A5043623805- Radiomics and Machine Learning in Medical Imaging
- MRI in cancer diagnosis
- Artificial Intelligence in Healthcare and Education
- Acute Ischemic Stroke Management
- Advanced X-ray and CT Imaging
- Cerebrovascular and Carotid Artery Diseases
- AI in cancer detection
- Meningioma and schwannoma management
- Advanced MRI Techniques and Applications
- IgG4-Related and Inflammatory Diseases
- Ultrasound and Hyperthermia Applications
- Glioma Diagnosis and Treatment
- Advanced Neuroimaging Techniques and Applications
- Ultrasound Imaging and Elastography
- Multiple Sclerosis Research Studies
- Vascular Malformations Diagnosis and Treatment
- Systemic Sclerosis and Related Diseases
- Bone Tumor Diagnosis and Treatments
- Glaucoma and retinal disorders
- Ocular Diseases and Behçet’s Syndrome
- Oral and Maxillofacial Pathology
- Retinal and Optic Conditions
- Venous Thromboembolism Diagnosis and Management
- Salivary Gland Tumors Diagnosis and Treatment
- Neuroendocrine Tumor Research Advances
Fondation Ophtalmologique Adolphe de Rothschild
2018-2025
Université Paris Cité
2019-2025
Paris Cardiovascular Research Center
2019-2025
Inserm
2019-2025
Fondation de Rothschild
2019-2025
Assistance Publique – Hôpitaux de Paris
2016-2022
Sorbonne Paris Cité
2019-2022
Délégation Paris 5
2019
Sorbonne Université
2014-2019
Hôpital Foch
2017
Objectives To assess the influence of gray-level discretization on inter- and intra-observer reproducibility texture radiomics features clinical MR images. Materials methods We studied two independent MRI datasets 74 lacrymal gland tumors 30 breast lesions from different centers. Two pairs readers performed three two-dimensional delineations for each dataset. Texture were extracted using softwares (Pyradiomics an in-house software). Reproducible selected a combination intra-class correlation...
Background The interpretation of radiographs suffers from an ever-increasing workload in emergency and radiology departments, while missed fractures represent up to 80% diagnostic errors the department. Purpose To assess performance artificial intelligence (AI) system designed aid radiologists physicians detection localization appendicular skeletal fractures. Materials Methods AI was previously trained on 60 170 obtained patients with trauma. were randomly split into 70% training, 10%...
Over a third of minor stroke patients experience post-stroke cognitive impairment (PSCI), but no validated tools exist to identify at-risk early. This study investigated whether disconnection features derived from infarcts and white matter hyperintensities (WMH) could serve as markers for short- long-term decline in first-ever ischemic patients. First-ever (NIHSS ≤ 7) were prospectively followed at 72-h, 6 months, 36 months with tests brain MRI. Infarct WMH volumes semi-automatically...
Abstract To evaluate the relative contribution of different Magnetic Resonance Imaging (MRI) sequences for extraction radiomics features in a cohort patients with lacrimal gland tumors. This prospective study was approved by Institutional Review Board and signed informed consent obtained from all participants. From December 2015 to April 2017, 37 lesions underwent MRI before surgery, including axial T1-WI, Diffusion-WI, coronal DIXON-T2-WI post-contrast DIXON-T1-WI. Two readers manually...
Background and Purpose— We aimed to study the intrarater interrater agreement of clinicians attributing DWI-ASPECTS (Diffusion-Weighted Imaging–Alberta Stroke Program Early Computed Tomography Scores) DWI-FLAIR Imaging–Fluid Attenuated Inversion Recovery) mismatch in patients with acute ischemic stroke referred for mechanical thrombectomy. Methods— Eighteen raters independently scored anonymized magnetic resonance imaging scans 30 participants from a multicentre thrombectomy trial, 2...
There are no current recommendations on which machine learning (ML) algorithms should be used in radiomics. The objective was to compare performances of ML radiomics when applied different clinical questions determine whether some strategies could give the best and most stable regardless datasets. This study compares nine feature selection combined with fourteen binary classification ten These datasets included features diagnosis for classifications including COVID-19 pneumonia or sarcopenia...
Objectives Distinguishing benign from malignant orbital lesions remains challenging both clinically and with imaging, leading to risky biopsies. The objective was differentiate using radiomics on 3 T magnetic resonance imaging (MRI) examinations. Materials Methods This institutional review board–approved prospective single-center study enrolled consecutive patients presenting an lesion undergoing a MRI prior surgery December 2015 July 2019. Radiomics features were extracted 6 sequences...
Anterior ischemic optic neuropathy (AION) is the most common cause of acute in older patients. Distinguishing between arteritic AION (A-AION) and nonarteritic (NA-AION) paramount for improved patient management.The aim this study was to evaluate 3-dimensional high-resolution vessel wall (HR-VW) magnetic resonance imaging (MRI) at 3 T discriminate A-AION from NA-AION.This prospective single-center approved by a national research ethics board included 27 patients (17 10 NA-AION) with 36 AIONs...
Background Orbital tumors present a diagnostic challenge due to their varied locations and histopathological differences. Although recent advancements in imaging have improved diagnosis, classification remains challenge. The integration of artificial intelligence radiology ophthalmology has demonstrated promising outcomes. Purpose This study aimed evaluate the performance machine learning models accurately distinguishing malignant orbital from benign ones using multiparametric 3 T magnetic...
MR imaging is the key examination in follow-up of patients with MS, by identification new high-signal T2 brain lesions. However, identifying lesions when scrolling through 2 images can be difficult and time-consuming. Our aim was to compare an automated coregistration-fusion reading approach standard multiple sclerosis during imaging.This prospective monocenter study included 94 (mean age, 38.9 years) treated for MS dimethyl fumarate from January 2014 August 2016. One senior neuroradiologist...
Multinodular and vacuolating neuronal tumor of the cerebrum is a rare supratentorial brain described for first time in 2013. Here, we report 11 cases infratentorial lesions showing similar striking imaging features consisting cluster low T1-weighted high T2-FLAIR signal intensity nodules, which referred to as multinodular posterior fossa unknown significance. No relationship was found between location lesion clinical symptoms. A hypointense central dot sign present images 9/11 (82%)...