|P1. Application Specific Processors for Signal Processing in Biomedicine
P2. Verification, test generation and fault diagnosis
P3. Optical and bioimpedance methods in cardiovascular diagnostics
P4. Evaluation of mental disorders using EEG analyser
P5. Reliable and disturbance free monitoring of dialysis
P6. Embedded Instrumentation Platform for Board Testing
P7. Semiconductor devices and physical defects
Figure illustrates the cooperation links between the research fields of CEBE and the project flows from basic research towards applications.
For example, the path of the project P4 (EEG Analyzer) shown by red colour starts with Brain Studies (basic research), continues through Signal Processing (algorithm research) up to Dependable Design (implementation of the Analyzer). The circuits of the Analyzer are verified and the implemented board will be tested. The feadback from the verification and test results may be used for redesign of the circuit to improve testability or dependability. In all these covered research fields, novel ideas, methods, algorithms and prototype tools for automatization of the desing work are used.
The projects P1, P3 and P4 are devoted to developing different signal sensoring and interpretation methods (TM teams), developing of novel signal processing algorithms (DE team) and related new specific signal processor architectures (DCE design team). The outputs of these projects are demonstrators or prototype boards.
The project P2 (lead by DCE verification team) is tightly coupled with the design process in projects P1, P3 and P4, contributing with design verification and error diagnosis (debugging). In the project P2 also the testability analysis of the designs is carried out, and BIST structures will be developed which results in corresponding redesign cycles for improving the testability and dependability.
The project P6 (lead by DCE test team) is also tightly coupled with the previous projects having the goal of developing test strategies for board level testing and test generation or implementing the tests developed in P2.
In the project P5, a synergy between the partners of DCE and TM is created by using the diagnostic methods and algorithms common in testing of electronic hardware in diagnosing and estimating the dialysis adequacy to secure high care quality of end stage renal disease patients
In the project P7, different research activities in a very deep physical level of GaAs and SiC based power p-n- and Schottky structures are carried out between DE and several external partners of CEBE. There is going on as well an internal cooperation in CEBE between DE and DCE Test and Diagnosis groups regarding pre-analysis of dynamic characteristics for semiconductor circuits. The knowledge derived from the related simulation and measurements would help to specify more exactly defect models for transistor circuits, especially in the context of delay faults, and to improve the strategies for testing physical defects with complex behavior.
Application Specific Processors for Signal Processing in Biomedicine
Application Specific Processors for Signal Processing in Biomedicine. Possible architectural solutions of the reprogrammable sampling module have been analyzed and implemented. The solutions took into account performance requirements for the pre-processing unit. Different implementation possibilities using reconfigurable technology (FPGA) were explored. Among architectural solutions, enhancements for signal processing were explored and as a result, alias effect reducing sampling strategy with supporting architecture was developed.
Verification, test generation and fault diagnosis.
The goal of the project is to verify the designs of signal processors developed in CEBE and carry out the testability analysis, fault simulation and test generation. A converter was developed for creating SSBDD based diagnostic models of the processor architectures developed in P1. The VHDL designs of 8 signal processors developed in P1 were converted into the diagnostic models. The testability of the processors and the potential quality of testing the processors by traditional BIST methods was analyzed and suggestions for improving the testability were worked out. A useful synergy was achieved here by cooperative design of bio-signal processors which will have a practical use in the medical field and in the same time can be used as a family of benchmark systems for evaluating the properties of new algorithms in the field of digital test [1,2].A new marker was developed for measuring the testability of digital circuits. Differently from the known testability measures based on simulation of circuits, the proposed marker is calculated as a function of the structure of re-converging fan-outs, and in this sense can be regarded as the measure of structural complexity of the circuit as well. The new measure outperforms the known simulation based testability measures in the speed of calculation.
The aim of this project is to develop a smart optical sensor for cardiovascular activity monitoring at different tissue layers and to obtain the pulse wave signal from artery. Photoplethysmography (PPG) is a noninvasive optical technique for monitoring mainly blood volume changes in the examined tissue. However, different important physiological parameters, such as oxygen saturation, heart and breathing rate, dynamics of skin micro-circulation, vasomotion activity etc., can be extracted from the registered PPG signal. The sensor consists of number of light emitting sources with different wavelengths and photodiodes. Photodiodes are located to different distances from the LEDs. The sensor enables to combine different LED and photodiode optopairs. The signal can be obtained from the different layers due to the optical properties of the tissue. Compared to the available sensors, the system enables to select the optimal LED (light emitting diode) and photo detector couple in order to obtain the pulse wave signal from the interested blood vessels with the highest possible signal to noise ratio.
The objective is to use the sensor for various application areas, such as sleep research, heart fibrillation and atherosclerosis detection, and others.
Evaluation of mental disorders using EEG analyser
A method for determining human depression by measuring and analysis of the EEG signals of the brain was developed. The method is targeted for real-time in-field diagnostics of human brain depressive and potentially other mental disorders, which are related to similar imbalances. The analysis is based on an original algorithm for calculation of spectral asymmetry index (SASI) . The ongoing research has revealed a good potential of the method to determine also psychological stress and influence of microwave radiation.
In cooperation between BME and DCE, the SASI algorithm was implemented as a FPGA-based prototype of an EEG Analyzer device . The analyzer is aimed at early detection of brain disorders and can be used for preventing screening of people or monitoring special groups of high-risk or high-stress workers (military personnel, police, rescue workers). Compared to existing EEG analyzers, the developed one has advantages in high detectability of disorders and simplicity of implementation (one EEG channel, quick non-invasive method). The proposed portable EEG analyzer device will be evaluated in the North Estonia Medical Center.
Reliable and disturbance free monitoring of dialysis
In clinical practice, many events like patient blood pressure changes, needle displacement and concentrate depletion can trigger dialysis machine alarm, which will stop the treatment and give wrong information during optical dialysis dose monitoring. The competences of BME in biosensorics and RES in digital design and dependability formed the creative and synergetic basis for developing new competitive bioengineering solutions within optical dialysis monitoring. The research results in test and diagnosis developed by RES allows coping with the complexity of dependency problems in new innovative applications proposed by BME. New methods (AVRG, SMART and SIF) for accurate dialysis dose evaluation and extrapolation by means of Kt/V from online UV-absorbance measurements were proposed. The algorithms have a significantly positive effect on removal the disturbances and data visualization for the doctors showing substantial improvement on both chart readability and measurement precision. The output of the monitor has higher reliability and helps to avoid false prescriptions and high-quality medical care can be offered for the patients. The final aim of the dialysis monitoring technique will be to deliver a platform for technology transfer realised as a practical toolkit for the clinicians helping obtain adequate dialysis targets meeting the individual needs of each patient and leading to "personalized healthcare" within haemodialysis.
Embedded Instrumentation Platform for Board Testing
A new board-level test conception based on FPGA-enabled synthesizable embedded instruments. The purpose of intelligent embedded instrument is to carry out a vast portion of test application related procedures, perform measurement and configuration of system components thus minimizing the usage of external test equipment. FPGA-enabled embedded instruments enable significant reduction of test costs by replacing traditional test and measurement equipment and facilitate dynamic high-speed and at-speed testing, thus significantly improve JTAG test quality.
The cooperation of CEBE with Testonica Lab and Göpel Electronic GmbH (Germany) in the field of Decision Diagram Theory has resulted in two products „Microprocessor models for testing" and „Embedded instrumentation IPs for testing of electronic boards and chips." The licences are being sold in the markets all over the world.
The BERT (Bit-Error-Rate-Test) algorithms were developed and implemented as a test equipment for certification and testing of communication lines at the order of CERN for using in the Large Hadron Collider.
Semiconductor devices and physical defects
Nowadays in the commercial production of SiC Schottky and GaAs p-n diodes with reverse voltage up to 600V and forward current up to 20A are available. They are mostly used in pulsed voltage converter instead of fast recovery silicon diodes with p-n junction.
The goal of investigations in the field of energy management applications is to develop new type smart converters (including voltage multiplier in input, intelligent transformer and bidirectional energy flow conversion (AC/DC) modules) using new effective semiconductor switching devices. The scientific aim to solve described goal is to introduce new solutions for the joining technologies of solids (e.g. metal-semiconductor or semiconductor-semiconductor contacts) and upgrade the theoretical models describing the behavior of the semi-wide and wide bandgap materials based Schottky and combined interfaces and structures.