Article
|
Open Access
Featured
-
-
Article
| Open AccessA machine and human reader study on AI diagnosis model safety under attacks of adversarial images
While active efforts are advancing medical AI model development and clinical translation, safety issues of medical AI models have emerged. Here, the authors investigate the effects on an AI model and on human experts of potential fake/adversarial images for breast cancer diagnosis.
- Qianwei Zhou
- , Margarita Zuley
- & Shandong Wu
-
Article
| Open AccessA deep learning framework identifies dimensional representations of Alzheimer’s Disease from brain structure
Alzheimer’s disease is heterogeneous in its neuroimaging and clinical phenotypes. Here the authors present a semi-supervised deep learning method, Smile-GAN, to show four neurodegenerative patterns and two progression pathways providing prognostic and clinical information.
- Zhijian Yang
- , Ilya M. Nasrallah
- & Balebail Ashok Raj
-
Article
| Open AccessArtificial intelligence system reduces false-positive findings in the interpretation of breast ultrasound exams
Ultrasound is an important imaging modality for the detection and characterization of breast cancer, but it has been noted to have high false-positive rates. Here, the authors present an artificial intelligence system that achieves radiologist-level accuracy in identifying breast cancer in ultrasound imaging.
- Yiqiu Shen
- , Farah E. Shamout
- & Krzysztof J. Geras
-
Article
| Open AccessAutomated bone mineral density prediction and fracture risk assessment using plain radiographs via deep learning
Dual-energy X-ray absorptiometry and the Fracture Risk Assessment Tool are recommended tools for osteoporotic fracture risk evaluation, but are underutilized. Here, the authors present an opportunistic tool to identify fractures, predict bone mineral density and evaluate fracture risk using plain pelvis and lumbar spine radiographs.
- Chen-I Hsieh
- , Kang Zheng
- & Chang-Fu Kuo
-
Article
| Open AccessDeep learning-based transformation of H&E stained tissues into special stains
Performing multiple histological stains on a biopsy can be costly and time consuming. Here the authors present a method for the digital transformation of H&E stained tissue into special stains (e.g., PAS, Masson’s Trichrome and Jones silver stain), and demonstrate that it improves diagnoses over the use of H&E only.
- Kevin de Haan
- , Yijie Zhang
- & Aydogan Ozcan
-
Article
| Open AccessAdvancing diagnostic performance and clinical usability of neural networks via adversarial training and dual batch normalization
Unmasking the decision making process of machine learning models is essential for implementing diagnostic support systems in clinical practice. Here, the authors demonstrate that adversarially trained models can significantly enhance the usability of pathology detection as compared to their standard counterparts.
- Tianyu Han
- , Sven Nebelung
- & Daniel Truhn
-
Article
| Open AccessReal-world stress resilience is associated with the responsivity of the locus coeruleus
Individuals vary considerably in how they are affected by stress. Here, the authors show that the severity of psychopathological symptoms triggered by prolonged real-life stress relate to fMRI-measured responsivity of the human brainstem arousal system and associated pupil responses.
- Marcus Grueschow
- , Nico Stenz
- & Birgit Kleim
-
Article
| Open AccessMulti-task weak supervision enables anatomically-resolved abnormality detection in whole-body FDG-PET/CT
Computational decision support systems could provide clinical value in whole-body FDG PET/CT workflows, but labeled data is scarce and PET/CT imaging exams are cumbersome. Here, the authors describe a weak supervision framework that extracts regional abnormality labels from free-text radiology reports.
- Sabri Eyuboglu
- , Geoffrey Angus
- & Matthew P. Lungren
-
Article
| Open AccessPredicting treatment response from longitudinal images using multi-task deep learning
Radiographic imaging is routinely used to evaluate treatment response in solid tumors. Here, the authors present a multi-task deep learning approach that allows simultaneous tumor segmentation and response prediction from longitudinal images in a multi-center study on rectal cancer.
- Cheng Jin
- , Heng Yu
- & Ruijiang Li
-
Article
| Open AccessAntibody-guided in vivo imaging of Aspergillus fumigatus lung infections during antifungal azole treatment
Invasive pulmonary aspergillosis is a life-threatening fungal lung disease devoid of specific rapid diagnosis and with limited therapeutic options. Here, the authors show how state-of-the-art imaging approaches can enable specific diagnosis and therapy monitoring of this infection.
- Sophie Henneberg
- , Anja Hasenberg
- & Nicolas Beziere
-
Article
| Open AccessEnsembled deep learning model outperforms human experts in diagnosing biliary atresia from sonographic gallbladder images
It is still challenging to make accurate diagnosis of biliary atresia (BA) with sonographic gallbladder images particularly in rural areas without relevant expertise. Here, the authors develop a diagnostic deep learning model which favourable performance in comparison with human experts in multi-center external validation.
- Wenying Zhou
- , Yang Yang
- & Luyao Zhou
-
Article
| Open AccessBedside functional monitoring of the dynamic brain connectivity in human neonates
Monitoring of cerebral function in human neonates remains challenging. Here, the authors propose a bedside monitoring technique using functional ultrasound to identify markers of cerebral activity based on intrinsic functional connectivity for early brain function monitoring.
- Jerome Baranger
- , Charlie Demene
- & Mickael Tanter
-
Article
| Open AccessA scalable physician-level deep learning algorithm detects universal trauma on pelvic radiographs
Pelvic radiographs (PXRs) are essential for detecting proximal femur and pelvis injuries in trauma patients, but none of the currently available algorithms can detect all kinds of trauma-related radiographic findings. Here, the authors develop a multiscale deep learning algorithm trained with weakly supervised point annotation.
- Chi-Tung Cheng
- , Yirui Wang
- & Le Lu
-
Article
| Open AccessDevelopment and evaluation of an artificial intelligence system for COVID-19 diagnosis
In some contexts, rapid detection of COVID-19 from CT scans can be crucial for optimal patient management. Here, the authors present a Deep Learning system for this task with multi-center data, human reader comparison and age stratified results.
- Cheng Jin
- , Weixiang Chen
- & Jianjiang Feng
-
Article
| Open AccessEarly prediction of disease progression in COVID-19 pneumonia patients with chest CT and clinical characteristics
Early identification of COVID-19 patients at risk of progression may facilitate more individually aligned treatment plans. Here the authors develop an online nomogram incorporating CT severity score and clinical characteristics for early predicting the disease progression risk among COVID-19 pneumonia patients.
- Zhichao Feng
- , Qizhi Yu
- & Wei Wang
-
Article
| Open AccessCeramic resonators for targeted clinical magnetic resonance imaging of the breast
Here, the authors present a concept for targeted clinical magnetic resonance imaging for relatively small targets in the body. They use an artificial resonator for spatial redistribution and passive focusing of the radiofrequency magnetic flux and demonstrate feasibility for targeted breast imaging.
- Alena Shchelokova
- , Viacheslav Ivanov
- & Anna Andreychenko
-
Article
| Open AccessA neurovascular high-frequency optical coherence tomography system enables in situ cerebrovascular volumetric microscopy
High resolution intravascular imaging in the brain is limited by the high tortuosity of the vasculature. Here the authors present a fiber optic imaging technology using high-frequency optical coherence tomography (HF-OCT) to provide volumetric high resolution images in the highly tortuous cerebral vasculature.
- Giovanni J. Ughi
- , Miklos G. Marosfoi
- & Ajit S. Puri
-
Article
| Open AccessDisentangling molecular alterations from water-content changes in the aging human brain using quantitative MRI
MRI has been used for the non-invasive imaging of human brain aging but is sensitive to both tissue molecular composition and water content. Here the authors present a quantitative MRI method that discriminates these factors and describe region-specific changes in the molecular composition of the aging human brain.
- Shir Filo
- , Oshrat Shtangel
- & Aviv A. Mezer
-
Article
| Open AccessA zwitterionic near-infrared fluorophore for real-time ureter identification during laparoscopic abdominopelvic surgery
Iatrogenic injury of the ureters is a feared complication of laparoscopic abdominal surgery. Here the authors present the NIR fluorophore ZW800-1 as an intraoperative imaging agent for ureter mapping, showing its safety, pharmacokinetic properties, and efficacy in healthy volunteers and patients undergoing abdominopelvic surgery.
- Kim S. de Valk
- , Henricus J. Handgraaf
- & Alexander L. Vahrmeijer
-
Article
| Open AccessWeakly supervised classification of aortic valve malformations using unlabeled cardiac MRI sequences
The availability of labelled training data is one of the practical obstacles towards wide application of machine learning models in medicine. Here the authors develop a weakly supervised deep learning model for the classification of aortic malformations using unlabelled cardiac MRI sequences from the UK biobank.
- Jason A. Fries
- , Paroma Varma
- & James R. Priest
-
Article
| Open AccessCortex-wide neural interfacing via transparent polymer skulls
Imaging the mouse brain using glass cranial windows has limitations in terms of flexibility and long-term imaging. Here the authors engineer transparent polymer skulls that can fit various skull morphologies and can be implanted for over 300 days, enabling simultaneous high resolution brain imaging and electrophysiology across large cortical areas.
- Leila Ghanbari
- , Russell E. Carter
- & Suhasa B. Kodandaramaiah
-
Article
| Open AccessWhy rankings of biomedical image analysis competitions should be interpreted with care
Biomedical image analysis challenges have increased in the last ten years, but common practices have not been established yet. Here the authors analyze 150 recent challenges and demonstrate that outcome varies based on the metrics used and that limited information reporting hampers reproducibility.
- Lena Maier-Hein
- , Matthias Eisenmann
- & Annette Kopp-Schneider
-
Article
| Open AccessSmartphone app for non-invasive detection of anemia using only patient-sourced photos
Anemia has a global prevalence of over 2 billion people and is diagnosed via blood-based laboratory test. Here the authors describe a smartphone app that can estimate hemoglobin levels and detect anemia by analyzing pictures of fingernail beds taken with a smartphone and without the need of any external equipment.
- Robert G. Mannino
- , David R. Myers
- & Wilbur A. Lam
-
Article
| Open AccessImplementation and benchmarking of a novel analytical framework to clinically evaluate tumor-specific fluorescent tracers
Fluorescent tracers are being tested in clinical trials to improve detection of tumor margins, but procedures are not standardised. Here, the authors develop an analytical framework that is compatible with the workflow in the operating theatre, and show that it leads to an 88% increase in intraoperative detection of tumor margins in patients with breast cancer.
- Marjory Koller
- , Si-Qi Qiu
- & Gooitzen M. van Dam
-
Article
| Open AccessSelf-decoupled radiofrequency coils for magnetic resonance imaging
Conventional coil arrays require complex decoupling technologies to reduce electromagnetic coupling between coil elements. Here, the authors report a self-decoupled RF coil design that achieves high inter-coil isolation between adjacent and non-adjacent elements and mixed arrays of loops and dipoles
- Xinqiang Yan
- , John C. Gore
- & William A. Grissom
-
Article
| Open AccessHeterogeneity in tumor chromatin-doxorubicin binding revealed by in vivo fluorescence lifetime imaging confocal endomicroscopy
The engagement of DNA-binding drugs to their target is difficult to study, particularly in vivo. Here the authors develop an in vivo fluorescence lifetime imaging confocal laparo/endomicroscope to show intra- and inter-tumor heterogeneity in doxorubicin binding to peritoneal metastases, which depends on the route of administration.
- Hugh Sparks
- , Hiroshi Kondo
- & Erik Sahai
-
Article
| Open AccessZika virus infection in pregnant rhesus macaques causes placental dysfunction and immunopathology
Zika virus infection during pregnancy can result in birth defects, but underlying pathogenesis at the maternal-fetal interface is unclear. Here, the authors use non-invasive in vivo imaging of Zika-infected rhesus macaques and show that infection results in abnormal oxygen transport across the placenta.
- Alec J. Hirsch
- , Victoria H. J. Roberts
- & Daniel N. Streblow
-
Article
| Open AccessPD-L1 is an activation-independent marker of brown adipocytes
Current approaches to visualise brown adipose tissue (BAT) rely primarily on markers that reflect its metabolic activity. Here, the authors show that PD-L1 is expressed on brown adipocytes, does not change upon BAT activation, and that BAT volume in mice can be measured by PET-CT with a radiolabeled anti-PD-L1 antibody.
- Jessica R. Ingram
- , Michael Dougan
- & Hidde L. Ploegh
-
Editorial
| Open AccessFostering reproducible fMRI research
The validity of conclusions drawn from functional MRI research has been questioned for some time now. Nature Neuroscience and Nature Communications are committed to working with neuroimaging researchers to improve the robustness and reproducibility of their work.
-
Article
| Open AccessCovalent assembly of nanoparticles as a peptidase-degradable platform for molecular MRI
Iron oxide microparticles (MPIO) are better MRI contrast agents than nanoparticles, but are of limited clinical use as they are not degradable and so risk toxicity. Here the authors present an iron oxide microparticle MRI contrast agent with peptide linkers that enable degradation into non-toxic nanoparticlesin vivo.
- Francisco Perez-Balderas
- , Sander I. van Kasteren
- & Benjamin G. Davis
-
Article
| Open AccessQuantifying cerebral contributions to pain beyond nociception
Pain is affected by cerebral processes in addition to afferent nociceptive input. Here the authors develop an fMRI-based signature that predicts pain independent of the intensity of nociceptive signals and mediates the pain-modulating effects of several cognitive interventions.
- Choong-Wan Woo
- , Liane Schmidt
- & Tor D. Wager
-
Article
| Open AccessDietary fatty acid metabolism of brown adipose tissue in cold-acclimated men
Brown adipose tissue (BAT) takes up and burns fatty acids for thermogenesis in mice. Here the authors use PET to show that, in humans, cold stimulation increases BAT dietary fatty acid uptake from plasma and oxidative metabolism, although, unlike mice, human BAT takes up less fatty acids than other metabolic tissues.
- Denis P. Blondin
- , Hans C. Tingelstad
- & André C. Carpentier
-
Article
| Open AccessTransportable hyperpolarized metabolites
MRI imaging can be significantly enhanced by injecting highly magnetized chemical agents, but the short magnetization lifetime requires processing at the point of use. Here, the authors demonstrate a method that could extend the lifetime from seconds to hours, enabling remote preparation.
- Xiao Ji
- , Aurélien Bornet
- & Sami Jannin
-
Article
| Open AccessMultiparametric imaging with heterogeneous radiofrequency fields
Magnetic resonance fingerprinting (MRF) requires a uniform B1+ radiofrequency field. Here the authors present plug-n-play MRF, a technique that enables multiparametric imaging with heterogeneous B1+fields, and demonstrate its utility in the presence of RF distortion caused by a metallic orthopaedic implant.
- Martijn A. Cloos
- , Florian Knoll
- & Daniel K. Sodickson
-
Article
| Open AccessAccurate spike estimation from noisy calcium signals for ultrafast three-dimensional imaging of large neuronal populations in vivo
Two-photon laser scanning microscopy allows functional calcium imaging of large neuronal populations in vivo, but the recorded signals typically suffer from low signal to noise. Here the authors develop an algorithm, MLspike, which estimates action potentials from noisy calcium signals, and benchmark it against existing methods.
- Thomas Deneux
- , Attila Kaszas
- & Ivo Vanzetta
-
Article
| Open AccessHigh-throughput dual-colour precision imaging for brain-wide connectome with cytoarchitectonic landmarks at the cellular level
High-throughput imaging methods for brain-wide connectome mapping with precise location reference have been lacking. Here authors report a method that allows simultaneous acquisition of fluorescently labelled neurons and cytoarchitectural landmarks in the same mouse brain at the single-cell resolution.
- Hui Gong
- , Dongli Xu
- & Qingming Luo
-
Article
| Open AccessHigh-throughput automated home-cage mesoscopic functional imaging of mouse cortex
Functional imaging in awake head-fixed mice is a widely used technique to study neural responses. Here the authors report on an open source, fully automated unsupervised system for training mice to self initiate head fixation to enable stable mesoscopic functional imaging of cortical functional connectivity.
- Timothy H. Murphy
- , Jamie D. Boyd
- & Jeff M. LeDue
-
Article
| Open Access4D Visualization of replication foci in mammalian cells corresponding to individual replicons
Whether replication happens at individual replicons or in replication factories is a controversial debate. Here the authors use super-resolution microscopy and analysis of replication fork speed to present evidence in favour of replicons.
- V. O. Chagin
- , C. S. Casas-Delucchi
- & M. C. Cardoso
-
Article
| Open AccessWhole-central nervous system functional imaging in larval Drosophila
To understand how neuronal networks function, it is important to measure neuronal network activity at the systems level. Here Lemon et al. develop a framework that combines a high-speed multi-view light-sheet microscope, a whole-CNS imaging assay and computational tools to demonstrate simultaneous functional imaging across the entire isolated Drosophilalarval CNS.
- William C. Lemon
- , Stefan R. Pulver
- & Philipp J. Keller
-
Article
| Open AccessLarge-scale genomics unveil polygenic architecture of human cortical surface area
How genetic variation contributes to brain morphology is still poorly understood. Here Chenet al. combine brain imaging with single-nucleotide polymorphism data to discover that a substantial degree of cortical variation is derived from underlying genetic differences.
- Chi-Hua Chen
- , Qian Peng
- & Anders M. Dale
-
Article |
Mesoscale infraslow spontaneous membrane potential fluctuations recapitulate high-frequency activity cortical motifs
The brain exists in a state of constant activity but little is known about very low frequency forms of activity. Here, the authors use high-speed, wide-field, voltage-sensitive dye imaging to investigate the presence and functional structure of infraslow spontaneous activity in anaesthetized and awake mouse cortex.
- Allen W. Chan
- , Majid H. Mohajerani
- & Timothy H. Murphy
-
Article
| Open AccessStereomicroscopic 3D-pattern profiling of murine and human intestinal inflammation reveals unique structural phenotypes
The gut epithelium is damaged in inflammatory bowel disease, but capturing such lesions by histology can be difficult. Here, the authors use stereomicroscopy to visualize different 3D inflammatory structures and associated microbes in humans and in 16 genetic mouse models relevant to intestinal inflammation.
- Alex Rodriguez-Palacios
- , Tomohiro Kodani
- & Fabio Cominelli
-
Article
| Open AccessIdentifying active vascular microcalcification by 18F-sodium fluoride positron emission tomography
Atherosclerotic plaques with macrocalcification are stable, whereas microcalcification is a key feature of rupture-prone plaques. Here the authors show that18F-NaF PET/CT imaging can distinguish between macro- and microcalcification providing a potential, non-invasive imaging technique to identify patients with high-risk atheroma.
- Agnese Irkle
- , Alex T. Vesey
- & Anthony P. Davenport
-
Article
| Open AccessDirect wavefront sensing for high-resolution in vivo imaging in scattering tissue
Direct wavefront sensing with laser ‘guide stars’ is used in astronomy and microscopy to correct for optical aberrations. Wang et al.use near-infrared guide stars to extend this approach to the highly scattering mouse brain, allowing high-resolution fluorescence imaging at 700μm depth.
- Kai Wang
- , Wenzhi Sun
- & Na Ji
-
Article |
Redox-responsive branched-bottlebrush polymers for in vivo MRI and fluorescence imaging
Multimodal and stimuli-responsive imaging agents can yield more biological information than more typical single-mode or inert imaging probes. Here, the authors have made a dual-modal MRI-fluorescence probe and demonstrate its ability to image redox status in vivo.
- Molly A. Sowers
- , Jessica R. McCombs
- & Jeremiah A. Johnson
-
Article |
A long Stokes shift red fluorescent Ca2+ indicator protein for two-photon and ratiometric imaging
Current calcium-sensitive probes based on red fluorescent proteins are unsuitable for two-photon excitation at the near-infrared wavelengths commonly used for green fluorescent probes. Wu et al. use a structure-guided approach to engineer a red fluorescent probe with optimal two-photon excitation at these wavelengths.
- Jiahui Wu
- , Ahmed S. Abdelfattah
- & Robert E. Campbell
-
Article
| Open AccessGraphene-based carbon-layered electrode array technology for neural imaging and optogenetic applications
Monitoring neuronal activity in the rodent in vivobrain is commonly done using micro-electrode arrays but these devices are not normally compatible with optical technologies. Here the authors design a transparent and flexible electrode array based on graphene that allows them to combine electrophysiological recordings with optogenetic and imaging experiments.
- Dong-Wook Park
- , Amelia A. Schendel
- & Justin C. Williams
-
Article |
Transparent and flexible low noise graphene electrodes for simultaneous electrophysiology and neuroimaging
Monitoring neuronal activity of large populations of neurons at high-temporal and spatial resolution is important to understand neurophysiology but requires improved tools and methods. Here the authors develop a transparent and flexible electrode based on graphene that allows them to combine electrophysiological recordings with calcium imaging.
- Duygu Kuzum
- , Hajime Takano
- & Brian Litt
-
Article
| Open AccessFunctional ultrasound imaging of intrinsic connectivity in the living rat brain with high spatiotemporal resolution
Functional connectivity of brain networks is poorly understood, in part, due to limited imaging approaches. Here, the authors use ultrasound imaging to study functional connectivity in the adult rat brain in vivo, allowing for the identification of highly contrasted intrinsic connectivity patterns.
- Bruno-Félix Osmanski
- , Sophie Pezet
- & Mickael Tanter