Alzheimer's Dementia (AD), one of the age-related neurological disorders, causes loss of cognitive functions and seriously affects the daily life of patients. Electroencephalogram (EEG) is one of the most frequently used clinical tools to investigate the effects of AD on the brain. In the proposed study, a time-frequency representation and deep feature extraction based model is introduced to distinguish EEG segments of control subjects and AD patients. TF representations of EEG segments are obtained using high-resolution SynchroSqueezing Transform (SST), and conventional short-time Fourier transform (STFT) methods. The magnitudes of SST and STFT are used for deep feature extraction. Various classifiers are used to classify the extracted features to distinguish the EEG segments of control subjects and AD patients. STFT based deep feature extraction approach yielded better classification results than that of the SST method.
In this study, it was aimed to design smart walking stick in purpose of detect obtracles that cause danger for blind people. The selected sensors and electronic components were programmable with Arduino. Ultrasonic distance sensor was used for detect obtracles in determined closest distance. A buzzer was used to provide audible alert when obstracles were detected via ultrasonic distance sensor. Additionaly, experimential studies were applied on GPRS sensor which is used for send locational information via SMS to caretaker of blind person when she or he lose the way.
The Auto Train Brain mobile app improves dyslexia symptoms, according to a clinical trial. There was just one unique neurofeedback user interface in the original mobile app, which gave visually and vocally rewarding feedback to the subject via a colorful arrow on the screen. Later, new modules are added to the app in response to end-user requests, such that users can choose from a variety of “youtube” videos and begin neurofeedback sessions with both visual and vocal rewards, or they can choose from a “storyteller” and begin neurofeedback sessions with only vocal rewards. In this study, we looked into whether the multimodal neurofeedback method is more effective in improving sample entropy in the gamma band in the left temporal region for children with dyslexia.
The aim of this study is to determine the success of different parameter optimization methods in a classification done by using a new Convolutional Neural network (CNN) proposed in this study on a new original dataset consisting of lung CT images collected by the researchers from ill/healthy people in order to diagnose COVID-19. To achieve this purpose, the researcher prepared a new data set involving computerized tomography images collected from Erzurum City Hospital Emergency Department for the diagnosis of COVID-19 after the necessary permissions were taken. This new data set consists of 1081 lung CT images, 568 of which belong to 313 people infected with COVID-19. The results obtained for the proposed CNN model through different optimization methods were compared. Precision, recall and f1-score parameters were used while evaluating the results. In our dataset, the highest performance values for the proposed CNN model were achieved by AdaMax optimizer as follows: %89.86 accuracy, %86.51, precision, %95.61 recall and %87.33 f1-score. The classification accuracy values achieved by other optimizers were %89.40 (RMSProp), %87.10 (Adam), %81.11 (SGD).
Pain is widely acknowledged to be a complicated experience. Pain is also a symptom of joint inflammation in arthritis, such as rheumatoid arthritis and osteoarthritis. Machine learning (ML) for classification uses nonparametric supervised learning techniques called decision trees. By gaining knowledge from decision rules created from the characteristics of the supplied data, they may be utilized to forecast the target variable. We applied a machine learning decision model to anticipate j oint pain (target variable) while considering clinical biochemistry. A total of 650 patients were included who visited orthopedic OPD with joint swelling or myalgia. Decision Tree was trained tested, and cross-validated with supervised learning. The model was evaluated along with the selected features/attributes (age, gender, uric acid, & CRP). 44% of patients were diagnosed with joint pain. The decision tree model yielded an accuracy of94% and a validation accuracy of 96%. Uric acid was strongly correlated with joint pain. Early ML-based joint pain identification will help avert more significant orthopedic issues.
As noise corruption is an inevitable issue for all imaging technologies, this problem causes serious difficulties in analyzing the biological fine-details of fluorescence microscopy images. While Gaussian only, Poisson only and mixture of Poisson- Gaussian can generally be observed, the mixed-noise is more prominent in fluorescence microscopy. In this paper, a novel patch-based denoiser-learning approach is proposed for the images captured by fluorescence microscopy. The developed algorithm mainly builds upon linear-embeddings of neighboring image patches, and it learns a linear transformation between noisy and clean intrinsic geometric properties of patch-spaces. Experimental results demonstrate that the proposed “Neighbor Linear- Embedding Denoising” (NLED) has competitive performance both visually and statistically when compared to other algorithms in literature, for noise corrupted fluorescence microscopy images.
Attention Deficit Hyperactivity Disorder (ADHD) is a neurological disease that typically appears in childhood. The disease has three main symptoms in children: inattention, hyperactivity, and impulsivity. Treatment of the disease is based on behavioral studies; however, there is no definitive diagnosis method. Hence, the electroencephalography (EEG) signals of ADHD subjects are often investigated to understand changes in the brain. In the proposed study, it is aimed to process and reduce the EEG data of ADHD and control subjects (CS) by using the Douglas―Peucker algorithm and to investigate the effects of the algorithm on EEG signal analysis. EEG data obtained from 18 control subjects (4 boys, 14 girls, mean age 13) and 15 ADHD patients (7 boys, 8 girls, mean age 12) are collected. By using reduced EEG data; time features such as energy, skewness, kurtosis, mean absolute deviation (MAD), root mean square (RMS), peak to peak (PTP) value, Hjorth parameters, and non-linear features such as largest Lyapunov Exponent (LLE), correlation dimension (CD), Hurst exponent (HE), Katz fractal dimension (KFD), Higuchi fractal dimension (HFD), are calculated to examine different signal characteristics. Extracted features are used to distinguish the EEG data of ADHD and CS by using various machine learning algorithms.
Thrombogenesis and infections due to pathogen transmission caused by medical equipment interfaces are one of the biggest health problems that threaten human life. One of the recent approaches to overcome these problems is superhydrophobic surfaces with low surface energies and micro/nano surface roughness. Here, medical equipment interfaces such as gloves, lancets, surgical drapes and gowns have been made superhydrophobic. These superhydrophobic surfaces exhibited high blood repellency with a static contact angle of 172° against blood, allowing continuous blood flow without leaving any residue on the surface. Furthermore, these coatings perfectly suppressed the biofilm formation by preventing the adhesion of the most common S. aureus and E. coli bacterial species to the surfaces. Superhydrophobic material, which has over 100% biocompatibility in vitro, also exhibited high hemocompatibility, giving hope for its applicability to a number of in-body medical device interfaces such as catheters, artificial vessels, and implants.
Wrist rehabilitation robots are fundamental for helping patients with stroke or wrist injuries and also decrease the workload of physiotherapists. Though wrist rehabilitation robots are essential, recent wrist rehabilitation robots have shortcomings such as heavy weight, immobility, costliness, etc. To remedy these shortcomings, in this study, we developed and produced a low-cost and mobile robotic device for these patients with partial paralysis. The device is designed to assist the patient to perform wrist exercises comfortably in the home environment without being dependent on rehabilitation centers and/or physiotherapists. Therefore, we offer the device driven by a 3D-printed mechanism in two degrees of freedom. In addition, we analyzed and simulate the robot via finite element analysis and Solidworks respectively. The results indicate the robot provides enough force, torque, and range of motion. For these reasons, the robot can be used as a compact and lighter robot in hardware and a cheaper robot in cost. So it is feasible and affordable for real-time application in wrist rehabilitation.
Infectious diseases are serious threat to the human health and safety. Traditional methods to overcome this challenge are less effective due to emergence of resistant strains. Therefore, low cost, efficient novel antibacterial materials may be powerful alternative to combat bacteria. In this study, we present a single step, in-situ fabrication approach to antibacterial surfaces. Herein, copper nanoparticles were in-situ grown and the morphology of the paper surface was characterized with scanning electron microscopy (SEM). Energy-dispersive X-ray spectroscopy (EDX) was used to understand the composition of the surface and EDX mapping for location of the elements. It was seen that, Cu nanoparticles successfully grew on the surface and distributed homogeneously. Furthermore, feasibility of the approach for various application area was demonstrated in the glove, fabric, paper and PDMS surfaces. Moreover, mechanical stability of the surfaces was evaluated. As an advantage of in situ copper growth, the surfaces retained their antibacterial activity after mechanical abrasion and washing. Obtained results show that, developed antibacterial surface fabrication approach may provide a powerful, effective, low-cost and stable platform to combat bacteria.
The infant mortality rate is an important indicator of maternal and infant health, as well as global health status in general. In 2018, annual infant deaths were announced at 4.0 million. The main causes must be pointed out for decreasing infant deaths and unexplained death situations (SIDS). In-fants' thermoregulation systems malfunctioning or hypothermia-hyperthermia situations are proposed as the primary or ulti-mate causes of SIDS. On these foundations, before generating a model, developed thermoregulation system models were examined. Using infant-specific parameters and more precise ambient conditions, the model depth was increased. To investigate the thermal distribution of the infant body, the infant body was divided into 7 compartments, which were then sliced into layers. The thermoregulation system of infants was simulated using thermodynamic equations and generated equations. The model's accuracy was tested by comparing its findings to those from experiments and data from the literature. The results unmistakably demonstrated that the model's data and those gathered from the real system and the literature are in good agreement.
Representing data in array form is not always an efficient way. Sometimes it can cause loss of bias information that is inherent in data. With the help of taking an image to data matrix form as input instead of flattening to an array, deep learning methods, especially convolutional neural networks, are more successful than traditional machine learning techniques in terms of accuracy rate in image processing. Besides, graphs are a very efficient way to represent problems such as social networks, protein interface prediction, and images. Graph Neural Networks (GNNs) are deep learning methods that apply convolution logic to data like a graph, and many applications are efficiently applied. That is why representing histopathological images with graphs can be an advantage to know the connection between cores. The study uses GNNs to classify tissue types in the Chaoyang dataset. First, the superpixel graph is constructed from an image, and then GNNs models are applied to the constructed graph dataset. Experimental results present better accuracies than the compared literature methods.
Magneto-acousto-electrical tomography with Magnetic Field Measurements (MAET) is a new hybrid imaging modality to image electrical impedance property of body tissues. This modality couples the electrical impedance mapping property of the electrical impedance tomography and high spatial resolution of sonography. In this modality, magnetic field of induced Lorentz currents is detected by receiver coils. Electrical impedance map of the body is reconstructed using voltages induced on the receiver coils. The main aim of this study is to offer a method to use air-cored multi-wound coils with the high-quality factors for MAET with magnetic field measurement signals and remove the coil's transfer function's effects on the measured signals. For this purpose, a simulation study was conducted in COMSOL. In this simulation study ultrasound transducer, coils, and phantom's geometry and properties used in experiments were modeled. The forward problem of the MAET method with magnetic field measurement is solved and the MAET signal is obtained. In addition, a method is proposed to numerically estimate the transfer function obtained by using the electrical circuit diagram of the coil. Finally, the artifact tail on the obtained MAET signal due to the high-quality the factor of the coil is removed using the general inverse Wiener Filter.
The Ministry of Health Turkish Medicines and Medical Devices Agency has been conducting studies since 2015 on the inspection and certification of the institutions that carry out the testing, control and calibration activities of medical devices in health institutions, and the certification of the personnel who perform the calibration measurements in these institutions through trainings. Medical devices that are functionally similar to each other were grouped under the name of authorization groups and certification trainings were started, and establishment authorization continued on the basis of these authorization groups. While there were 7 authorization groups announced to be implemented at the beginning of the process, 5 additional authorization groups whose studies were completed within the elapsed time were included in the system. In this paper, it is aimed to inform biomedical graduates who want to take part in the testing, control and calibration activities of medical devices by explaining these additional authorization groups.
A variety of artificial intelligence (AI) approaches are applied for the classification of hand movements in systems that use electromyography (EMG), which measures the electrical activity of muscles. In AI approaches, machine learning (ML) is frequently preferred and researched for this classification issue. In this study, hand gesture classification was performed with ML algorithms using EMGs of 10 hand movements. Features were extracted from the time domain (TD), frequency domain (FD), time-frequency domain (TFD) (via Wavelet-based Synchrosqueezing Transform), and Fractional Fourier Transform (FrFT) domain. After training 31 ML models with all features, Subspace k-Nearest Neighbor (kNN), which is ensemble-based learning, was determined as the best model. This model was trained with different feature and channel combinations, and the classification performances were examined as channel-based and domain-based, separately. In all cases, an accuracy of 97.10% was obtained as the highest via the TD-FD-FrFT domain feature combination, including all channels. When all the results are examined, an alternative classification approach is presented to the literature by proving that the computational load decreases while the accuracy value increases by determining and utilizing the channels and features that contain more related information about hand movement.
This paper demonstrates an objective method to measure arterial resistance and compliance in humans from ar-terial vascular impedance. Conventional approaches use linear velocity to calculate the impedance spectra. We have presented a method to compute volume-flow using the area of cross-section and doppler shifted frequency, synchronously recorded along with the intra-arterial pressure. Arterial vascular impedance is calculated as the ratio of arterial pressure to volume-flow. A three-element Windkessel model has been employed to estimate the lumped parameters representing the arterial tree.
With this study, a new test system was designed to be used in the sterilization of all critical medical materials used in the health sector. A new type of oxygen molecule allatrope (YOMA) from reactive oxygen species (ROS) was developed for use in the system, tested and compared with other sterilization systems. Before the test, 12 different materials, each of the same size (2 cm2), were sterilized in an autoclave at 120 °C, 1 atm pressure for 20 minutes. In the second stage, all surfaces of the samples were contaminated with the 3M Attest 1264 Geobacillus stearothermophilus biological indicator, which was cultured in an oven at 37 °C. In the designed system, all test samples were taken into sterile petri dishes, each sample test material was placed on the solid medium surface and then taken into the desiccator. It was checked whether there was bacterial growth as a result of the 18- 24 hour incubation of the samples in an oven at 37 °C for 30, 45 and 60 minutes. It has been determined that the system, which was designed and tested with spore-forming bacteria, has a sterilization rate of 100% and there is no growth on the materials. As a result, compared to the systems that sterilize medical equipment with gas; A system that does not need additional consumables, does not harm equipment, is environmentally friendly, consumes less energy, can be sterilized faster and at lower temperatures with YOMA has been developed and its effectiveness, effect and reliability have been tested.
Internet-of-Things technology (IoTs) have accelerated biosensor applications in all fields. Loop-mediated isothermal amplification (LAMP)-based biosensor technologies in conjunction with smartphone detection have been adequate to cover the demands of mobile diagnostics. The ease of use, affordability, portability, high sensitivity, flexibility, and specificity demands of point-of-care detection can be achieved by low-cost electronic components, 3-dimensional printing technologies, capturing images of calorimetrically detected readouts made our system a promising approach for real-time point-of-detection in the field. In this study, we implemented a cloud service to our LAMP-based biosensor. We previously performed bacteria detection using colony-based LAMP device and now distributed the optical readouts of the assay using smartphones. We transferred the obtained image and results of the assays through cloud. Our user-friendly interface simplifies the data processing, it directly digitized the readouts and eliminates the need of data interpretation.
With 2.3 million women diagnosed with breast cancer and 685.000 deaths globally in 2020, breast cancer is the most prevalent form of cancer worldwide. However, 7.8 million women are alive after being diagnosed with breast cancer in the past five years, indicating that breast cancer is survivable if diagnosed early. One in eight women will have invasive breast cancer during her lifetime. Therefore, it is vital that regular screenings are conducted for early detection and, thus, survival. Previous studies have been completed on the Breast Cancer Coimbra Data Set from the UCI repository to identify biomarkers with reliable confidence intervals for further investigation with less emphasis on a perfect model due to lack of enough data. Here, Deep Feature Synthesis and Conditional Generative Adversarial Networks (CTGANs) are implemented for data generation and to increase the performance of the models in terms of specificity, accuracy, and sensitivity. Logistic Regression outperforms the other classifiers with 100% specificity, accuracy, and sensitivity when only Deep Feature Synthesis is used. Performance decrease of models, when introduced to data CTGANs-generated data, indicates that CTGANs do not generate similar enough data. However, the large range of variability in the model's performance shows that BMI, Resistin, Age, and Glucose are not trustable biomarkers for early-stage breast cancer.
Wound is formed by the disruption of the anatomical and functional integrity of living tissue and its healing is a dynamic process. The skin has the ability to heal itself, but the formation of bacterial infections during this process delays healing time or sometimes results in unsuccessful healing. Wound dressings offer promising treatments to prevent the formation of bacterial infection by providing antibacterial activity in the wound area. Similar to wound dressings, light applications at certain wavelengths and power densities accelerate the healing process and may have similar antibacterial activity. In the presence of a photosensitizer, the antibacterial activity of light applications may become more effective. In this study, PLGA-based wound dressings (WDs) were produced with chlorin e6 and abietic acid by a thin film casting method to be concomitantly applied with light applications in future studies. SEM analyses, swelling, and degradation tests were performed for the characterization of PLGA-based wound dressings. All WDs were created with smooth surfaces that degrade at the same rate and some WDs reach maximum swelling after 24 hours. Hence, the production of thin film WDs that can be combined with antibacterial treatment modalities is important and necessary for the elimination of infection and improvement of wound healing.
