A5027336767- Muscle activation and electromyography studies
- Advanced Neural Network Applications
- Brain Tumor Detection and Classification
- Hemodynamic Monitoring and Therapy
- Motor Control and Adaptation
- Medical Imaging and Analysis
- AI in cancer detection
- Stroke Rehabilitation and Recovery
- Tactile and Sensory Interactions
- Heart Rate Variability and Autonomic Control
- Parkinson's Disease Mechanisms and Treatments
- Medical Image Segmentation Techniques
- Neurological disorders and treatments
- Image Retrieval and Classification Techniques
- Cerebral Palsy and Movement Disorders
- Voice and Speech Disorders
- Machine Learning and ELM
- Non-Invasive Vital Sign Monitoring
- Ergonomics and Musculoskeletal Disorders
- Hand Gesture Recognition Systems
University of Belgrade
2018-2025
University of Belgrade - School of Electrical Engineering
2019-2024
The classification of brain tumors is performed by biopsy, which not usually conducted before definitive surgery. improvement technology and machine learning can help radiologists in tumor diagnostics without invasive measures. A machine-learning algorithm that has achieved substantial results image segmentation the convolutional neural network (CNN). We present a new CNN architecture for three types. developed simpler than already-existing pre-trained networks, it was tested on T1-weighted...
Background: This study analyzed different classifier models for differentiating pancreatic adenocarcinoma from surrounding healthy tissue based on radiomic analysis of magnetic resonance (MR) images. Methods: We observed T2W-FS and ADC images obtained by 1.5T-MR 87 patients with histologically proven training validation purposes then tested the most accurate predictive that were another group 58 patients. The tumor segmented three consecutive slices, largest area interest (ROI) marked using...
The use of machine learning algorithms and modern technologies for automatic segmentation brain tissue increases in everyday clinical diagnostics. One the most commonly used image processing is convolutional neural networks. We present a new autoencoder tumor based on semantic segmentation. developed architecture small, it tested largest online database. dataset consists 3064 T1-weighted contrast-enhanced magnetic resonance images. proposed architecture’s performance using combination two...
We propose a novel system for measuring finger force profiles dexterity assessment. Developed application controls turning on/off of LEDs (each LED corresponds to one finger) signalize the user which should press corresponding strain gage sensor. Linearity, repeatability and sensitivity position pressure sensors were tested prove reliability. An example potential usage with assessment quantitative parameters is presented.
We propose a novel system for measuring finger force profiles dexterity assessment. The consists of software application and encased hardware. sensing part the ten high sensitivity strain gage sensors, one each finger. developed controls turning ON/OFF LEDs (each LED corresponds to finger) signalize user which should press corresponding sensor. Linearity, repeatability, position pressure sensors were tested prove reliability. potential usage with assessment quantitative parameters was...
Medical imaging is substantial in diagnosing and treatment of various types diseases, however, a single modality may not be sufficient. Therefore, gathering more than one to provide additional information needed. Thus, an open-source application for the registration two medical modalities using homography transformation was implemented. Using homography, mapping points magnetic resonance image corresponding computer tomography performed. The enables radiologist finely adjust by moving...
Piano training generally improves motor control of hands and fingers in individuals. The assessment fine implies manual dexterity quantification as well finger synergy analysis. In this paper we present a modification the previously developed strain-gauge based system for hand assessment. upgraded was used to compare strategy fingers' between pianists non-pianists. During experiment, examinee had track predefined trapezoidal force pattern real-time. Two tests were performed: first one...