A5003383147- Radiomics and Machine Learning in Medical Imaging
- Brain Tumor Detection and Classification
- Advanced Neural Network Applications
- Medical Imaging Techniques and Applications
- Advanced Radiotherapy Techniques
- Glioma Diagnosis and Treatment
- Advanced X-ray and CT Imaging
- Radiation Therapy and Dosimetry
- Brain Metastases and Treatment
- Medical Image Segmentation Techniques
- AI in cancer detection
- Radiation Detection and Scintillator Technologies
- Medical Imaging and Analysis
- Management of metastatic bone disease
- Radiation Effects and Dosimetry
- Cerebrospinal fluid and hydrocephalus
- Advanced MRI Techniques and Applications
- Digital Imaging for Blood Diseases
- Vascular Malformations Diagnosis and Treatment
- Fetal and Pediatric Neurological Disorders
- Microtubule and mitosis dynamics
- Intracranial Aneurysms: Treatment and Complications
- COVID-19 diagnosis using AI
- Image Retrieval and Classification Techniques
- Radiopharmaceutical Chemistry and Applications
National Research Nuclear University MEPhI
2016-2024
Burdenko Neurosurgery Institute
2015-2021
Centrum Gamma Knife
2019-2021
Research Center of Neurology
2021
The prevailing approach for three-dimensional (3D) medical image segmentation is to use convolutional networks. Recently, deep learning methods have achieved human-level performance in several important applied problems, such as volumetry lung-cancer diagnosis or delineation radiation therapy planning. However, state-of-the-art architectures, U-Net and DeepMedic, are computationally heavy require workstations accelerated with graphics processing units fast inference. scarce research has been...
Recent Metal Artifacts Reduction (MAR) methods for Computed Tomography are often based on image-to-image convolutional neural networks adjustment of corrupted sinograms or images themselves. In this paper, we exploring the capabilities a multidomain method, which consists both sinogram correction (projection domain step) and restored image (image-domain step). We formulate first step problem directly as inpainting, allows us to use specific field, such partial convolutions. Moreover, propose...
We systematically evaluate a Deep Learning model in 3D medical image segmentation task. With our model, we address the flaws of manual segmentation: high inter-rater contouring variability and time consumption process. The main extension over existing evaluations is careful detailed analysis that could be further generalized on other tasks. Firstly, analyze changes detection agreement. show reduces number disagreements by [Formula: see text] text]. Secondly, improves agreement from to...
The study purpose was to evaluate the impact of gamma knife radiosurgery (GKRS) alone on overall survival and rate intracranial recurrences in brain metastasis patients.Treatment outcomes 502 patients (211 males 291 females with 2782 metastases (BMs)) were retrospectively reviewed. Most (n=142; 28.2%) diagnosed breast cancer. Multiple BMs detected 259 (51.6%). median total tumor volume ВM number 5.9 cm3 (0.09-44.5 cm3) 4 (1-36), respectively. mean marginal radiation dose 21 Gy (15-24 Gy)....
Cerebral arteriovenous malformations (AVMs) are the congenital anomalies of development cerebral vessels during embryonic period. The conventional therapy for AVMs currently includes endovascular management, microneurosurgical resection, and stereotactic irradiation.A total 315 patients with brain were subjected to radiotherapy in 2005-2011. 238 (76%) had previous subarachnoid hemorrhage (SAH) within different time (6 months 5 years) before therapy; 214 (68%) headaches; 113 (36%) focal...
Classification-based image retrieval systems are built by training convolutional neural networks (CNNs) on a relevant classification problem and using the distance in resulting feature space as similarity metric. However, practical applications, it is often desirable to have representations which take into account several aspects of data (e.g., brain tumor type its localization). In our work, we extend classification-based approach with multitask learning: train CNN MRI scans heterogeneous...
Hypothalamic hamartoma (HH) is a dysplastic lesion fused with hypothalamus and followed by epilepsy, precocious puberty behavioral disorders. Up to 50% of patients become free seizures after surgery, but various complications occur in 1/4 cases. Radiofrequency thermocoagulation, laser interstitial thermal therapy stereotactic radiosurgery (SRS) are alternative treatment options.To define the indications for SRS HH clarify irradiation parameters.Twenty-two epilepsy underwent at Moscow...
Download This Paper Open PDF in Browser Add to My Library Share: Permalink Using these links will ensure access this page indefinitely Copy URL DOI
Magnetic resonance imaging (MRI) data is heterogeneous due to differences in device manufacturers, scanning protocols, and inter-subject variability. A conventional way mitigate MR image heterogeneity apply preprocessing transformations such as anatomy alignment, voxel resampling, signal intensity equalization, denoising, localization of regions interest. Although a pipeline standardizes appearance, its influence on the quality segmentation other downstream tasks deep neural networks has...
Magnetic resonance imaging (MRI) data is heterogeneous due to differences in device manufacturers, scanning protocols, and inter-subject variability. A conventional way mitigate MR image heterogeneity apply preprocessing transformations such as anatomy alignment, voxel resampling, signal intensity equalization, denoising, localization of regions interest. Although a pipeline standardizes appearance, its influence on the quality segmentation other downstream tasks deep neural networks has...
Purpose/Objective: For stereotactic radiosurgery (SRS) of multiple brain metastases, several different systems have been used in the last decades; they include GammaKnife, Cyberknife, Tomotherapy and linear accelerators (linacs).Recently, a number works on comparison techniques published.However, none these Elekta linacs software.The purpose this project was to evaluate new single-isocenter technique for SRS metastases using VersaHD linac compare it with Leksell GammaKnife (LGK) based...
In this work machine learning approach was used to predict the patients overall survival after Gamma Knife radiosurgery. We constructed regression and multiclass classification models time interval from onset of oncological disease unfavourable outcome patient's class. The were built on data 916 with 26 different primary features. train set included known clinical outcomes (445 patients). median deviation in determining cancer date death 1.4 months. According model most significant feature...
Glioma is one of the most common primary tumors among adults. Glioblastoma multiforme (GBM) aggressive form glioma with very poor prognosis. The median patient survival about 15 months. Treatment requires a complex approach combining surgical resection, chemotherapy and radiation therapy. Definition tumor border important step therapy treatment planning. rapid development diagnostic methods made it possible to address this challenging task. However, optimal volume still matter debate due...