Biomedical Engineering

Subtask 1 Biomedical technologies

Main activity 1 Clinical applications of diagnostics and therapy in surgery (Ronja, Bibi)

The main goal of the activity is automatic segmentation of liver tissue from CT images of the patient and creation of a suitable 3D model for its use in consequent processes. This issue will be solved in collaboration with physicians from FNO, VP7.

A software tool will be developed for automatic segmentation of liver tissue from image data, including the possibility of subsequent validation of the results by a physician. It will be possible to segment important parts of liver tissue such as liver parenchyma, tumours, and blood vessels. Technically, however, it will be possible to segment any part that is represented by a sufficient amount of validated data. The primary focus will be on segmentation of vascular structures, whose quality segmentation poses the greatest technical challenge. The tool will include the aforementioned ability to validate segmentation results by a physician and collect validated data. The collection of this data will allow the improvement of the whole automatic segmentation system. The software solutions will be implemented as a functional system that will connect FNO and VŠB and thus provide fast and individualized creation of liver tissue models from image data. The solution will also include efficient use of available HPC technologies available at the IT4I site.

The created 3D models will be subsequently used in a virtual reality environment and at the same time in this environment the possibility of data analysis in the form of preoperative preparations over individual patients will be created. The specific tools that will be possible and desirable to use in the VR environment will be consulted with doctors from FNO.

The resulting 3D models created through automatic segmentation will most likely be too complex for use in augmented reality devices, so part of the preparation of the final 3D model will involve optimizing the model to reduce its complexity while preserving important details.

RONJA flex - Radiofrequency surgical instrument for superficial and subsurface hemispheric application

The activity represents the use of technical devices to stop bleeding after traumatic events or during surgical procedures and is a necessity nowadays. With the advancement of electrical engineering, these technical tools can be improved and refined to make the work of doctors as easy as possible, while at the same time ensuring the improvement of human life.

Electrocoagulation devices use high-frequency energy - RF to achieve a coagulation effect in tissue.  We may also encounter the naming of these devices as radiofrequency - RF because of the use of the same frequency band as for radio signal transmissions. Radiofrequency radiation is defined by the Institute of Electrical and Electronics Engineers - IEEE as ranging from 3 kHz to 300 GHz and is absorbed by biological systems that contain water containing free ions.

In the presented part of the project the influence of the type of radiofrequency instruments will be investigated when used on selected tissues and surgical procedures. The knowledge gained from surgical procedures and postoperative conditions will be applied in the development and design of original radiofrequency surgical instruments.

Taking into account the knowledge and experience gained from the use of existing radiofrequency instruments in liver surgery, new radiofrequency surgical instruments called RONJA (Radiofrequency Operative Liver Ablation Instrument) have been developed. The instruments are designed both for superficial and subsurface application with effective electrode placement - RONLINE and for superficial and subsurface hemispheric application - RONJA ICS. Research and testing of the new Ronja ICS flex tool with a variable hemispherical tool size for removal of subsurface neoplasms of liver tissue is planned.

Bibi – Biopsy Bloodless incision

The aim of the project is to verify the functionality of a newly designed tool for liver tissue biopsy with the possibility of radiofrequency treatment of the puncture channel after injection. Another aim of this project is to compare the newly designed biopsy tool with the standard method of liver biopsy. The method to verify the feasibility of liver biopsy with radiofrequency treatment of the puncture canal is an experimental in vivo study in laboratory pigs. The experimental study will be conducted on a set of 8 laboratory pigs. The pig as an experimental animal was chosen due to its easy availability, economic considerations, and relatively undemanding husbandry conditions. All phases of this study will be conducted in an approved animal facility. A set of 8 pigs will be randomly divided into two groups A (n=4) and B (n=4). Group A will undergo liver biopsy with a newly designed biopsy instrument under laparoscopic control with 10 liver samples taken each time and subsequent coagulation of the puncture channels using radiofrequency energy. Laboratory pigs in group B will undergo 10 standard liver biopsies under laparoscopic visualization.

The project is focused on the design, implementation, and in vitro and in vivo testing of a high-frequency bioptic needle with high-frequency current injection treatment to stop possible bleeding.

At present, both in operating theatres and in clinical outpatient clinics, biopsy needles are used, guided through the abdominal wall into the liver with a non-negligible risk of subsequent bleeding into the abdominal cavity. Especially in outpatient procedures where the needle puncture is guided by ultrasound, the possibility of post-sampling bleeding rates is nil.

The project will focus on the use of electrosurgical, especially radiofrequency methods for coagulation of the biopsy needle injection site.  For this purpose, a prototype of a new biopsy needle will be designed with emphasis on ergonomics and ease of use. On the other hand, a needle introducer with a stepping mechanism will be designed to enable precise depth of sample removal and subsequent radiofrequency treatment of the tissue after sample removal. Furthermore, the sampling needle and the principle of RF energy delivery to the sampling site will be developed.   The development of the prototype instrument will include the design of the shape and delivery of the leads to the sampling site, as well as the external source of RF current to stimulate the tissue.

In the course of the solution, the prototype production of the tool and electronics by a reputable prototype manufacturer is envisaged, so that this product can be modified, in case of successful tests, for possible subsequent preclinical tests on animals.

Outcomes: the project will result in a two- to three-part biopsy needle concept using high-frequency tissue excitation leading to coagulation. In the final phase of the project, the high-frequency biopsy needle will be tested for functionality and usability on porcine liver and the results will be documented in detail. Subsequently, in case of positive results, intellectual property protection in the form of a patent will be proceeded with, and the results will be published in a peer-reviewed journal or at a conference.

The activity will be implemented in stages:

• Analysis of existing methods in surgery - liver biopsy and resection
• Prototype development for preclinical animal testing
  • Selection of suitable materials for the production of the bioptic instrument
  • Selection of suitable materials for the production of the bioptic instrument
  
  
• Development of an improved prototype based on further measurements
  • In vivo studies of the functionality of the biopsy tool on tissue samples
  • Fine-tuning the prototype design based on the results of the in vivo study
  • Production of a prototype of an innovative bioptic tool
  • Performing biopsies in the vivarium in collaboration with VP7
  
  
• Validation of results, treatment of intellectual property, publishing
• Preparation of a clinical trial proposal

Main activity 2 Clinical applications of diagnostics and therapy in gastroenterology

This activity is aimed at basic research and deals with the design of a methodology for measuring and evaluating the electrical activity of the stomach to support the diagnosis of the physiological state of the stomach. The output of the activity is a measurement methodology consisting of a defined minimized electrode distribution, sensing rules and preprocessing methods, a methodology for the evaluation of the measured data describing the procedures for the extraction of classifiers and an implemented fuzzy expert system designed for their evaluation. To provide test data for the analysis, design and validation of the principles and procedures in the design of both methodologies, several experimental measurements of gastric electrical activity were performed on volunteers. The proposed fuzzy expert system was developed based on the statistical evaluation of the measured data and clinical experts were consulted. The reliability of the system was verified by simulations with artificial data and evaluation of real signals, and the system passed in both cases and can be used for further validation in clinical practice.

The research will include in vitro abdominal electro gastrographic signal measurement to verify the correlation between food intake and satiety and long-term EGG recording. The aim is to address the issue in collaboration with the OU and FNO Faculty of Medicine into a technical solution for food restriction, measurement and stimulation of gastric symptoms and research into alternative approaches to obesity.

The activity will be implemented in stages:

• Analysis of existing methods in cardiology and surgery
  • A survey of current methods, techniques and innovations recently used
  • Initial measurements and verification on models
  • Modelling the electrical activity of the stomach on the surface of the human body
    
    
• Analysis of electrode distribution, number, measurement parameters
• Development of a prototype electrogastrography measurement device
• Development of a method for gestroenterological monitoring of food intake
• Validation of results, treatment of intellectual property, publishing

Main activity 3 Sensors for diagnostics and therapy

The activity includes two lines of research, mainly whole-body analysis (impedance and tomographic) and the development of testers and simulators, in this case in the field of cardiotocography.

The activity will be implemented in stages:

• Analysis of existing medical instrumentation simulators and testers
• Analysis of availability of tester/simulator parts
• Simulator and tester development
• Validation of results, treatment of intellectual property, publishing

Subtask 2 Telemedicine technologies and systems

The aim of the sub-programme is to develop new telemedicine solutions for remote health monitoring of selected target groups in various areas of telemedicine application.  The activities focus on the development and testing of new telemedicine technologies that enable the sensing of biological signals and other manifestations of the human body, such as physical activity or behavioural manifestations and their changes, using measurements of biological and non-biological signals provided by the environment of the monitored person. In the area of new technology development, the focus is on non-intrusive wearable technologies and methods for non-contact measurement of biological signals. It focuses on the application of telemedicine solutions in various fields of medicine where long-term monitoring of biological signals is appropriate, especially in chronically ill patients and the elderly. The activity also includes development in the field of application of telemedicine to the therapeutic process, in the form of development of telemedicine solutions in the field of rehabilitation, orthotics, prosthetics and other therapeutic areas. The activity also includes the development of algorithms that enable automated evaluation of the detected manifestations of the monitored persons both in the short-term and in the long-term monitoring using artificial intelligence approaches.

Main activity 1 Therapeutic telemedicine

The development of new biotelemetric sensory systems focused on therapy, the implementation of technical solutions in standard therapeutic aids, provides therapists with a better overview of the rehabilitation process. They can thus adapt it more frequently to the individual without the need for frequent personal visits and patient checks. These solutions provide the patient with information about the course of treatment, the correctness and effectiveness of the therapeutic process, thus increasing confidence in the therapist and increasing motivation to undergo the rehabilitation process.

Within the activity we will focus on the development of sensory systems suitable for implementation in various types of rehabilitation aids, which will allow to monitor the course of the rehabilitation process in the home environment. It also includes research and development of supporting software programs that will enable automatic evaluation of the rehabilitation process and the associated automatic adjustment of the rehabilitation plan. In view of the need to support patient motivation to take an active approach to the rehabilitation itself, methods to support patient motivation will be developed and applied in cooperation with experts in the field of psychology.

The second area of development in this activity is the development of new sensory systems for orthotics and possibly prosthetics. For successful therapy with orthotic devices, it is necessary that the device is actively and long-term used. Despite the best efforts of orthotic manufacturers, the use of these devices can be uncomfortable for the patient and may result in the device not being actively used. It is advisable to monitor the types of activities performed - walking, positioning of the limb and the intensity of the activities performed - for example, the number of steps. It is desirable that the lifetime of the sensor is at least in the order of months without the need for recharging. At the same time, it is advisable to inform the patient, but also the attending physician or orthotist, in an appropriate way about the success of the treatment. The development of such sensor systems for smart braces in combination with innovative manufacturing methods has significant application potential and added value for manufacturers, users, and professionals.

The activity will be implemented in stages:

• Technology for monitoring therapy in the home environment
  • Development of new sensor solutions
  • Development of rehabilitation management systems in the home environment
  • Validation of developed solutions in a clinical setting

• Smart Orthotics
  • Sensory solutions for monitoring and controlling treatment with orthotic devices
  • Laboratory verification and validation against precision monitoring systems
  • Validation in clinical trials
  
  
• Validation of results, treatment of intellectual property, publishing

Main activity 2 Diagnostic telemedicine

The field of diagnostic telemedicine is currently very well developed, especially in the field of home diagnostic medical devices but also in the field of wearable devices for non-diagnostic monitoring of biological signals. However, for almost all known implemented technical solutions it is necessary to ensure the interaction of the monitored person to varying degrees. From active independent measurement of biological signals to at least the need for regular charging of wearable devices. Within the project we will focus on the development of completely contactless sensor systems for monitoring basic biological signals, namely radar systems. Known technical solutions in the field of radar diagnostics are based on the principles of impulse wideband radars (IR-UWB) or frequency modulated continuous wave radars (FMCW) using millimeter waves. Their basic shortcomings are due to their limited detection range, where distances over higher units of meters lead to a significant reduction in detection capability, and at the same time to the impossibility of effective suppression of artifacts in measurements caused by the movement of the person or persons within the detection area. The aim of the research activity is the development of such radar systems that allow to monitor selected biological signals with adequate quality in the natural environment of the monitored person's dwelling without the necessity of restricting the movement activities of the monitored person. The research will focus on technical solutions and possibilities of using MIMO antenna systems, but at the same time on modern algorithms for preprocessing of measured signals, eliminating random motion activities. The sensor systems thus developed will be subsequently tested and validated in various application scenarios ranging from monitoring diabetics, people with metabolic disorders to remote home care applications for the elderly or people in recovery.

The second area of development of telemetric diagnostic solutions within this project is in close connection with physical activities and their influence on selected diseases, such as manifestations of chronic diabetes. It involves the development of a unique wearable diagnostic device for measuring shear forces between the limb and the footwear. Currently there are wearable measuring devices in the form of sensory shoe insoles that can measure the force exerted by the foot on the pad only in the perpendicular direction - measuring the pressure exerted by the foot on the pad. The measurement of shear forces (the force between the pad and the sole of the foot in the plane of the pad) is currently only possible under static conditions on stationary measuring devices. A new technical solution will be proposed to measure shear forces during gait in the diagnostic environment of a movement laboratory or subsequently a doctor's office.  The proposed technical solution consists of the mechanical design of a specific sandwich shoe insole, which will allow the implementation of sensors enabling the measurement of shear forces between the sole of the foot and the footpad. At the same time, the technical solution will incorporate inertial sensors that will allow accurate detection of the progress of the executed foot/pad/boot movement in space, as well as an innovative solution for pressure measurement. The newly developed measurement system will enable early diagnosis and prediction of soft tissue damage in the foot, for example in diabetes melitus but also in other applications.

The activity will be implemented in stages:

• Development of methods for non-contact monitoring of biological signals
  • Experimental facility for radar systems for monitoring biological systems
  • Development of new analysis methods
  • Experimental verification in living labs
    
    
• Cooperation with VP4
• Validation of results, treatment of intellectual property, publishing

Main activity 3 Activities of daily living

Monitoring activities of daily living is a complex technical task that combines different sensor systems and approaches to analyze the measured signals. It has a wide application area.  It ranges from pure telemetry systems for monitoring the behavioural evolution of persons affected by chronic diseases, detection of acute behavioural changes caused by sudden changes in health status, detection of dangerous situations such as a fall in the home environment, to home automation systems providing quality of life enhancement for disabled persons or the elderly. Therefore, this sub-activity is crucial for the collaboration through the different work packages of the LERCO project.

The development of solutions for the monitoring of daily life activities requires special test environments, in the form of living laboratories, which allow high-quality verification of the proposed technical solutions and algorithms in an almost natural environment, allowing at the same time precise measurement of the monitored parameters. The project will make use of the living laboratories of the University of Science and Technology, which allow both the development and subsequent validation of new technical solutions and long-term tests in a monitored real-world living laboratory.

The first area being developed in this activity is a methodology for monitoring and evaluating the daily rhythms of people in applications that are jointly defined with the staff of the other work packages of the LERCO project. These are the areas of monitoring the level of daily activities that influence a healthy lifestyle with respect to the surrounding environment and its quality or suitability for, for example, outdoor sports activities. Another scenario is the monitoring and support of people with metabolic disorders to maintain the effectiveness of their treatment. Another area under development is monitoring the development of chronic diseases in the elderly population. A separate challenge is in the area of fall detection using non-contact measurement systems that are easily implemented in the homes of seniors or persons at risk of falls, and at the same time in hospital settings.  All of the above application areas require specific sensor solutions and new algorithms for automated decision-making systems.

At the same time, the activity will develop new sensory solutions that fit into the area of monitoring non-biological variables, more specifically the monitoring of the environment and its qualities that are affected by human activities. This will be followed by the development of home automation elements that enable people with disabilities to control their homes.

The activity will be implemented in stages:

• Assistive technologies for improving quality of life and preventing crisis situations in the home (hospital) environment
  • Development of a methodology for the analysis of circadian rhythms and personal behaviour to predict the development of chronic diseases
  • Development of systems for fall detection in the home environment
  • Development of systems for fall detection in the hospital environment
  • Experimental verification in living laboratories
  • Support for cooperation with VP4
  
  
• Activity monitoring by sensing non-biological parameters
• Validation of results, treatment of intellectual property, publishing

Subtask 3 Cardiovascular system

Main activity 1 New methods for determining cardiac output (CO)

The aim of the research is the development and validation of a new diagnostic method for the determination of cardiac output based on the dilution principle. The development of the method also includes the development of technical means, namely a continuous sensor for measuring glucose concentration and a catheter for the possibility of measurement in the bloodstream. According to early research, the method achieves higher accuracy than existing dilution methods. The activity will include a comparison of the method with conventional measurement methods, CO, and its preclinical validation in an animal model. The activity also includes the development of algorithms for accurate determination of cardiac output from measured data.

The activity will be implemented in stages:

• Analysis of technical solutions for continuous glucose measurement
  • Design and implementation of a sensor for continuous measurement of whole blood glucose
  • Design of the technical solution of the sensor construction
  • Design of sensor production technology
  • Prototype production for sensor production
  • Development of equipment for processing measured sensor data
  • Laboratory verification of the sensor
  • Sensor optimization based on laboratory tests
  
  
• Validation of a prototype sensor in an animal model
• Continuous sensor optimization for online application in the catheter
  • Miniaturization of the sensor
  • Technical design of the catheterisation system
  • Verification of the catheterisation system in laboratory conditions
  
  
• Validation of the Glycodilution method in an animal model
• Validation of results, treatment of intellectual property, publishing

Main activity 2 Measurement and processing of ECG in radiodiagnostics

The main activity deals with two areas.

Development of polymer and capacitive electrodes

Nowadays, there are so-called MRI-compatible electrodes (mostly carbon) and conductors (optical fibres) that are not ferromagnetic. On the other hand, these electrodes and conductors also have their limitations. In general, when the ECG signal is sensed during an MRI examination, interference signals are induced into the measured ECG signal, which can be completely distorted by this interference. We aim to develop new biopotential electrodes and conductors whose materials would allow safe and high-quality ECG signal sensing during the examination not only in MRI but also in other imaging modalities. One of these potentially suitable materials is the conducting polymers polyaniline and polypyrrole. Both polymers allow sensing and conduction of biopotentials (their conductivity is at the level of semiconductors), do not contain metal nanoparticles and are non-toxic. In addition, they can be polymerized onto almost any hydrophilic object and made into a polymer biopotential electrode or conductor. Thus, the aim of this activity is to analyse polymer electrodes and verify them with standard electrodes on the quality of the sensed signal in imaging and the degree of image artefacts generated.

Diagnosis of myocardial infarction

Insufficient blood supply to the heart - coronary heart disease (CHD) is one of the most common causes of death worldwide. For this reason, there is a need to address the issue of vectorcardiography (VCG) as another electrocardiographic diagnostic method and its use for the automated diagnosis of IHD. VCG offers the possibility of a non-invasive, accessible and inexpensive examination that can detect proper cardiac function, highlight warning signals, evaluate the status of coronary artery disease during routine clinical examination and provide refining information in addition to the commonly used 12-lead ECG method. The ECG describes the electrical space of the heart (EPS) quantitatively by a set of features that are suitable for further processing and automatic evaluation of IHD using computational techniques and classification algorithms. The main objective is to design an algorithm for automatic classification of patients with MI based on VCG recordings using methods of the widely developed Laufberger octant theory of VCG for quantitative description of EPS and using cybernetic approaches for signal processing by digital filtering (FIR), wavelet transform (WT), statistical analysis for selection of significant features of EPS and classification methods of loglinear modeling (LR) and artificial intelligence (MLP). One of the other goals is to create a database of VCG records with validated case physiology and thus to extend the classification capabilities of the proposed algorithms to recognize acute and stable forms of IHD with specification of myocardial status, extent and localization of IHD. Algorithm optimization is possible at all stages of the classification process discussed below, from measurement, data preprocessing, feature selection and significant EPS features to classification methods, in order to refine the electrocardiographic diagnosis.

The activity will be implemented in stages:

• Development of polymer and capacitive electrodes
• Analysis of materials for the production of dry and non-metallic electrodes
  • Production of prototype electrode systems and their verification with floating electrodes
  • Modelling of the influence of electromagnetic field on biosignals sensed by dry and capacitive electrodes
  • Implementation of electrodes in selected clinical applications
  
  
• Diagnosis of myocardial infarction
  • Analysis of ECG lead transformation methods
  • Statistical analysis of transformation methods
  • Development and construction of an instrument for experimental measurement of VCG records
  • Database of physiological VCG
  • Development of algorithms for VCG signal preprocessing
  • Development of algorithms for VCG symptom analysis
  • Extraction of VCG features for IM prediction
  • Verification of the significance of the analysed VCG
  • Design of classification methods to support diagnosis
  • Algorithm optimization
  
  
• Validation of results, treatment of intellectual property, publishing

Subtask 4 Medical data processing

The research project is primarily focused on the use of recent intelligent methods of image data processing in selected biomedical applications, in particular machine learning, image registration and segmentation methods for automatic analysis of image signals. This project will address methods used to create a Brain Computer Interface (BCI) system for rehabilitation purposes, the development of an autonomous biometric system for oncology patient recognition, the development of a system for automatic diagnosis of articular cartilage, and the creation of a system for liver virtualization during surgical procedures.

Main activity 1 Development of algorithms based on fuzzy logic

One of the essential operations in image processing is segmentation, which leads to the identification and consistent quantification and classification of objects of interest from medical images. These methods have the potential to fully replace the subjective annotation of tissues of interest in the context of clinical diagnostics. Within this activity, modern methods of detection and recognition of selected biological structures from medical images will be developed. The prerequisite is the development of modern regionally oriented methods with elements of artificial intelligence, using the principles of fuzzy logic and machine learning, which have the potential of automatic detection of selected biological tissue, as well as the interpretation of the region of the tissue. A prerequisite is the development of hybrid methods that, in addition to the above-mentioned detection, will also allow quantification of tissue characteristics based on selected geometrically and luminosity-oriented features. The development of these methods will also include robust testing of segmentation quality compared to the gold standard, as well as robustness of methods against additive deterministically determined noise with dynamic intensity for objective assessment of segmentation robustness under variable image conditions.

The activity will be implemented in stages:

• Review of the current state of knowledge of modelling of biological tissue from medical images
• Design and implementation of algorithms for detection and modeling of biological tissue from medical images
  • Implementation of algorithms based on fuzzy logic for detection and quantification of biological tissue from medical images
  • Testing algorithms on real data
  • HW implementation of solutions for acceleration of calculations
    
    
• Validation of algorithms for accuracy evaluation
  • Data collection for algorithm validation
  • Definition of gold standards for evaluating the effectiveness of algorithms
  • Objective testing and evaluation of algorithms against the gold standard
    
    
• Validation of results, treatment of intellectual property, publishing

Main activity 2 Augmented reality in surgery

The application of augmented reality methods in surgery represents one of the modern trends for planning liver surgery. These methods should be able to virtualize the operating space, areas of interest, overall morphological structure of the liver and surgical tools that are used in surgery to simulate surgery and individual procedures to minimize harm to the patient. The aim of this activity is to design and implement a complex software system that can simulate the procedures that will be performed within the operation. In the first stage of the activity, a comprehensive in-depth search of the current state of knowledge of scientific methods and procedures that are used in augmented reality applications for planning therapeutic procedures of surgery in clinical practice will be carried out with the aim of optimal design of the augmented reality system for liver surgery. The assumption is that the system will use image data from clinical CT scans and real records from the surgery. The aim is to create intelligent registration algorithms for registering these records, where the surgeon will be able to simultaneously virtualize the morphological details of the area of interest within the real scene he sees during surgery, including the simulation of surgical instruments. Part of this SW solution will be artificial intelligence methods, which should automatically extract the liver area from individual image modalities, these models will then be used for the aforementioned registration of image domains. This intended tool, which enables virtualization of operational performance, will have a wide range of applications in the simulation of operational performance, which predetermines the wide possibilities of applications.

The activity will be implemented in individual stages (in some phases in close cooperation with the VP7):

• Analysis of the current state of augmented reality in surgery
  • Defining Literature Search Criteria
  • Research of the current state of knowledge for augmented reality in surgery and current HW resources
    
    
• System for augmented reality in surgery
  • Design of augmented reality algorithms in surgery
  • Design and implementation of a system for the implementation of augmented reality in surgery on a model
    
    
• Creation of a 3D model of the liver
  • Creating automatic 3D reconstruction using neural networks
  • Data transfer back with ePACS
    
    
• Optimization of algorithms, finalization of 3D liver model
  • Optimization of algorithms and imaging software
  • Displaying the resulting 3D model
    
    
• Imaging an augmented reality system
  • View a finalized 3D model in augmented reality mode
  • Automatic segmentation of the actual image
  • Algorithmic alignment of virtual and real images
  • Precise targeted display of the virtual image to the right place in the real image
  • Continuous research into the possibilities of automatic detection of distortion and changes in a real object and corresponding model changes
    
    
• Validation of results, treatment of intellectual property, publishing

Main activity 3: Image processing (Radiotherapy patient ID)

The aim of the activity is the design and implementation of an autonomous biometric system with the aim of recognizing a patient who has indicated therapeutic oncological treatment. It often happens that patients undergoing oncological procedures are apathetic, suffer from poor orientation, and also have difficulty communicating with medical staff. These stimuli lead to the development of an autonomous biometric recognition system that would be able to classify and recognize a specific patient based on unique patient markers. This system will be connected to a database of data of a specific patient, which will make it easier for medical staff to correctly and quickly set up the appropriate procedures within the indicated oncological treatment. The first part of the activity will be a search of the current state of knowledge within the classification of facial markers with the aim of selecting optimal methods for the design of the biometric system. It is assumed that the system will be set up for face classifications from the camera system, which will be located in the reception department. As part of the activity, a robust objectification analysis of the system will be carried out, where great emphasis will be placed on robustness in the context of possible visual changes in facial features during oncological treatment to ensure high recognition efficiency under variable conditions. The robustness of the system is a major advantage in contrast to conventional methods of facial classification, which often use standardized images and are difficult to respond to the dynamics of changes in facial markers. This complex biometric system will be implemented into an autonomous HW solution based on FPGA architecture with the aim of accelerating calculations and simple installation without the need to solve the reliability and computing capabilities of the computer.

The activity will be implemented in individual stages (in some phases in close cooperation with the VP7):

• Analysis of the current state of knowledge for patient ID in radiotherapy
  • Definition of criteria for search studies
  • Creation of review for fNIRs and patient ID analysis
  
  
• Design of a software system for classifying patients on oncology
  • Research of classification methods for biometric identification of patients in oncology
  • Implementation of appropriate classification algorithms for patient identification
  • Analysis of efficiency and robustness of proposed classification methods
  
  
• Implementation of the system for patient ID in oncology
  • Creation and data collection for patient ID testing in oncology
  • HW implementation of algorithms for acceleration of calculations
  • Testing of a comprehensive solution in real conditions

• Validation of results, treatment of intellectual property, publishing

Main activity 4: Real-time processing of biological signals

The aim of the activity is to design an autonomous system to create an interface between electrical signals originating in the brain and selected rehabilitation procedures with the aim of optimizing the therapeutic effect and controlling these procedures. The major application area of this system will be the optimization of therapeutic procedures in patients with deficit after strokes, spinal cord injuries, degenerative diseases of the nervous system, such as ALS, and also as an extended treatment of muscular dystrophy.

The activity will be implemented in individual stages:

• Analysis of existing recent methods for real-time processing of biological signals
• Design and implementation of a model of processing biological signals in real time
  • Creation of a software model
  • Pilot testing of the model on real data
  • HW implementation of algorithms
• Verification and testing of the model for real use
  • Design of standardized verification procedures
  • Testing the model against standards
  • Evaluation of the accuracy and robustness of the model
  
  
• Validation of results, treatment of intellectual property, publishing