RESUMEN

Bioelectrical impedance techniques have been useful in various applications, including body composition analysis, impedance plethysmography, impedance cardiography, lung ventilation, perfusion, and tissue characterization. Electrical impedance methods have also been useful in characterizing different foods like meat, fruits, and beverages. However, the temperature of tissue samples can change their dielectric properties, affecting their impedance. This research investigated the effects of temperature on the impedance of various biological tissues over the frequency range of 10 Hz to 5 MHz. Freshly excised animal tissues (lamb, cow, chicken), fish, fruits, and plants were considered as biological samples. The samples were placed in a test cell and submerged in a water bath heated by a hot plate to vary the temperature. Impedance measurements were conducted using a bioimpedance spectrometer in 2 °C steps within the temperature range of 20 °C to 50 °C. Impedance values decreased with increased temperature across all measurement frequencies for all biological samples. Curve fitting indicated that impedance decreased linearly with temperature, with a mean correlation coefficient of 0.972 for all samples. For all biological samples under investigation, the relative impedance change ranged from -0.58% to -2.27% per °C, with a mean and standard deviation of (-1.42±0.34) %/°C. On average, animal samples exhibited a higher relative temperature coefficient of -1.56% per °C (±0.41) across the frequency range, compared to -1.31% per °C (±0.26) for fruit and vegetable samples. Additionally, the relative temperature coefficient values were generally higher at lower frequencies than at higher frequencies. The findings of this research can be valuable for studies or biomedical applications involving variable tissue temperatures.

RESUMEN

Every one of us has a unique manner of communicating to explore the world, and such communication helps to interpret life. Sign language is the popular language of communication for hearing and speech-disabled people. When a sign language user interacts with a non-sign language user, it becomes difficult for a signer to express themselves to another person. A sign language recognition system can help a signer to interpret the sign of a non-sign language user. This study presents a sign language recognition system that is capable of recognizing Arabic Sign Language from recorded RGB videos. To achieve this, two datasets were considered, such as (1) the raw dataset and (2) the face-hand region-based segmented dataset produced from the raw dataset. Moreover, operational layer-based multi-layer perceptron "SelfMLP" is proposed in this study to build CNN-LSTM-SelfMLP models for Arabic Sign Language recognition. MobileNetV2 and ResNet18-based CNN backbones and three SelfMLPs were used to construct six different models of CNN-LSTM-SelfMLP architecture for performance comparison of Arabic Sign Language recognition. This study examined the signer-independent mode to deal with real-time application circumstances. As a result, MobileNetV2-LSTM-SelfMLP on the segmented dataset achieved the best accuracy of 87.69% with 88.57% precision, 87.69% recall, 87.72% F1 score, and 99.75% specificity. Overall, face-hand region-based segmentation and SelfMLP-infused MobileNetV2-LSTM-SelfMLP surpassed the previous findings on Arabic Sign Language recognition by 10.970% accuracy.

Asunto(s)

RESUMEN

Accurate assessment of Respiratory Rate (RR) is the most important mechanism in detecting pneumonia in low-resource settings. Pneumonia is a disease with one of the highest mortality rates among young children under five. However, the diagnosis of pneumonia for infants remains challenging, especially in low- and middle-income countries (LMIC). In such situations, RR is most often measured manually with visual inspection. Accurate RR measurement requires the child to remain calm without any stress for a few minutes. The difficulty in achieving this with a sick child in a clinical environment can result in errors and misdiagnosis, even more so when the child is crying and non-cooperating around unfamiliar adults. Therefore, we propose an automated novel RR monitoring device built with textile glove and dry electrodes which can make use of the relaxed posture when the child is resting on the carer's lap. This portable system is non-invasive and made with affordable instrumentation integrated on customized textile glove. The glove has multi-modal automated RR detection mechanism that simultaneously uses bio-impedance and accelerometer data. This novel textile glove with dry electrodes can easily be worn by a parent/carer and is washable. The real-time display on a mobile app shows the raw data and the RR value, allowing a healthcare professional to monitor the results from afar. The prototype device has been tested on 10 volunteers with age variation of 3 years to 33 years, including male and female. The maximum variation of measured RR with the proposed system is ±2 compared to the traditional manual counting method. It does not create any discomfort for either the child or the carer and can be used up to 60 to 70 sessions/day before recharging.

RESUMEN

Harnessing the inherent anti-spoofing quality from electroencephalogram (EEG) signals has become a potential field of research in recent years. Although several studies have been conducted, still there are some vital challenges present in the deployment of EEG-based biometrics, which is stable and capable of handling the real-world scenario. One of the key challenges is the large signal variability of EEG when recorded on different days or sessions which impedes the performance of biometric systems significantly. To address this issue, a session invariant multimodal Self-organized Operational Neural Network (Self-ONN) based ensemble model combining EEG and keystroke dynamics is proposed in this paper. Our model is tested successfully on a large number of sessions (10 recording days) with many challenging noisy and variable environments for the identification and authentication tasks. In most of the previous studies, training and testing were performed either over a single recording session (same day) only or without ensuring appropriate splitting of the data on multiple recording days. Unlike those studies, in our work, we have rigorously split the data so that train and test sets do not share the data of the same recording day. The proposed multimodal Self-ONN based ensemble model has achieved identification accuracy of 98% in rigorous validation cases and outperformed the equivalent ensemble of deep CNN models. A novel Self-ONN Siamese network has also been proposed to measure the similarity of templates during the authentication task instead of the commonly used simple distance measure techniques. The multimodal Siamese network reduces the Equal Error Rate (EER) to 1.56% in rigorous authentication. The obtained results indicate that the proposed multimodal Self-ONN model can automatically extract session invariant unique non-linear features to identify and authenticate users with high accuracy.

Asunto(s)

RESUMEN

A real-time Bangla Sign Language interpreter can enable more than 200 k hearing and speech-impaired people to the mainstream workforce in Bangladesh. Bangla Sign Language (BdSL) recognition and detection is a challenging topic in computer vision and deep learning research because sign language recognition accuracy may vary on the skin tone, hand orientation, and background. This research has used deep machine learning models for accurate and reliable BdSL Alphabets and Numerals using two well-suited and robust datasets. The dataset prepared in this study comprises of the largest image database for BdSL Alphabets and Numerals in order to reduce inter-class similarity while dealing with diverse image data, which comprises various backgrounds and skin tones. The papers compared classification with and without background images to determine the best working model for BdSL Alphabets and Numerals interpretation. The CNN model trained with the images that had a background was found to be more effective than without background. The hand detection portion in the segmentation approach must be more accurate in the hand detection process to boost the overall accuracy in the sign recognition. It was found that ResNet18 performed best with 99.99% accuracy, precision, F1 score, sensitivity, and 100% specificity, which outperforms the works in the literature for BdSL Alphabets and Numerals recognition. This dataset is made publicly available for researchers to support and encourage further research on Bangla Sign Language Interpretation so that the hearing and speech-impaired individuals can benefit from this research.

Asunto(s)

RESUMEN

Probing deep regions of the lung using electrical impedance is very important considering the need for a low cost and simple technique, particularly for the low and medium income countries. Because of complexity and cost, Electrical Impedance Tomography is not suitable for this envisaged application. The simple Tetrapolar Impedance Measurement (TPIM) technique employing four electrodes is the age old technique for bioelectrical measurements. However, it has its limitations in respect of organ localisation and depth sensitivity using skin surface electrodes. Recently, a new 6-electrode TPIM with two current electrodes but two pairs of appropriately connected potential electrodes positioned on the front and back of the thorax, proposed by one of the authors, came with a promise. However, this work gave a qualitative proposal based on concepts of physics and lacked a quantitative evaluation. In order to evaluate the method quantitatively, the present work employed finite element method based COMSOL Multiphysics software and carried out simulation studies using this new 6-electrode TPIM and compared the results with those from 4-electrode TPIM, with electrodes applied either on the front or at the back of the thorax for the latter. Initially, it carried out a sensitivity distribution study using a simple rectangular volume conductor which showed that the 6-electrode TPIM gives better depth sensitivity throughout the lung region. Next it used a near life like thorax model developed by another of the authors earlier. Using this model, extensive studies were carried out to quantify the overall sensitivity over a target lung region, the contribution of the target lung to the total measured impedance, and several other parameters. Through these studies, the 6-electrode TPIM was established on a stronger footing for probing deep regions of the lungs.

RESUMEN

Objective.Pneumonia is the single largest cause of death in children worldwide due to infectious diseases. According to WHO guidelines, fast breathing and chest indrawing are the key indicators of pneumonia in children requiring antibiotic treatments. The aim of this study was to develop a video based novel method for simultaneous monitoring of respiratory rate and chest indrawing without upsetting babies.Approach.Respiratory signals, corresponding to periodic movements of chest-abdominal walls during breathing, were extracted by analyzing RGB (red, green, blue) components in video frames captured by a smartphone camera. Respiratory rate was then obtained by applying fast Fourier transform on the de-noised respiratory signal. Chest indrawing was detected by analysing relative phases of regional chest-abdominal wall mobility. The performance of the developed algorithm was evaluated on both healthy and pneumonia children.Main results.The proposed method can measure respiratory rate with an overall mean absolute error of 1.8 bpm in the range 18-105 bpm. Phase difference between regional chest wall movements in the chest indrawing (pneumonia) cases was found to be 143 ± 23.9 degrees, which was significantly higher than that in the healthy cases 52.3 ± 32.6 degrees (p< 0.001).Significance.Being non-intrusive and non-subjective, this computer-aided method can be useful in the monitoring for respiratory rate and chest indrawing for the diagnosis of pneumonia and its severity in children.

Asunto(s)

RESUMEN

Electrical impedance measurements of biological tissue have many potential applications and tetrapolar impedance measurement (TPIM) with four electrodes is traditionally used which eliminates high skin contact impedance. A linear array of four electrodes for TPIM on the horizontal plane of a cylindrical volume conductor of diameter D, where the length of the array is πD/2 with potential electrodes near the centre of the array, will give a high sensitivity near the surface which reduces rapidly with depth. A recently proposed six-electrode variation of TPIM uses an additional pair of potential electrodes on the opposite side of the volume conductor in the same horizontal plane around the circumference, with the expectation that the sensitivity of the deeper regions will thereby be enhanced. The present work carries out a finite element simulation (using COMSOL) and an experimental phantom study (saline phantom) to quantitatively evaluate the improvement obtained by this new method. The new configuration doubled the sensitivity at the central region, which was reasonably uniform over a wider zone, gradually increasing towards the potential electrodes on both sides. This would be useful for a range of biological studies of deep body organs such as lungs, stomach, and bladder. where the respective external body shapes may be approximated by an oval cylinder and where electrical impedance techniques have shown promise.

Asunto(s)

RESUMEN

For probing deep organs of the body using electrical impedance, the conventional method is to use Electrical Impedance Tomography (EIT). However, this would be a sophisticated machine and will be very expensive when a full 3D EIT is developed in the future. Furthermore, for most low income countries such expensive devices may not deliver the benefits to a large number of people. Therefore, this paper suggests the use of simpler techniques like Tetrapolar Impedance Measurement (TPIM) or Focused Impedance Method (FIM) in probing deeper organs. Following a method suggested earlier by one of the authors, this paper studies the possibility of using TPIM and FIM for the stomach. Using a simplified model of the human trunk with an embedded stomach, a finite element simulation package, COMSOL, was used to obtain transfer impedance values and percentage contribution of the stomach region in the total impedance. For this work, judicious placement of electrodes through qualitative visualizations based on point sensitivity equations and equipotential concepts were made, which showed that reasonable contribution of the stomach region is possible through the use of TPIM and FIM. The contributions were a little over 20% which is of similar order of the cross-sectional area percentage of the stomach with respect to that of the trunk. For the case where the conductivity of the stomach region was assumed about 4 times higher, the contributions increased to about 38%. Through further studies this proposed methods may contribute greatly in the study of deeper organs of the body.

RESUMEN

Tetra-polar electrical impedance measurement (TPIM) with a square geometry of electrodes is useful in the characterization of epithelial tissues, especially in the detection of cervical cancer at precancerous stages. However, in TPIM, the peak planar sensitivity just below the electrode surface is almost zero and increases to a peak value at a depth of about one third to one half of the electrode separation. To get high sensitivity for the epithelial layer, having thicknesses of 200 µm to 300 µm, the electrode separation needed is less than 1 mm, which is difficult to achieve in practical probes. This work proposes a conical conducting layer in front of a pencil like probe with a square geometry of TPIM electrodes to create virtual electrodes with much smaller separation at the body surface, thus increasing the sensitivity of the epithelial tissues. To understand the improvements, if any, 3D sensitivity distribution and transfer impedance were simulated using COMSOL Multiphysics software for a simplified body tissue model containing a 300 µm epithelial layer. It has been shown that fractional contribution of an epithelial layer can be increased several times placing a cylindrical conducting layer in between the tissue surface and the electrodes, which can further be enhanced using a conical conducting layer. The results presented in this paper can be used to choose an appropriate electrode separation, conducting layer height and cone parameters for enhanced sensitivity in the epithelial layer.