Searched for: person:wisnit01 or marshk01 or Resibb01 or masura01 or chodoj01 or fausta04 or willic43 or hr18 or shepht01 or osorir01 or boutaa01 or mittem01 or spruit01 or shaoy01 or raoj03 or gurinl01 or ouedrs02 or nsr3 or jeanlg01 or fieree01 or madelg01 or storee01 or kenows01
active:yes
exclude-minors:true
Rep2Rep-Shuffle: Shuffled Repetition-to-Repetition Learning for Noise-Adaptive Self-Supervised MRI Denoising
Luo, Renqing; Janjušević, Nikola; Jain, Rajan; Pei, Haoyang; Wang, Yao; Madelin, Guillaume; Feng, Li
PURPOSE/OBJECTIVE:MRI applications with low SNR, such as non-proton imaging and low-field imaging, often require multiple signal averages to achieve adequate image quality at the cost of longer scan times. This work proposes Rep2Rep-Shuffle, a self-supervised learning framework that improves MRI denoising performance in repeated-acquisition imaging with limited training datasets. METHODS:Rep2Rep-Shuffle extends our previously proposed repetition-to-repetition learning (Rep2Rep)-based self-supervised denoising approach with a new shuffled repetition pairing strategy. Instead of using a fixed split between training input and target repetitions, different repetition combinations are randomly generated during training epochs for each dataset, which increases the effective training space when training datasets are limited. In addition, the framework also extends from slice-wise 2D to 3D denoising by exploiting correlations across all spatial dimensions. The proposed method was tested in vivo using sodium MRI and further validated using synthetic noisy MRI datasets. Different training strategies for both 2D and 3D denoising were compared. RESULTS:In vivo sodium MRI results demonstrated improved denoising and better preservation of anatomical structures with 3D Rep2Rep-Shuffle compared with fixed pairing (Rep2Rep) and 2D denoising. Reader evaluation further confirmed that 3D Rep2Rep-Shuffle received higher scores for different assessment categories. In synthetic experiments, 3D Rep2Rep-Shuffle achieved the best performance among self-supervised methods and approached the performance of supervised learning. CONCLUSION/CONCLUSIONS:Rep2Rep-Shuffle outperforms Rep2Rep for MRI denoising by increasing training diversity through shuffled repetition pairing and by leveraging volumetric spatial correlations. The method is particularly suited for low-SNR MRI applications that acquire multiple repetitions but have limited training datasets.
PMID: 42972488
ISSN: 1522-2594
CID: 6075443
Male breast cancer cases in the World Trade Center survivors
Alptekin, Ramazan; Durmus, Nedim; Wang, Ziyue; Ashebir, Yibeltal A; Degirmen, Serhat; Goren, Emre; Wilson, Leigh; Reibman, Joan; Shao, Yongzhao; Arslan, Alan A
BACKGROUND/UNASSIGNED:The collapse of the World Trade Center (WTC) towers on September 11, 2001 resulted in an unprecedented environmental disaster that released carcinogenic particulates and chemicals. Exposure to these toxicants has been associated with increased cancer incidence among responders and community members. While female breast cancer has been characterized in this population, male breast cancer (MBC) remains underexplored. OBJECTIVE/UNASSIGNED:To describe the demographic, clinical, and exposure characteristics of MBC cases among community members enrolled in the WTC Environmental Health Center (WTC EHC). METHODS/UNASSIGNED:Eligible participants included men with primary breast cancer who met WTC exposure criteria and provided informed consent from the WTC EHC as of December 31, 2024. Demographic, exposure, tumor characteristics and molecular subtypes were analyzed descriptively. RESULTS/UNASSIGNED:Among 1,575 WTC EHC breast cancer cases, 22 were male, with 16 consented for the analysis. The median age at diagnosis was 64 years, with a median latency of 15.4 years post-9/11. Most tumors were ductal, moderately differentiated, early stage (I-II), and Luminal A. Most patients were Non-Hispanic White and overweight or obese. CONCLUSIONS/UNASSIGNED:Although MBC is rare, its occurrence in an environmentally exposed civilian cohort highlights potential long-term effects of WTC-related toxicants and supports continued surveillance.
PMID: 42975957
ISSN: 2154-4700
CID: 6075451
International Retrospective Observational Study of Continual Learning for AI on Endotracheal Tube Placement from Chest Radiographs
Chen, Emma; Saenz, Agustina; Banerjee, Oishi; Marklund, Henrik; Zhang, Xiaoman; Johri, Shreya; Zhou, Hong-Yu; Luo, Luyang; Adithan, Subathra; Wu, Kay; Dogra, Siddhant; Janapa Reddi, Vijay; Buensalido, Dominic; Kavnoudias, Helen; Kloeckner, Roman; Müller, Lukas; Salinas-Miranda, Emmanuel; Veloza Vega, Maria José; Kolck, Johannes; Penzkofer, Tobias; Ueda, Daiju; Walston, Shannon L; Quispe-Cornejo, Armin Alvaro; Salvagno, Michele; Lee, Christopher; Fournier, Jonathan; Castillo, Rosa Patricia; Luna, Cibele; Bahramipour, Tara; Zuback, Amanda; Braren, Rickmer; Jiraskova, Petra; Wannasopha, Yutthaphan; Khumrin, Piyapong; Wei, Desmond Lim Shi Wei; Hallinan, James Thomas Patrick Decourcy; Jiao, Zhicheng; Yi, Thomas; Plasencia Martinez, Juana Maria; Casado Alarcon, Nuria Isabel; Fellner, Franz A; Niedermair, Julian F; Wu, Derek; Kim, Dongkeun; Haubold, Johannes; Heiliger, Lars; Pérez-Chada, Daniel; Pratesi, Pablo; Cummings, Ryan; Razavian, Narges; Oikonomou, Anastasia; Tran, William T; Küstner, Thomas; Afat, Saif; Dula, Adrienne N; Rousseau, Justin F; Hržić, Franko; Fuchsjäger, Michael; Rajpurkar, Pranav
BACKGROUND:Medical artificial intelligence (AI) models often underperform when deployed at new hospitals despite strong performance during development, creating a need for effective adaptation strategies that maintain institutional privacy. Continual learning (i.e., repeatedly retraining a model at each new hospital where it is deployed) can improve performance during and after deployment. METHODS:To determine whether continual learning could improve model performance across diverse clinical settings without requiring data exchange between hospitals or complex computational setups, we performed a retrospective study of a convolutional neural network model for endotracheal tube (ETT) assessment, comparing three deployment approaches (original model deployment, hospital-specific fine-tuning, and continual learning across hospitals). The study included 2313 intensive care unit chest radiographs across 23 hospitals spanning 12 countries and 5 continents, with 1 hospital held out for external validation. Participants were adults (≥18 years of age) with ETTs placed in 2021 (or earlier if a hospital found too few cases from 2021), identified through standardized key word searches in hospital systems and verified by radiology reports where available. Each hospital contributed approximately 100 anteroposterior radiographs, with 50 used for training, as well as associated deidentified reports and metadata. Each image was preprocessed by the contributing hospital to remove identifiable features and was independently reviewed and annotated by two radiologists. RESULTS:Continual learning models achieved a significantly lower average ETT-carina error of 10.58 mm (standard deviation [±SD], 12.98 mm), compared with 12.49 mm (SD±, 15.37 mm) for models fine-tuned only at their deployment hospital and 16.39 mm (±SD, 21.56 mm) for the original model. In addition, continual learning models outperformed the original model at all hospitals, and they outperformed models fine-tuned only at their deployment hospital at 21 of 22 hospitals. CONCLUSIONS:Despite using only 50 training data points from each hospital, continual learning surpassed traditional fine-tuning as a method for improving medical AI model generalization.
PMID: 42932524
ISSN: 2836-9386
CID: 6074723
Brain and Cardiometabolic Effects of Major Weight Loss in People With Obesity: Preliminary Cohort Findings
Gaggi, Naomi L; Hurwitz, Brian; Bernal, Jennifer; O'Brien, Courtney; Ren-Fielding, Christine; Lofton, Holly; Chen, Jenny; Saperstein, Aviezer; Lee, Sang Han; Chang, Yun-Juan; Yates, Kathy F; Fieremans, Els; Osorio, Ricardo; Convit, Antonio
OBJECTIVE/UNASSIGNED:Given the overlap between obesity-related metabolic and vascular abnormalities and Alzheimer's disease (AD) risk factors, it is critical to determine whether weight loss improves brain and cardiovascular health. This study examined whether ≥ 40% excess-weight loss is associated with improvements in cardiometabolic risk, brain structure, and cognition. METHODS/UNASSIGNED:The study was a longitudinal study of individuals with obesity undergoing metabolic and bariatric surgery or medical/behavioral weight-loss treatment. Participants were classified at follow-up as weight-loss (WL; ≥ 40% excess-weight loss) or non-weight-loss (nWL; < 40%). All completed medical evaluations, structural MRI, and cognitive testing at baseline and 1 year. RESULTS/UNASSIGNED:At 1 year, the WL group exhibited improvements in BMI, HbA1C, inflammatory, metabolic, and lipid profiles. MRI-based volumetrics revealed increases in subcortical gray-matter volume and preservation of AD-vulnerable regions. Cognitive performance remained unchanged. CONCLUSIONS/UNASSIGNED:Achieving ≥ 40% excess-weight loss was associated with improvements in metabolic health and brain structure, including favorable subcortical volume increases. These findings underscore the potential of sustained weight reduction to mitigate risks linked to metabolic syndromes and neurodegeneration. Longer-term follow-up is warranted to determine whether continued metabolic improvements lead to detectable cognitive gains and additional brain-health benefits beyond 1 year.
PMCID:13720752
PMID: 42964797
ISSN: 2055-2238
CID: 6074823
Nerve growth factor metabolic dysfunction in Down's syndrome brains
Iulita, M Florencia; Do Carmo, Sonia; Ower, Alison K; Fortress, Ashley M; Flores Aguilar, Lisi; Hanna, Michael; Wisniewski, Thomas; Granholm, Anne-Charlotte; Buhusi, Mona; Busciglio, Jorge; Cuello, A Claudio
Basal forebrain cholinergic neurons play a key role in cognition. This neuronal system is highly dependent on nerve growth factor (NGF) for its synaptic integrity and the phenotypic maintenance of its cell bodies. Basal forebrain cholinergic neurons progressively degenerate in Alzheimer's disease and Down's syndrome, and their atrophy contributes to the manifestation of dementia. Paradoxically, in Alzheimer's disease brains, the synthesis of NGF is not affected and there is abundance of the NGF precursor (proNGF). We have shown that this phenomenon is due to a deficit in NGF's extracellular metabolism which compromises proNGF maturation and exacerbates its subsequent degradation. We hypothesized that a similar imbalance should be present in Down's syndrome. Using a combination of quantitative reverse transcription-polymerase chain reaction, enzyme-linked immunosorbent assay, western blotting and zymography, we have investigated signs of NGF metabolic dysfunction in post-mortem brains from the temporal (n = 14), frontal (n = 34) and parietal (n = 20) cortex obtained from Down's syndrome subjects and age-matched controls (age range 31-68 years). We further examined primary cultures of human fetal Down's syndrome cortex (17-21 gestational age weeks). We report a significant increase in proNGF levels in human Down's syndrome brains, with a concomitant reduction in the levels of plasminogen and tissue plasminogen activator mRNA as well as an increment in neuroserpin expression; enzymes that partake in proNGF maturation. Down's syndrome brains also exhibited elevated zymogenic activity of MMP9, the major NGF-degrading protease. Our results indicate a failure in proNGF maturation in Down's syndrome brains and a likely enhanced proteolytic degradation of NGF, changes which can compromise the trophic support of basal forebrain cholinergic neurons. The alterations in proNGF and MMP9 were also present in cultures of Down's syndrome fetal cortex; suggesting that this trophic compromise may be amenable to rescue, before frank dementia onset. Our study thus provides a novel paradigm for cholinergic neuroprotection in Alzheimer's disease and Down's syndrome.
PMID: 42841319
ISSN: 1460-2156
CID: 6073965
Oscillometry Detects Early Lung Injury and Predicts Future Abnormal Spirometry in World Trade Center Survivors
Seok, Eunsil; Siu, Katherine; Wang, Yuyan; Wang, Ziyue; Lee, Myeonggyun; Oppenheimer, Beno W; Goldring, Roberta M; Shao, Yongzhao; Reibman, Joan; Liu, Mengling
PMID: 42853668
ISSN: 1535-4970
CID: 6074054
How to PERSEVERE: Study protocol for a national, virtual, randomized controlled trial of a peer mentor-supported educational intervention for family caregivers of people with Lewy body dementia
Ayele, Wdasie; Hemm, Jessica; Chen, Sarah Mitchell; Tharp, Marla E; Koliqi, Viosa; Mariani, Diane; Seshadri, Sandhya; Stevens, Evelyn; Fargo, Keith; Otto, Robin; Pensyl, Claire E; Manak, Thomas J; Bentley, Cintra D; Miller, Allegra; Voss, Margaret; Pierce, Carmen M; Ouyang, Bichun; Chodosh, Joshua; Fleisher, Jori E
BACKGROUND:Family caregivers of people living with Lewy body dementia (PLBD) experience significant strain and adverse health outcomes that consequently affect the health of their loved ones with Lewy Body Dementia (LBD). Effective LBD caregiver interventions are needed. Building on pilot work, our objective is to test the effects of PERSEVERE (Peer Mentor Support and Caregiver Education)- a 12-week peer mentor-supported, LBD-specific educational intervention -on caregiver knowledge, attitudes, strain, and PLBD quality of life and healthcare utilization in a national, virtual, randomized controlled trial. METHODS:Partnering with LBD caregiver advisors and advocacy organizations, we are recruiting and enrolling 180 seasoned caregivers (having cared for a PLBD with ≥3 years of cognitive symptoms) and 502 novice LBD family caregivers (<3 years of cognitive symptoms). Seasoned caregivers train as Mentors. Novice caregivers are randomized 1:1 in cohorts of 100 to (1) the Mentee group receiving weekly Mentor calls covering a novel LBD curriculum focused on symptom management, future planning, and caregiver support, or (2) the attention-control group receiving pre-existing resources for self-directed learning. Primary outcomes include change from baseline in caregiver knowledge and attitudes at 12 weeks (post-intervention) and caregiver strain at nine months (follow-up). Secondary outcomes include engagement with materials and/or mentor, satisfaction, future interest, and change in PLBD quality of life and PLBD healthcare utilization at nine months. DISCUSSION/CONCLUSIONS:A peer mentor-delivered LBD-focused curriculum may improve caregiver and PLBD outcomes in a cost-effective, scalable manner. PERSEVERE has experienced robust national enrollment and retention within the first 18 months of recruitment, with outcomes forthcoming. Trial Registration https://clinicaltrials.gov/study/NCT06389032, registered on April 24, 2024.
PMID: 42778090
ISSN: 1559-2030
CID: 6073546
Diet-Induced Hyperhomocysteinemia in C57BL/6 Mice: Simultaneous Evaluation of Plasma Homocysteine, Methionine, and Cysteine by LC-MS/MS and Assessment of Associated Oxidative Stress in Brain Homogenates
Valle, Maria Luisa; De Jesus, Christopher Lawrence; Getaneh, Bitseat; Erdjument-Bromage, Hediye; Loveland, James; William, Christopher; Arienzo, Donatello; Neubert, Thomas A; Rostagno, Agueda; Ghiso, Jorge
Hyperhomocysteinemia (HHcy), a risk factor for cardiovascular, neurodegenerative, and metabolic dysfunctions in humans, has been recapitulated in murine models by genetic and dietary modifications. The relevance of HHcy in clinical and experimental settings has led to the development of various analytical techniques with varying sensitivity and sample volume requirements. This study uses a fit-for-purpose, small-volume LC-MS/MS workflow to quantify plasma homocysteine (Hcy), methionine (Met), and cysteine (Cys) in a diet-induced HHcy mouse model. Using these assays that overcome limitations imposed by sample availability in small animals, our study evaluated circulating levels of these amino acids in C57BL/6 mice fed a HHcy-inducing diet rich in Met and lacking vitamins B6, B9, and B12. The increased Hcy and decreased Cys plasma concentrations observed in these mice correlated with decreased glutathione concentrations in brain homogenates and concomitant increase in the activity of the antioxidant enzyme Glutathione-S-Transferase and in the levels of the lipid peroxidation marker malondialdehyde. Overall, our work links plasma HHcy to the induction of brain oxidative stress, supporting the role of dietary modifications as complementary therapeutic strategies for neurodegenerative conditions.
PMCID:13603794
PMID: 42792201
ISSN: 2076-3921
CID: 6073592
Semaphorin 3G (SEMA3G) is a highly selective marker of cerebral amyloid angiopathy
Balcomb, Kaleah; Buchanan, Jessica; Strobbe, Aysha; Claus, Daphne; Faustin, Arline; Schneider, Julie; Wisniewski, Thomas; Sunde, Margaret; Drummond, Eleanor
Biomarkers of cerebral amyloid angiopathy (CAA) are critically needed. We recently identified semaphorin 3G (SEMA3G) as a novel protein selectively enriched in CAA. Here, we aimed to determine if SEMA3G was a selective marker of CAA in a large cohort of human brain tissue spanning multiple neurodegenerative diseases and three brain regions, and to determine if SEMA3G directly interacts with amyloid beta (Aβ). Multiplexed immunofluorescence showed that SEMA3G significantly accumulated only in CAA+ blood vessels in the brain in all cases. We also showed that SEMA3G preferentially associated with Aβ40, AβpS8 and AβpE3, but not Aβ42. Thioflavin T assays and transmission electron microscopy showed that SEMA3G directly interacted with Aβ and slowed Aβ aggregation in vitro, and that this effect was more pronounced for Aβ40 than Aβ42. Together our results demonstrate that SEMA3G is a highly specific marker of CAA in the brain.
PMCID:13596215
PMID: 42779634
ISSN: 2692-8205
CID: 6073549
Objective and Subjective Multidimensional Sleep Health in Women After Myocardial Infarction
Liu, Olivia C; Arabadjian, Milla; Hausvater, Anais; Park, Chorong; Shallcross, Amanda J; Kalinowski, Jolaade; Johnson, Dayna A; Reynolds, Harmony R; Spruill, Tanya M
BACKGROUND:Poor sleep after acute cardiovascular events is associated with increased mortality. Women are more likely than men to report sleep disturbances after cardiovascular events, but other sleep metrics linked to cardiometabolic health remain underexplored. OBJECTIVES/OBJECTIVE:We characterize multiple dimensions of sleep and examine factors associated with poor sleep among women with prior myocardial infarction (MI). METHODS:Participants completed 7 nights of wrist actigraphy, which measured sleep duration, sleep efficiency, and wake after sleep onset (WASO), and sleep diary, which measured self-reported sleep quality using a Likert scale. Multivariable regression modeled associations between sociodemographic, clinical, and psychosocial characteristics and sleep measures. RESULTS:Among 94 post-MI women (median days since MI 93; IQR: 69-709), mean age was 60 years, and 67.0% identified as non-Hispanic White. The mean sleep duration was 405.1 minutes (approximately 6.8 hours), sleep efficiency 86.3%, WASO 61.6 minutes, and self-reported sleep quality 3.3, indicating fair-to-good sleep quality. Overall, 64.9% of women had short (<7 hours) or long (>9 hours) sleep duration, 33% low sleep efficiency (<85%), and 88.3% long WASO (>30 minutes). In exploratory analyses, associations with poor sleep included identifying as other than non-Hispanic White (shorter sleep duration, lower efficiency, and longer WASO), being nonpartnered (shorter sleep duration), hypertension (lower sleep efficiency), and higher depressive symptoms (lower sleep quality). CONCLUSIONS:Many women with prior MI had suboptimal sleep (short or long sleep duration, low sleep efficiency, and/or long WASO). Our results highlight the importance of discussing multiple dimensions of sleep with women post-MI to optimize cardiovascular health.
PMID: 42748745
ISSN: 2772-963x
CID: 6072921