Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Volume 1 Issue 4, April 2022

Deep learning predicts survival after ischemic heart disease

Popescu et al. develop a deep learning approach that combines neural networks and survival analysis from cardiac magnetic resonance images and clinical covariates for patients with ischemic heart disease to predict arrhythmic sudden death.

See Popescu et al. and News & Views by Krittanawong

Image: Kimberly Koury Popescu Cover Design: Bethany Vukomanovic

Comment & Opinion

  • After three decades of work, hypothesis-generating genomic approaches have led to the identification of several intracranial aneurysm risk loci and Mendelian mutations, involving several unexpected genes. These findings opened the door for exciting opportunities, unraveling the genomic architecture of brain aneurysms. The field is now ripe to face the next set of surprises in this long journey.

    • Tanyeri Barak
    • Murat Günel
    Comment

    Advertisement

Top of page ⤴

Research Highlights

Top of page ⤴

News & Views

  • A combined imaging–clinical risk prediction model with the use of deep learning seems a promising approach for predicting sudden cardiac death in patients with ischemic cardiomyopathies. Deep-learning-guided clinical trials will be needed to translate this model into clinical practice.

    • Chayakrit Krittanawong
    News & Views
  • Loss of transcription factor SMAD3 changes the smooth muscle cell (SMC) to a unique remodeling SMC phenotype and points to a potential role for SMAD3 in the inhibition of macrophage recruitment and outward remodeling of the aortic wall.

    • Marie Jose Goumans
    • Paul H. A. Quax
    News & Views
  • Genome-wide association studies of magnetic resonance imaging (MRI) of diastolic heart function shed light on the underlying molecular mechanisms and support a causal role of diastolic function for the development of heart failure.

    • J. Gustav Smith
    • Olof Gidlöf
    News & Views
Top of page ⤴

Research Briefings

  • Global or macrophage-specific knockout of Trpm2, which encodes the calcium-permeable ion channel TRPM2, protects mice against atherosclerosis induced by a high-fat diet. Mechanistically, activation of TRPM2 and the scavenger receptor CD36 promote the transformation of macrophages into inflammatory foam cells, thereby accelerating the development and progression of atherosclerosis.

    Research Briefing
Top of page ⤴

Reviews

  • In this Review, Banning and colleagues summarize diagnostic techniques that assess microcirculation in each specific organ, critically appraise all the evidence that supports the systemic and multi-organ nature of microvascular dysfunction and focus on current and emerging interventions for the treatment of microvascular dysfunction.

    • Maria Emfietzoglou
    • Dimitrios Terentes-Printzios
    • Adrian P. Banning
    Review Article
  • In this Review, Bowers et al. discuss how the development of therapeutics to combat cardiac diseases, specifically fibrosis, relies on a deeper understanding of how the cardiac extracellular matrix is intertwined with signaling processes that underlie cardiac cell activation and behavior.

    • Stephanie L. K. Bowers
    • Qinghang Meng
    • Jeffery D. Molkentin
    Review Article
Top of page ⤴

Research

  • Cheng et al. show that smooth muscle cell (SMC)-specific deletion of Smad3 influences the fate of de-differentiated SMCs in atherosclerotic plaques in vivo, promoting both a pro-remodeling SMC transition phenotype and expansion of the SMC-derived chondromyocyte population. These cellular changes are associated with increased outward remodeling and plaque calcification.

    • Paul Cheng
    • Robert C. Wirka
    • Thomas Quertermous
    Article
  • Popescu et al. developed a deep learning approach that blends neural networks and survival analysis to predict patient-specific survival curves from raw contrast-enhanced cardiac magnetic resonance images and clinical covariates for patients with ischemic heart disease to offer accurate arrhythmic sudden death predictions.

    • Dan M. Popescu
    • Julie K. Shade
    • Natalia A. Trayanova
    Article Open Access
  • Zong and colleagues reveal a critical role for the ion channel TRPM2 in macrophages through mediating reactive oxygen species production, inflammasome activation, oxidized LDL uptake and subsequently inflammatory responses, which they show is mediated by CD36 activity, thereby establishing a mutually regulating and positive feedback mechanism between CD36 and TRPM2 in atherogenesis.

    • Pengyu Zong
    • Jianlin Feng
    • Lixia Yue
    Article
  • O’Regan and colleagues use deep-learning cardiac motion analysis in participants of the UK Biobank to measure diastolic functional traits and perform a genome-wide association study to generate insights into the genetic and environmental factors that influence diastolic function.

    • Marjola Thanaj
    • Johanna Mielke
    • Declan P. O’Regan
    Article Open Access
  • Huang et al. show that myocardial infarction (MI)-associated vasculature is structurally and functionally abnormal, impeding vessel function and cardiac repair in mice. Analyses of the transcriptome of the cardiac endothelium after MI identify a PDGF–NF-κB–HIF-1α Snail axis responsible for mesenchymal transformation of endothelial cells and show that genetic ablation or targeted disruption of PDGF signaling normalizes vasculature and improves cardiac function recovery after MI.

    • Menggui Huang
    • Fan Yang
    • Yanqing Gong
    Article
  • Despite an emerging role for cerebrovascular endothelial cells in a range of neurological pathologies, AAV vector development to date has focused on tools designed to target neurons or astrocytes. Here, Krolak et al. describe a specific variant of AAV (AAV-BI30), with high specificity and efficacy for transduction of endothelial cells across the central nervous system.

    • Trevor Krolak
    • Ken Y. Chan
    • Benjamin E. Deverman
    Technical Report
Top of page ⤴

Amendments & Corrections

Top of page ⤴

Search

Quick links