In this study, a novel pipeline for retrieving and ranking DICOM images similar to a provided query image was presented. The aim of this study is to support cross-diagnosis through visual and semantic similarity. The presented approach combines deep learning-based feature extraction, metadata comparison, and vector database search. Embeddings are generated using pretrained neural networks and indexed with FAISS to enable efficient nearest-neighbor retrieval. Dimensionality reduction via PCA is applied to enhance computational efficiency. Retrieved results are initially ranked using embedding-based similarity, then refined by computing a joint similarity score that fuses image features with DICOM tag metadata. The final rankings are evaluated using Mean Reciprocal Rank (MRR) to assess retrieval performance. The presented pipeline demonstrates the feasibility of combining visual and semantic information to improve the relevance and interpretability of medical image retrieval.

Pedestrians, as vulnerable road users, pose safety challenges for autonomous vehicles (AVs). Their behavior, often unpredictable and subject to change, complicates AV-pedestrian interactions. To address this uncertainty, AV s can enhance safety by communicating their planned trajectories to pedestrians. In this research, we explore the interaction between pedestrians and autonomous vehicles within an industrial environment, focusing on how communicative behavior from the vehicles influences pedestrians' physiology. We investigate the possibility of mea-suring biosignals while participants wear a VR headset and experiment a pedestrian crossing. Our preliminary study reveals subtle variations in delta rhythms when users immersed in VR simulations interact with AV s that either provide or withhold additional information.

This thesis presents a real-time skin assessment system with the main aim of detecting inflammatory acne lesions and tracking skin conditions. A facial acne lesion detection algorithm was developed, using a pre-trained YOLOv8 model for lesion detection and a Mediapipe detector for face detection. The proposed solution aims to create a system used in smart mirrors to help users self-monitor and monitor their skin condition in real time without much involvement. The work shows that it is possible to effectively detect skin changes even in a motion situation, which could be a good direction for the development of solutions based on video analytics.

There are many approaches to non-invasively deter-mine blood pressure. Among them there are methods based on utilizing pulse transit time measured by means of photoplethys-mography. The variability of the blood pressure drop between two measurement sensors placed along the artery and its dependence on the selected parameters describing the cardiovascular system is presented in the paper. This pressure drop modifies the pressure pulse propagation velocity, thus also value of PTT. The properties of four fiducial points on the radial and brachial arteries were analyzed. Publicly available simulation data were used for the analyses. The best results were achieved for the point defined by the maximum of the first derivative of the signal.

In this work we analysed the application of con-volutional neural networks in motor imagery classification for the Brain Computer Interface (BCI) purposes. To increase the accuracy of classification we proposed the solution that combines the Common Spatial Pattern (CSP) with convolutional network (ConvNet). The electroencephalography (EEG) is one of the modalities we try to use for controlling the prosthetic arm. Therefor in this paper we exploited the subject dependent approach and show results for models trained individually for a particular subject. Although the ConvNets are design to work directly with EEG data, presented approach of joining CSP with ConvNet shows increase in accuracy of movement classification. In average, our approach resulted in ∼80% accuracy.

In this paper we present comprehensive system for communication with computer by gaze. One of the main assumptions behind this work was to provide solution that can be used with standard RGB webcam. The proposed comprehensive system included the eye tracking module and user interface for convenient gaze interaction with computer. As a result a fully functional application was developed. The average accuracy of the eye tracking module was 4.1°while precision was 0.042°.

This paper presents how neural networks can be utilized to improve the accuracy of reach and grab functionality of hybrid prosthetic arm with eye tracing interface. The LSTM based Autoencoder was introduced to overcome the problem of lack of accuracy of the gaze tracking modality in this hybrid interface. The gaze based interaction strongly depends on the eye tracking hardware. In this paper it was presented how the overall the accuracy can be slightly improved by software solution. The cloud of points related to possible final positions of the arm was created to train Autoencoder. The trained model was next used to improve the position provided by the eye tracker. Using the LSTM based Autoencoder resulted in nearly 3% improvement of the overall accuracy.

Pulse Wave Velocity (PWV) is measured and utilized in many clinical applications. Recently, a wide research has been led to develop a cuff-less and continuous blood pressure method basing on PWV. However, in this application a decision on choosing an appropriate fiducial point of pulse wave (PW) waveform is necessary and substantial. It would allow to measure time parameters necessary to determine PWV. An influence of sampling frequency and other preprocessing algorithms on an accuracy of fiducial points determination are discussed in the paper.

In this paper the simple architecture of neural network for hand gesture classification was presented. The network classifies the previously calculated parameters of EMG signals. The main goal of this project was to develop simple solution that is not computationally complex and can be implemented on microprocessors in low-cost 3D printed prosthetic arms. As the part of conducted research the data set EMG signals corresponding to 5 different gestures was created. The accuracy of elaborated solution was 90% when applied real time on data sampled with 1kHz frequency and 75% when applied real time on data acquired and process directly on microprocessor with lower,100Hz sampling frequency.

In this paper the main aspects of mechanical design behind the low-cost prosthetic arm are presented. The fundamentals of a proper design has been defined to obtain functional 3D printed 5 degree of freedom (DOF) prosthesis. The designed prosthetic arm is a part of the hybrid interface with eye tracking movement control. The main focus was to create affordable but usable prosthesis which corresponds in size and weights to the human arm. The iterative process (starting from the final segment of the arm) was used to design fully functioning arm. All the elements were evaluated regarding total weight and the maximum load that can be carried by the arm. The result of this work is a prototype that weighs below 6kg and has a range of motion comparable to the human's arm. Final product is able to freely move an object of a total weight of 1 kg. All the mechanical parts of the designed arm were 3D printed which therefore presented construction can be adopted by people with different disabilities and (when connected to interfaces like EEG, EMG or eye tracking) provide support in everyday life activities.

A non-invasive and continuous blood pressure estimation could allow a better diagnosis and an earlier detection of various diseases. It could be performed using a photoplethysmography. However, it requires that a relation between blood pressure and pulse transit time is known. Eleven theoretical formulas were used in obtaining the simulated and noise free synthetic data. Then, they were utilized in validation of three, commonly used, formulas for blood pressure estimation. Finally, the best one was indicated.

In this preliminary study we, investigate the possibility of user recognition techniques suitable on smart home devices like chairs, beds, aiming for low-power, high accuracy and quick response time. We propose the two well know technique: voice speaker recognition and accelerometer signal from device mounted on the chair, and the third one optical system basing on IR LED transmitter/receiver circuit. The preliminary results proved that the developed system based on measurement and analysis of the accelerometer signal is very promising and suitable for implementation in the final product. The optical system requires further work but potentially it can be used for user identification purposes.

This paper describes the possibility of monitoring the progress of knowledge and skills acquisition by the students of radiology. It is achieved by an analysis of a visual attention distribution patterns during image-based tasks solving. The concept is to use the eye-tracking data to recognize the way how the radiographic images are read by recognized experts, radiography residents involved in the training program, and untrained users who graduated from biomedical engineering. The results of research presented in this paper support the usefulness of earlier elaborated eye metrics, Visual Attention Time Product VATP, for skills measurement when using Electronic Medical Record for CT images evaluation. Moreover, it appears that it is possible to differentiate the level of residents skills in radiology structures recognition. Presented methodology can also be utilized in other areas where operator performs image related tasks.

EDR is an interesting measuring technique that allows an indirect assessment of respiratory activity. This is an alternative solution to direct methods that are based on the measurement of air flow, which require a specialized sensor or even a system. However, due to inter-personal anatomical differences, the optimal ECG lead (placement of the electrodes) ensuring the best EDR signal quality is not fixed. An influence of ECG lead geometrical relation to the heart axis on the EDR signal quality has been examined in the paper. It was found that optimal electrodes position strongly differs between individuals. For each person there were at least few leads with high correlation with reference signal (over 0.8). They allowed for respiration frequency estimation with almost 100% accuracy.

Impedance properties of bath solution are compared. Any ingredient such as soap, shampoo or bathing salt changes bath-solution's electrical properties eg. impedance. We have measured and analysed influence of typical bathing ingredients on electrical properties of the solution. We investigate impedance changes over frequency range from 100Hz tu 100kHz as it is most applicable for person detection. Electrical properties of bath solution knowledge is essential for person detection and biosignal acquisition of bathing person.

In this paper the possibility of person identification based on biosignal is investigated. The work focus on the analysis of the changes in intervals between successive R-waves of electrocardiogram (ECG) recorded by wearable electronics in form of a necklaces. The main idea behind this project is to find efficient tool which may prevent sudden consciousness loss episodes or even sudden death episodes related to rapid temperature changes. Proposed method relays on the estimation of the RR interval changes in time with three approximation functions. Then, the identification of a person is performed based on the parameters of the utilized estimation functions.

The estimation of systolic and diastolic blood pressure using artificial neural network is considered in the paper. The blood pressure values are estimated using pulse arrival time, and additionally RR intervals of ECG signal together with respiration signal. A single layer recurrent neural network with hyperbolic tangent activation function was used. The average blood pressure estimation error for the data obtained from 21 subjects from MIMIC database was equal to 2.490 mmHg with standard deviation equal to 1.063 mmHg for systolic blood pressure, and was equal to 1.330 mmHg with standard deviation equal to 0.627 mmHg for diastolic blood pressure using vanilla recurrent neural networks. Similar results were obtained for long short term memory cells. The simulation shows that taking into account pulse arrival time together with RR intervals and respiration signal gave better results than pulse arrival time alone.

The electromyographic activity of muscles was measured using a wireless biofeedback device. The aim of the study was to examine the possibility of creating an automatic muscle tension classifier. Several measurement series were conducted and the participant performed simple physical exercises - forcing the muscle to increase its activity accordingly to the selected scale. A small wireless device was attached to the electrodes placed on the patient's body in the area of biceps muscle. The patient body position, electrode placement and performed exercises were the features that as much as possible, minimized the impact of the surrounding muscles influence. The data were recorded and an analysis was made using QT /C++ environment. The exercises were designed to enable evaluation of muscle activity according to Lovett scale. The aim of the research described in this article was to help in the assessing of the muscle strength tension to assess progress in rehabilitation. The designed feed-forward neural network allowed classification of recorded signals with 78% accuracy.

In a smart home, health and well-being monitoring systems could be embedded in everyday devices providing a pervasive care. A home bathroom scale is an example of such a device, typically used to measure body weight and very often its composition (e.g. body water/fat percentage). In this paper, we analyzed a potential use of the bathroom scale to measure electrocardiogram (ECG) from electrodes located on the scale's tile. In particular we used both, simulations and real measurements to analyze the possibility of a such approach. We originally created a simple finite element model to analyze a role of a vector cardiogram projected on the foot-foot lead. We also analyzed results obtained during experiments in which took part (in total) 13 healthy volunteers. They were examined both, in sitting and standing positions. As a result we showed that the quality of the recorded ECG signal from feet is highly related to the electrical vector orientation of the individual's heart. We additionally compared the detected QRS complexes from hands-and feet-sampled ECG for sitting and standing positions. Results showed that for the sitting position the difference was only for 3.6% of all QRS complexes, while for standing one about 24.9%.

In this paper we propose to use temporal muscle contraction to perform certain actions. Method: The set of muscle contractions corresponding to one of three actions including “single-click”, “double-click” “click-n-hold” and “non-action” were recorded. After recording certain amount of signals, the set of five parameters was calculated. These parameters served as an input matrix for the neural network. Two-layer feedforward neural network with one hidden layer of 200 neurons was applied to classify gestures based on the input matrix. Results: The network was trained using the dataset consisted of 43 samples and then tested on the 34 samples dataset. All gestures from the test set were correctly classified.

In this paper we present the research conducted on synchronous measurements of biosignals. The experiment was conducted to evaluate the possibility of estimating vital signs based on eye tracking. Method: The eGlasses platform was used for acquisition of ECG, respiration rate, eye and pupil movement and blood pressure. Data were acquired in three 5 min. intervals during which a subject was performing certain tasks. The signals were filtered and resampled. The correlation between blood pressure, respiration and pupil dilation was calculated. Results: The measurements were conducted on 12 healthy volunteers. The obtained correlation coefficient was similar for all people and measurement positions. Conclusions: The electronic eyewear like eGlasses can be utilized to measure and process essential vital signs. Although some results are promising, estimating a value of blood pressure or hart rate based on data provided by eye tracking module requires some further studies.

Nearly 10% of all upper limb amputations concern the whole arm. It affects the mobility and reduces the productivity of such a person. These two factors can be restored by using prosthetics. However, the complexity of human arm makes restoring its basic functions quite difficult. When the osseointegration and/or targeted muscle reinnervation (TMR) are not possible, different modalities can be used to control the prosthesis. In this paper the usability of electromyography (EMG) signals for such a control is evaluated. Method: first, the types of operations performed by the prosthetic arm that could be handled by EMG module were defined. The raw EMG signal, corresponding to the predefined gesture, was acquired from the surface of trapezius muscle. The pattern recognition neural network was trained to classify gestures based on recorded RAW data. Results: The neural network was trained using 56 signals corresponding to performed gestures. Optimal performance was achieved for 29 training cycles. The network was tested using data set of 56 gestures. The designed network was tested on gestures recorded from 10 volunteers. The gestures were correctly classified with nearly 84% accuracy. Conclusions: The EMG analysis is a reliable modality when it comes to hybrid interfaces for control over prosthetic arm.

A technique for the acquisition of an increased number of pupil positions, using a combined sensor consisting of a low-rate camera and a high-rate optical sensor, is presented in this paper.The additional data are provided by the optical movement-detection sensor mounted in close proximity to the eyeball.This proposed solution enables a significant increase in the number of registered fixation points and saccades and can be used in wearable electronics applications where low consumption of resources is required.The results of the experiments conducted here show that the proposed sensor system gives comparable results to those acquired from a high-speed camera and can also be used in the reduction of artefacts in the output signal.

The paper presents an example of model-based estimation of blood pressure parameters (onset, systolic and diastolic pressure) from continuous measurements.First, the signal was low pass filtered and its quality was estimated.Good quality periods were divided into beats using an electrocardiogram.Next, the beginning of each beat of the blood pressure signal was approximated basing on the function created from the sum of two independent distributions: Gaussian and exponential.The nonlinear least square method was used to fit measurement data to the model.The initial conditions for the fitting procedure were selected for each beat on the basis of its parameters.Finally, the diastolic and systolic values of blood pressure and onset were determined.

Respiration rate is a very important vital sign.Different methods of respiration rate measurement or estimation have been developed.However, especially interesting are those that enable remote and unobtrusive monitoring.In this study, we investigated the use of smart glasses for the estimation of respiration rate especially useful for indoors applications.Two methods were analyzed.The first one is based on measurements of respiration-related body movements using an accelerometer.The second one uses the thermal camera to observe temperature changes in the nostril region.For both methods signals were extracted, filtered and processed using two different respiration rate estimators.Both methods were validated during experiments with the participation of volunteers using the respiration belt as a reference measurement method.Results proved that for both methods it is possible to reliable estimate the respiration rate with Root Mean Square Error lower than 2 breaths per minute, which is sufficient for medical screening.

In this preliminary study we analyzed the possibility of the reliable measurement of biomedical signals with some potential hardware extensions of smart glasses. Using specially designed experimental prototypes four category of biomedical signals were measured: electrocardiograms, electromyograms, electroencephalograms and respiration waveforms. Experiments with volunteers proved that using even simple construction of sensors it is possible to reliable measure biomedical signals with the quality enough for screening purposes as for the needs of simple interaction between an user and smart glasses.

In this paper we investigate the effectiveness of the interaction using eye tracking and electromyography. Smart glasses requires reliable interfaces for controlling the graphic content displayed directly in front of the user's eye. Presented research is related with the eGlasses project, which is focused on the development of an open platform in the form of multisensory electronic glasses and related interaction methods. One of the implemented interaction methods is the one based on eye tracking module. The eye tracking provides quite accurate pointing but there is always a matter of how to select and confirm while using the eye tracking. In this paper we combine the eye tracking with the EMG signal acquired from the skeletal muscles. Both, the advantages and limitation of the method are discussed.

In this paper we present the possibility of using eGlasses eye tracking module as an extension for Scratch programming tool which is a visual programming language supporting computer skills learning. The main concept behind this project is to setup the interface for rapid interaction design. Eye tracking is a powerful tool for hands free communication but for that requires a dedicated software. This software is rarely tailored for a specific needs of a potential user. It means that people who wants to explore gaze driven computer interfaces often need to create their own eye tracking software or need to have certain computer skills to be able to go through the software development kit (SDK) provided by the eye tracker vendors. It makes designing interaction by gaze quite challenging. The eGlasses eye tracking module was used to extend sensor values available in Scratch by the information about pupil position, gaze position and eye opened/eye closed signal. In this paper we have tested if the gaze data acquired by the eGlasses are reliable and can be used as an input signal for rapid interaction design using Scratch. The whole interface was validated regarding the system usability score (SUS). The purpose of our work is to support disabled children and their therapists but also to give a chance to a wider community, without computational skills, to easily check they ideas for gaze based interaction. The results of conducted studies with the participation of volunteers give strong foundations to support this statement.

In this study we evaluated system architecture for the use of smart glasses as a viewer of information, as a source of medical data (vital sign measurements: temperature, pulse rate, and respiration rate), and as a filter of healthcare information. All activities were based on patient/device identification procedures using graphical markers or features based on visual appearance. The architecture and particular use cases were implemented and verified using smart glasses prototypes developed under the eGlasses project and using a reference Health Level 7 Fast Healthcare Interoperability Resources (HL7 FHIR) server. The results show that information about the identified patient can be quickly retrieved from FHIR servers and annotated using voice recognition services. Smart glasses can be used in the measurement of vital signs of the observed patient, providing values of body temperature, pulse rate, and respiration rate by means of non-contact measurements. Such measurements are sufficiently reliable for medical screening and for fast data exchange using HL7 FHIR actions.

This paper presents the algorithm for validation of electrodes locations for the thoracic impedance measurements. In particular the presented algorithm was designed to perform the telemetric sleep apnea monitoring. One of the problems, during the clinical tests of a developed device, was to preserve the repeatability of measurements. It strongly depended on the appropriate electrodes placement on the examined person's thorax. It seems that the developed software, which allows inspecting the electrodes localization, significantly increases a possibility of impedance based sleep apnea monitoring at home. In this article, the properties of the evaluation algorithm of electrodes placement accuracy are presented.

In this paper we investigate the effectiveness of the gaze interaction within near-to-eye display. The practical aspect of the paper is about combining an eye tracker with smart glasses. Presented research is related with the eGlasses project, which is focused on the development of an open platform in the form of multisensory electronic glasses and related interaction methods. One of the implemented interaction methods is the one based on eye tracking module. Both, the advantages and limitation of the method are discussed. We also considering the calibration free method of fixation points estimation within the near-to-eye display.

An influence of head movements on the gaze estimation accuracy when using a head mounted eye tracking system is discussed in the paper. This issue has been examined for a multi-display environment. It was found that head movement (rotation) to some extent does not influence on the gaze estimation accuracy seriously. Acceptable results were obtained when using eye-tracker to communicate with a computer via in two displays simultaneously.

This paper presents the basic ideas of eye and gaze tracking in multiple-display environment. The algorithm for display detection and identification is described as well as the rules for gaze interaction in multi display environment. The core of the method is to use special LED markers and eye and scene tracking glasses. Scene tracking camera registers markers position which is then represented as a cloud of points. Analyzing the mutual positions of detected points the algorithm estimates screen/displays position. Display number is assigned based on special information marker. Described project shows the possibility of hands free interaction in MDE. It also presents how to register visual attention when information is dispersed on several screens.