Browsing by Author "Vujaklija, Ivan"
Now showing 1 - 20 of 90
Results Per Page
Sort Options
Item Adaptive HD-sEMG decomposition: towards robust real-time decoding of neural drive(Institute of Physics Publishing, 2024-04-01) Yeung, Dennis; Negro, Francesco; Vujaklija, Ivan; Department of Electrical Engineering and Automation; Bionic and Rehabilitation Engineering; University of BresciaObjective. Neural interfacing via decomposition of high-density surface electromyography (HD-sEMG) should be robust to signal non-stationarities incurred by changes in joint pose and contraction intensity. Approach. We present an adaptive real-time motor unit decoding algorithm and test it on HD-sEMG collected from the extensor carpi radialis brevis during isometric contractions over a range of wrist angles and contraction intensities. The performance of the algorithm was verified using high-confidence benchmark decompositions derived from concurrently recorded intramuscular electromyography. Main results. In trials where contraction conditions between the initialization and testing data differed, the adaptive decoding algorithm maintained significantly higher decoding accuracies when compared to static decoding methods. Significance. Using “gold standard” verification techniques, we demonstrate the limitations of filter re-use decoding methods and show the necessity of parameter adaptation to achieve robust neural decoding.Item Analysis and evaluation of systematic test management frameworks in agile embedded software development(2024-03-11) Liljeström, Oskar; Korhonen, Janne; Sähkötekniikan korkeakoulu; Vujaklija, IvanAgile software development practices, such as continuous integration and deployment, are becoming more common within the embedded software development area. These practices generally require extensive testing of the software throughout the entire development process. The particularities of testing embedded system software further complicate the process, as the tests are often hardware-specific and tests can take relatively long to execute. To make sufficient testing feasible, developers make use of various testing and automation tools to reduce the workload. However, because each tool serves a specific purpose, information related to the testing progress of a software project is easily split up over multiple plat-forms. This can make it hard to get a complete overview of the project’s test and requirements coverage, and gathering the data and reporting on it takes manual work and bears the risk of human error in the results. In order to ease the integration of different test management tools in-to the software testing workflow, this thesis develops a framework for automating the transfer of test results and requirements information between different testing tools. To evaluate the functionality of the pro-posed framework, the thesis implements a functioning system based on the framework guidelines and puts it to use in a live testing environment. The implemented system offers significant improvements to the speed and reliability of test status reporting and requirements validation, compared to previous solutions. A weak point in the developed framework is the configuration work required to set up the different tools before information can be shared between them. Future research could investigate whether it’s possible to also automate the configuration of one tool based on the configuration of another.Item Analysis and modernization of microwave radar receiver design(2022-08-22) Koski, Visa; Vujaklija, Ivan; Sähkötekniikan korkeakoulu; Vujaklija, IvanRadar (Radio Detection and Ranging) is a system that uses radio waves to detect and track targets. The first significant development of radar occurred before and during World War II when it provided a substantial advantage in the detection of air and sea targets. Typical marine surveillance radar systems use frequency mixing to convert the received signal to IF (intermediate frequency). Developments in semiconductor technology, that have taken place in recent decades, make it possible to digitize the intermediate frequency of the radar receiver more inexpensively and simply than before. This thesis aims to design a new receiver as a part of a larger modernization program for the existing marine surveillance radar system. The modernization program is based on the integration of a new digitizer unit with new requirements for the receiver. The goal is to rely on the existing front end, but conversely the characterization and measurements are needed in order to ensure it can operate as part of the new receiver. The benefit of this work is increased radar system performance and enabled further modernization of other components in the radar system. A system for tuning of IF of the front end is designed, and it consists of hardware and software for tune control. A simulation program is written in Python to test the new receiver without the need for all the electronics to be present. The simulation program is used to run tests on different setups and measure the performance of the tuning system. Furthermore, the real data has been collected on the same front end as the existing system. The front end performed as expected with lower than specified IF frequencies and a lower image rejection ratio. The tuning control logic was successfully implemented in Python and was tested as part of the system with the help of the simulation program. The tuning control worked as desired and could tune the intermediate frequency to the desired level in situations the system is expected to face. The simulations were performed using signal generator and magnetron as RF (radio frequency) sources.Item Anchoring systems for lower limb prosthetics(2022-09-16) Jensen, Nadja; Vujaklija, Ivan; Sähkötekniikan korkeakoulu; Forsman, PekkaItem Applications of Electrotactile Feedback for Prosthetics and Rehabilitation(2020-11-30) Siljamo, Oula; Vujaklija, Ivan; Sähkötekniikan korkeakoulu; Turunen, MarkusItem Artificial Intelligence Approaches for Prediction of Ground Reaction Forces During Walking(2020-09-17) Mård, Rudolf; Vujaklija, Ivan; Sähkötekniikan korkeakoulu; Turunen, MarkusItem Assistive multi-sensor framework for prevention and monitoring of pressure ulcers(2022-12-12) Farooqi, Shah Fahad; Gowrishankar, Ramyah; Sähkötekniikan korkeakoulu; Vujaklija, IvanPressure ulcers (PU) are injuries to the skin and soft tissues due to prolonged pressure and have a significant impact on the quality of a patient’s life. The common risk factors of PU are pressure, shearing forces, friction, moisture, temperature, and immobility. The use of low-tech and high-tech devices is effective in distributing the high pressure exerted on the patient’s body. However, patients can still develop PU because these devices do not provide any real-time information about the risk factors of PU development. With the help of sensors, smart beds are effective to monitor and prevent the development of PU but the prevalence of PU around the globe is still high due to the aging population. In this thesis, a multi-sensor framework is developed and tested to aid in the monitoring and prevention of PU. The proposed system is able to measure, process, and store real-time information about the numerous risk factors of PU. In addition, the framework does not need to be attached to the patient’s body in order to provide extra comfort. Different experiments are conducted to test the reliability and effectiveness of the proposed system. Experimental results indicate that the system is capable to provide the rate of repositioning, temperature, and pressure distribution.Item Audio feedback modulation for closed-loop myoelectric prosthesis control(2020-05-07) Granö, Henrik; Vujaklija, Ivan; Sähkötekniikan korkeakoulu; Turunen, MarkusItem Benchmark methodologies for the optimized physical synthesis of RISC-V microprocessors(2023-08-21) Palayath Godage Don, Shashika; Ganesha, Ajay; Sähkötekniikan korkeakoulu; Vujaklija, IvanAs technology continues to advance and chip sizes shrink, the complexity and design time required for integrated circuits have significantly increased. To address these challenges, Electronic Design Automation (EDA) tools have been introduced to streamline the design flow. These tools offer various methodologies and options to optimize power, performance, and chip area. However, selecting the most suitable methods from these options can be challenging, as they may lead to trade-offs among power, performance, and area. While architectural and Register Transfer Level (RTL) optimizations have been extensively studied in existing literature, the impact of optimization methods available in EDA tools on performance has not been thoroughly researched. This thesis aims to optimize a semiconductor processor through EDA tools within the physical synthesis domain to achieve increased performance while maintaining a balance between power efficiency and area utilization. By leveraging floorplanning tools and carefully selecting technology libraries and optimization options, the CV32E40P open-source processor is subjected to various floorplans to analyze their impact on chip performance. The employed techniques, including multibit components prefer option, multiplexer tree prefer option, identification and exclusion of problematic cells, and placement blockages, lead to significant improvements in cell density, congestion mitigation, and timing. The optimized synthesis results demonstrate a 71\% enhancement in chip design performance without a substantial increase in area, showcasing the effectiveness of these techniques in improving large-scale integrated circuits' performance, efficiency, and manufacturability. By exploring and implementing the available options in EDA tools, this study demonstrates how the processor's performance can be significantly improved while maintaining a balanced and efficient chip design. The findings contribute valuable insights to the field of electronic design automation, offering guidance to designers in selecting suitable methodologies for optimizing processors and other integrated circuits.Item Benefits of RISC-V as an educational tool(2023-06-03) Makkonen, Iikka; Vujaklija, Ivan; Sähkötekniikan korkeakoulu; Forsman, PekkaItem Biomechanical Analysis of Body Movements of Myoelectric Prosthesis Users During Standardized Clinical Tests(IEEE, 2023-03-01) Vujaklija, Ivan; Jung, Moon Ki; Hasenoehrl, Timothy; Roche, Aidan D.; Sturma, Agnes; Muceli, Silvia; Crevenna, Richard; Aszmann, Oskar C.; Farina, Dario; Department of Electrical Engineering and Automation; Bionic and Rehabilitation Engineering; Imperial College London; Medical University of Vienna; St John's Hospital; Chalmers University of TechnologyObjective: The objective clinical evaluation of user's capabilities to handle their prosthesis is done using various tests which primarily focus on the task completion speed and do not explicitly account for the potential presence of compensatory motions. Given that the excessive body compensation is a common indicator of inadequate prosthesis control, tests which include subjective observations on the quality of performed motions have been introduced. However, these metrics are then influenced by the examiner's opinions, skills, and training making them harder to standardize across patient pools and compare across different prosthetic technologies. Here we aim to objectively quantify the severity of body compensations present in myoelectric prosthetic hand users and evaluate the extent to which traditional objective clinical scores are still able to capture them. Methods: We have instructed 9 below-elbow prosthesis users and 9 able-bodied participants to complete three established objective clinical tests: Box-and-Blocks-Test, Clothespin-Relocation-Test, and Southampton-Hand-Assessment-Procedure. During all tests, upper-body kinematics has been recorded. Results: While the analysis showed that there are some correlations between the achieved clinical scores and the individual body segment travel distances and average speeds, there were only weak correlations between the clinical scores and the observed ranges of motion. At the same time, the compensations were observed in all prosthesis users and, for the most part, they were substantial across the tests. Conclusion: The sole reliance on the currently available objective clinical assessment methods seems inadequate as the compensatory movements are prominent in prosthesis users and yet not sufficiently accounted for.Item CAN Bus and CANopen Standard for Building and Home Automation, Analysis for Feasibility(2024-04-26) Kasela, Siim; Vujaklija, Ivan; Sähkötekniikan korkeakoulu; Ylirisku, SaluBuilding and home automation is an essential part of energy efficiency efforts around the world. Wired and wireless communication protocols help implement the functionalities necessary for these automation systems. Because of the significant popularity increase in IoT solutions, wireless field-level protocols have been the main focus of development. Consequently, the adoption and development of wired fieldbus solutions have slowed down. Therefore, this thesis determines if CAN bus combined with CANopen standard is feasible for wired building and home automation to offer a matured low-cost alternative for existing fieldbuses in the industry. The proposed wired solution could decrease the over-reliance on wireless technologies and complement those networks where possible. For this purpose, a proof-of-concept home automation system was developed using both custom-designed and commercial hardware with open-source software. The developed fieldbus network was additionally tested with a series of transmission tests to verify its performance. As a result, a CANopen fieldbus network was integrated into a well-known open-source IoT solution. Additionally, the underlying network performance was confirmed with some drawbacks noticed regarding the chosen microcontroller. This work presents the potential of CAN bus and CANopen standard in the context of building and home automation and justifies further research into the topic.Item Can Multi-DoF Training Improve Robustness of Muscle Synergy Inspired Myocontrollers?(IEEE, 2019-06-01) Yeung, Dennis; Farina, Dario; Vujaklija, Ivan; Department of Electrical Engineering and Automation; Bionic and Rehabilitation Engineering; Imperial College LondonNon-negative Matrix Factorization (NMF) has been effective in extracting commands from surface electromyography (EMG) for the control of upper-limb prostheses. This approach enables Simultaneous and Proportional Control (SPC) over multiple degrees-of-freedom (DoFs) in a minimally supervised way. Here, like with other myoelectric approaches, robustness remains essential for clinical adoption, with device donning/doffing being a known cause for performance degradation. Previous research has demonstrated that NMF-based myocontrollers, trained on just single-DoF activations, permit a certain degree of user adaptation to a range of disturbances. In this study, we compare this traditional NMF controller with its sparsity constrained variation that allows initialization using both single and combined-DoF activations (NMF-C). The evaluation was done on 12 able bodied participants through a set of online target-reaching tests. Subjects were fitted with an 8-channel bipolar EMG setup, which was shifted by 1cm in both transversal directions throughout the experiments without system retraining. In the baseline condition NMF performed somewhat better than NMFC, but it did suffer more following the electrode repositioning, making the two perform on par. With no significant difference present across the conditions, results suggest that there is no immediate advantage from the naïve inclusion of more comprehensive training sets to the classic synergy-inspired implementation of SPC.Item Clinical potential and challenges of invasive brain-computer interfaces(2020-12-18) Luoto, Antti; Vujaklija, Ivan; Sähkötekniikan korkeakoulu; Turunen, MarkusItem Co-adaptive control of bionic limbs via unsupervised adaptation of muscle synergies(IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC, 2022-08-01) Yeung, Dennis; Mendez Guerra, Irene; Barner-Rasmussen, Ian; Siponen, Emilia; Farina, Dario; Vujaklija, Ivan; Department of Electrical Engineering and Automation; Bionic and Rehabilitation Engineering; Imperial College London; Helsinki University Central HospitalObjective: In this work, we present a myoelectric interface that extracts natural motor synergies from multi-muscle signals and adapts in real-time with new user inputs. With this unsupervised adaptive myocontrol (UAM) system, optimal synergies for control are continuously co-adapted with changes in user motor control, or as a function of perturbed conditions via online non-negative matrix factorization guided by physiologically informed sparseness constraints in lieu of explicit data labelling. Methods: UAM was tested in a set of virtual target reaching tasks completed by able-bodied and amputee subjects. Tests were conducted under normative and electrode perturbed conditions to gauge control robustness with comparisons to non-adaptive and supervised adaptive myocontrol schemes. Furthermore, UAM was used to interface an amputee with a multi-functional powered hand prosthesis during standardized Clothespin Relocation Tests, also conducted in normative and perturbed conditions. Results: In virtual tests, UAM effectively mitigated performance degradation caused by electrode displacement, affording greater resilience over an existing supervised adaptive system for amputee subjects. Induced electrode shifts also had negligible effect on the real world control performance of UAM with consistent completion times (23.91±1.33 s) achieved across Clothespin Relocation Tests in the normative and electrode perturbed conditions. Conclusion: UAM affords comparable robustness improvements to existing supervised adaptive myocontrol interfaces whilst providing additional practical advantages for clinical deployment. Significance: The proposed system uniquely incorporates neuromuscular control principles with unsupervised online learning methods and presents a working example of a freely co-adaptive bionic interface.Item The Components and Requirements of Bionic Legs for Gait Rehabilitation(2020-12-07) Mattila, Ilona; Vujaklija, Ivan; Sähkötekniikan korkeakoulu; Turunen, MarkusItem Computer vision assisted control of upper limb prostheses(2023-09-09) Harju, Joona; Vujaklija, Ivan; Sähkötekniikan korkeakoulu; Forsman, PekkaLoss of a hand can be devastating to a person’s ability to function in everyday life. A common solution is the use of a prosthetic device. Most commercial prosthetic devices function by measuring electromyographic (EMG) signals from the skin surface of muscles. These signals are then interpreted into controlling one functionality of a prosthesis, therefore resulting in clumsy and time-consuming operation. Advanced myoelectric prostheses use classification methods to estimate a desired hand gesture by applying pattern recognition to a larger amount of EMG signals. However, the number of different grasp types is limited, and the movements must be sequential, therefore still lacking in intuitiveness. Additionally, operating the prosthetic is cognitively effort demanding. Alone, myoelectric control is difficult to use for smooth and intuitive control, however, augmenting additional modules for sensing the environment and the state of the device can aid in improving the control. This thesis discusses the approaches of computer vision assisted control of upper limb prostheses. Computer vision can be used to sense the environment and surroundings of the device and the user, allowing more accurate classification for choosing the correct hand gesture. In advanced multimodal approaches, computer vision can be used to achieve semi-autonomous control, resulting in less time- and effort-consuming operation of prosthetic devices. The thesis discussed three different researches based on their different approaches and found out that computer vision can indeed aid in achieving better control for the users by offering more intuitive and less effort-consuming control. However, the researches did not consider the implementation to real life scenarios of the systems. The design for actual usable and convenient devices utilizing computer vision still requires more research.Item Control of Exo-Skeleton for Error Augmentation Rehabilitation Protocols(2023-08-21) Kazmi, Syed; Vujaklija, Ivan; Sähkötekniikan korkeakoulu; Vujaklija, IvanItem Control strategies of actuated upper limb prostheses(2022-12-12) Lindblad, Ann-Sofi; Vujaklija, Ivan; Sähkötekniikan korkeakoulu; Turunen, MarkusItem Custom design of fine wire EMG recording set-ups(2023-01-16) Metsämäki, Daniel; Vujaklija, Ivan; Sähkötekniikan korkeakoulu; Vujaklija, IvanElectromyographic (EMG) research studies muscle action potentials with two different types of electrodes: surface and invasive. Although surface electrodes are easier to use and pose less risks, invasive applications are increasingly utilised as they offer better stability and higher sensitivity. A special case of invasive electrodes are fine wires, which are electrically conductive, biocompatible, flexible wires, that tend to offer high stability and agility for the long-term implantations. The objective of this master’s thesis was to investigate an optimal implantable fine wire electrode interface which would offer more sensitive and accurate EMG signal measurements. To achieve this, multiple fine wire designs with different properties were designed and prepared, and then tested based on their operability and functionality. The tests were initially performed in vitro with a ballistic gel and an external function generator, which offered a risk-free way to compare different fine wire designs with similar circumstances. Some of the more promising fine wire designs were then tested in vivo in a controlled laboratory environment. The main findings of the study suggest that an increased number of fine wires, as well as longer sensing tips, enhance the electrode set-up’s ability to detect signals. In turn, the diameter, which affects the detection surface, does not have a significant effect. Finally, the material of the wire affects the sensitivity and precision, but the differences may not be significant. In conclusion, to enhance the sensitivity and precision of the fine wire set-up, it is recommended to use at least a bipolar wire configuration with long detection tips. Many of the available fine wire materials seem to be highly applicable for EMG research, but the magnitude of difference in the detection ability remains unclear. Nevertheless, the accuracy of the results in this study was limited by several notable factors. Firstly, the wire materials were not available with the same diameters and insulation layers, which could have distorted the results. In addition, although in vitro measurements offer good reproducibility, it is very challenging to capture the complex nature of human tissue with artificial environment, which certainly affected the reliability of the results of the in vitro measurements. Of course, some of the fine wire set-ups were also tested in vivo to verify that the set-up functions in a general level with human tissue.
