subfields, and dentate gyrus, in addition to the external boundaries of the hippocampus apart from surrounding white matter and cerebrospinal fluid. The segmentation protocol is based on a novel histological reference data set labeled by multiple expert neuroanatomists. With broad participation of the research community, we voted on the segmentation protocol via online survey, which included detailed protocol information, feasibility testing, demonstration videos, example segmentations, and labeled histology. All boundary definitions were rated as having high clarity and reached consensus agreement by Delphi procedure. The harmonized body protocol yielded high inter- and intra-rater reliability. In the present paper we report the procedures to develop and test the protocol, as well as the detailed procedures for manual segmentation using the harmonized protocol. The harmonized protocol will significantly facilitate cross-study comparisons and provide increased insight into the structure and function of hippocampal subfields across the lifespan and in neurodegenerative diseases.

  PublisherOA PDFDOI

• Imaging Biomarkers for Neurodegenerative Diseases from Detailed Segmentation of Medial Temporal Lobe Subregions on in vivo Brain MRI Using Upsampling Strategy Guided by High-resolution ex vivo MRI

  ArXiv.org · 2025-04-25

  preprintOpen accessSenior author

  The medial temporal lobe (MTL) is a region impacted extensively and non-uniformly in early stages of Alzheimer's disease (AD). Regional MTL morphometric measures extracted from magnetic resonance imaging (MRI) are supportive features for the diagnosis of AD and related disorders (ADRD). Different MRI modalities have distinct advantages for MTL morphometry. Anisotropic T2-weighted (T2w) MRI is preferred for hippocampal subfields due to its higher contrast between hippocampal layers. Isotropic T1-weighted (T1w) MRI is beneficial for thickness calculation of extra-hippocampal subregions due to its stable image quality and isotropic resolution. We propose a multi-modality MTL segmentation algorithm that bridges the T1w and T2w modalities by bringing both to a nearly isotropic voxel space. Guided by high-resolution ex vivo 9.4T MRI, an upsampling model was designed for the ground truth segmentations. Combined with non-local means upsampling, this model was used to construct a nearly iso-tropic T1w and T2w MTL subregion segmentation training set, which was used to train a nnUNet model. Morphometric biomarkers extracted by this model were compared to those extracted using conventional models operating in anisotropic spaces on downstream tasks. Biomarkers extracted using the proposed model had greater ability to discriminate between individuals with mild cognitive impairment and cognitively unimpaired; and had great-er longitudinal stability. These findings suggest that the biomarkers derived from T1w and T2w MRI unsampled to nearly isotropic resolution have sig-nificant potential for improving disease diagnosis and monitoring disease progression in ADRD.

  PublisherOA PDFDOI

• Operationalizing postmortem pathology-MRI association studies in Alzheimer’s disease and related disorders with MRI-guided histology sampling

  Acta Neuropathologica Communications · 2025-05-28 · 3 citations

  articleOpen accessSenior author

  Postmortem neuropathological examination, while the gold standard for diagnosing neurodegenerative diseases, often relies on limited regional sampling that may miss critical areas affected by Alzheimer's disease and related disorders. Ultra-high resolution postmortem MRI can help identify regions that fall outside the diagnostic sampling criteria for additional histopathologic evaluation. However, there are no standardized guidelines for integrating histology and MRI in a traditional brain bank. We developed a comprehensive protocol for whole hemisphere postmortem 7T MRI-guided histopathological sampling with whole-slide digital imaging and histopathological analysis, providing a reliable pipeline for high-volume brain banking in heterogeneous brain tissue. Our method uses patient-specific 3D printed molds built from postmortem MRI, allowing standardized tissue processing with a permanent spatial reference frame. To facilitate pathology-MRI association studies, we created a semi-automated MRI to histology registration pipeline and developed a quantitative pathology scoring system using weakly supervised deep learning. We validated this protocol on a cohort of 29 brains with diagnosis on the AD spectrum that revealed correlations between cortical thickness and phosphorylated tau accumulation. This pipeline has broad applicability across neuropathological research and brain banking, facilitating large-scale studies that integrate histology with neuroimaging. The innovations presented here provide a scalable and reproducible approach to studying postmortem brain pathology, with implications for advancing diagnostic and therapeutic strategies for Alzheimer's disease and related disorders.

  PublisherOA PDFDOI

• Medial temporal lobe Tau-Neurodegeneration <i>mismatch</i> from MRI and plasma biomarkers identifies vulnerable and resilient phenotypes with AD

  medRxiv · 2025-08-19 · 1 citations

  preprintOpen access

  Abstract While tau pathology is closely associated with neurodegeneration in Alzheimer’s disease (AD), our prior work using multi-modality imaging revealed that mismatch between tau (T) and neurodegeneration (N) may reflect contributions from non-AD processes. The medial temporal lobe (MTL), an early site of AD pathology, is also a common target of co-pathologies such as limbic-predominant age-related TDP-43 encephalopathy neuropathologic change (LATE-NC), often following an anterior–posterior atrophy gradient. Given the susceptibility of MTL to co-pathologies, here we explored T-N mismatch specifically within MTL using plasma ptau 217 and MTL morphometry for identifying vulnerabilities and resilience in cognitively impaired or unimpaired AD patients. We parcellated the MTL into 100 spatially contiguous segments and calculated their T-N mismatch using plasma ptau 217 as a measure for T and thickness as a marker of N. Based on these mismatch profiles, we clustered 447 amyloid-positive individuals from ADNI cohort into data-driven T-N phenotypes. We characterized the T-N phenotypes by examining their cross-sectional and longitudinal atrophy both within the MTL and across the whole brain, as well as cognitive trajectories. This framework was replicated in an independent cohort and finally translated to a real-world clinical sample of 50 patients undergoing anti-amyloid therapy. Clustering identified three T-N phenotypes with different MTL T-N mismatch profiles, atrophy patterns, and cognitive outcomes, despite comparable AD severity. The “canonical” group, characterized by low T-N residuals (N ∼ T), showed AD-like neurodegeneration patterns. The “vulnerable” group, characterized by disproportionately greater neurodegeneration than tau (N &gt; T), showed atrophy primarily in the anterior MTL that extended into temporal-limbic regions, both in cross-sectional and longitudinal analyses. This group also exhibited neurodegeneration that preceded estimated tau onset and experienced faster cognitive decline across multiple domains, aligning with the typical characteristics of mixed LATE-NC with AD. In contrast, the “resilient” group (N &lt; T) showed minimal atrophy and preserved cognitive function. These phenotypes were reproducible in an independent research cohort. Importantly, in a feasibility study applying the model developed from ADNI to a clinical cohort of patients receiving lecanemab, we identified vulnerable individuals with LATE-like atrophy patterns. This highlights its potential utility for identifying individuals with co-pathology in clinical settings. Our findings demonstrate that T-N mismatch within MTL using MRI and plasma biomarkers can reveal AD groups with varying vulnerability/resilience, with the vulnerable group displaying structural and cognitive outcomes suggestive of LATE-NC. This approach offers a cost-effective strategy for clinical trial stratification and precision medicine for AD therapeutics.

  PublisherOA PDFDOI

• Effects of APOE Genotypes on Periventricular White Matter Cerebral Blood Flow in Cognitively Intact Adults (P3-3.004)

  Neurology · 2025-04-07

  article

  This study aimed to investigate the association between APOE genotype and cerebral blood flow (CBF) in periventricular white matter (PVWM) as a potential marker of small vessel disease (SVD) in cognitively intact (CI) individuals.

  PublisherDOI

• Harmonized Protocol for Segmentation of the Hippocampal Tail on High-Resolution <i>in vivo</i> MRI from the Hippocampal Subfields Group (HSG)

  bioRxiv (Cold Spring Harbor Laboratory) · 2025-11-20

  preprintOpen access

  The hippocampus is a heterogeneous structure with cytoarchitectonically distinct subfields that exhibit heterogeneous lifespan trajectories and are differentially susceptible to diseases. Advances in high-resolution imaging have accelerated research on these structures, yet variability in segmentation protocols limits cross-study comparability. The Hippocampal Subfields Group (HSG) is an international consortium addressing this challenge by developing a reliable, accessible, and freely available segmentation protocol for high-resolution T2-weighted 3 tesla MRI scans (http://www.hippocampalsubfields.com). Here, we present the harmonized protocol for the posterior portion of the hippocampus (the "tail"), complementing the previously established "body" protocol, and with an anterior "head" protocol under development. The tail protocol provides standardized definitions of the external boundaries for the posterior-most extent of the hippocampus, facilitating consistent segmentation from surrounding tissues. The research community was extensively involved through an online survey that incorporated comprehensive protocol details, feasibility assessments, tutorial videos, and illustrative segmentations. Through this collaborative process, consensus emerged to exclude subfield labeling in the hippocampal tail due to limited visibility of internal landmarks and substantial anatomical variability in this region. All proposed boundary guidelines were deemed clear and agreed upon via a Delphi procedure. The harmonized tail protocol has high intra- (Averaged ICC(2,1) > 0.98; Averaged Dice Similarity Coefficient = 0.92) and inter-rater reliability (Averaged ICC(2,k) > 0.98; Averaged Dice Similarity Coefficient = 0.86) and offers a practical framework for replicable segmentation. By establishing standardized guidelines, this protocol enhances comparability of findings across developmental, aging, and clinical research and is compatible with ongoing technological advances.

  PublisherOA PDFDOI

• Challenges and best practices when using ComBAT to harmonize diffusion MRI data

  Scientific Reports · 2025-11-24 · 3 citations

  articleOpen access

  Over the years, ComBAT has become the standard method for harmonizing MRI-derived measurements, with its ability to compensate for site-related additive and multiplicative biases while preserving biological variability. However, ComBAT relies on a set of assumptions that, when violated, can result in flawed harmonization. In this paper, we thoroughly review ComBAT's mathematical foundation, outlining these assumptions, and exploring their implications for the demographic composition necessary for optimal results. Through a series of experiments involving a slightly modified version of ComBAT called Pairwise-ComBAT tailored for normative modeling applications, we assess the impact of various population characteristics, including population size, age distribution, the absence of certain covariates, and the magnitude of additive and multiplicative factors. Based on these experiments, we present five essential recommendations that should be carefully considered to enhance consistency and supporting reproducibility, two essential factors for open science, collaborative research, and real-life clinical deployment.

  PublisherOA PDFDOI

• Relationships of PGRN with sTREM2 in AD continuum and non-AD pathophysiology and their reciprocal roles in modulating amyloid pathology: two population-based study

  Translational Psychiatry · 2025-07-08 · 3 citations

  articleOpen access

  . These proteins are mainly enriched in immune processes and neural plasticity. These findings suggest that the interplay between lysosome function and microglia-related neuroinflammation plays key roles in amyloid metabolism.

  PublisherDOI

Recent grants

• NIH Grant K25AG027785

  NIH · $854k · 2012

• NIH Grant R01EB014346

  NIH · $2.0M · 2017

• NIH Grant R03EB008200

  NIH · $158k · 2010

• Ex Vivo Imaging of the Aging Brain to Discover Morphology/Pathology Associations

  NIH · $2.1M · 2023–2026

• NIH Grant R21NS061111

  NIH · $379k · 2010

Frequent coauthors

• David A. Wolk

  California University of Pennsylvania

  282 shared

• Sandhitsu R. Das

  Alzheimer’s Disease Neuroimaging Initiative

  275 shared

• Robin de Florès

  Centre Hospitalier Universitaire de Caen

  216 shared

• Laura Wisse

  Lund University

  211 shared

• Long Xie

  Siemens Healthcare (United States)

  144 shared

• Ranjit Ittyerah

  University of Pennsylvania

  117 shared

• David J. Irwin

  University of Pennsylvania

  91 shared

• Murray Grossman

  87 shared

Labs

• PICSL LabPI