Journal of Medical Engineering and Technology最新文献

Objective: Diabetic Peripheral Neuropathy (DPN) is the most common prolonged complication of diabetes. A nerve reaches to the hands, legs, and a foot is damaged due to excessive glucose level. This leads to the loss of sensation, numbness and pain in the feet, legs or hands. Currently available devices are expensive, take more time and need more expertise to operate them to detect the level of DPN. This study is designed to detect the level of diabetic peripheral neuropathy (DPN) from first joint of index finger using a novel 128-Hz electronic tuning fork prototype which is capable of performing accurate vibration perception duration (VPD).

Methods: A total of 169 diabetic patients were recruited from the secondary author's practice for assessment of level of DPN with our device. All the patients were enrolled according to an approved protocol. Patient places index finger on the tip of our device in such a way that the tip covers the first joint of index finger. Our device then provides the vibration of desired frequency and voltage to the index finger via tactile platform and patient starts feeling the vibration. Depending on the vibration perception duration (VPD) for which the patient feels the vibration, 4 levels of DPN i.e. Normal, Mild, Moderate and Severe are calculated. Three repeated measurements were taken from all 169 patients.

Results: Our device detected 74 DPN patients (6 severe, 26 moderates, 42 mild) and 89 normal (no DPN) patients. The mean of vibration perception duration (VPD) was 6.8 s, with a standard deviation (SD) of ± 0.84 s of all 169 patients. Mean VPD of severe, moderate, mild and normal level of DPN patients was 1.73 (mean SD = 0.7 s), 5.82 (mean SD = 0.84 s), 8.32 (mean SD = 1 s) and 11.3 s (mean SD = 0.84 s), respectively. Considering the Biothesiometer as the reference standard, our results were compared against it and our device's result accuracy was > 92%.

Conclusion: VPD was a sensitive measure of a detection of level of DPN. The device is compact, handy, easy to use and takes only few seconds to diagnose the level of DPN level in diabetic patients.

It further zeroes in on the forecasting of the effects of yoga on CVI with the aid of a broad dataset including demographic background, basic case severities, and yoga practice details. Through careful feature engineering, the machine learning algorithms foresee such eventualities as the changes in the symptom severity and overall improvements in well-being. This predictive model has the potential to transform personalised treatment approaches in CVI by providing specific yoga practice recommendations, optimising therapeutic methods, and enhancing the effective utilisation of health resources. It is also emphasised that ethical considerations, patient preferences, and safety issues are of utmost importance and must be ensured in any responsible clinical implementation. Integrating MLPC with optimisation systems holds great promise as a novel approach. This integration is likely to provide a befitting platform for the customised management of CVI and give essential insights for ongoing and future healthcare service practices. Certainly, results across VCSS-PRE and VCSS-1 revealed remarkable performance that the MLPC+MGO model achieved in prediction and classification. The results depict that this model ensured impressive levels of both Accuracy and Precision through all the layers of the MLPC. On that account, the first layer obtained top results, with a result of 0.957 Accuracy and 0.961 Precision for VCSS-PRE, and even more at results of 0.971 Accuracy and 0.973 Precision for VCSS-1.

This study presents an algorithm for classifying individuals into four hypertension categories (healthy, prehypertension, Stage 1, and Stage 2) using indices computed from photoplethysmographic (PPG) and anthropometric data. The dataset includes 219 individuals (115 women, 104 men, ages 21-86), with resting PPG signals, body mass index (BMI), age, weight, height, and resting heart rate. Key features (PPGAI, Ab, and Ad indices) were computed from the PPG signal. After dimensionality reduction through stepwise linear regression, the most informative predictors of hypertensive stages were identified for model training. Seven machine learning models, including Support Vector Machine (SVM), K-Nearest Neighbours, Logistic Regression, Random Forest, Naive Bayes, Linear Discriminant Analysis, and Quadratic Discriminant Analysis, were evaluated using leave-one-out cross-validation and the most accurate one was selected for final classification. The Linear SVM showed the best performance, correctly classifying 71.3%, 67.1%, 38.2%, and 55% of healthy, prehypertensive, Stage 1, and Stage 2 subjects, respectively. However, in a preliminary screening scenario aimed at prompting clinical follow-up for positive cases, the algorithm flagged 76.5% of prehypertensive, 97.1% of Stage 1, and 100% of Stage 2 individuals as belonging to one of the three hypertensive categories. Nonetheless, additional training data are needed to improve the model's accuracy.

The most common cause of dementia, which includes significant cognitive impairment that interferes with day-to-day activities, is Alzheimer's Disease (AD). Deep learning techniques performed better on diagnostic tasks. However, current methods for detecting Alzheimer's disease lack effectiveness, resulting in inaccurate results. To overcome these challenges, a novel deep ensemble architecture for AD classification is proposed in this research. The proposed model involves key phases, including Preprocessing, Segmentation, Feature Extraction, and Classification. Initially, Median filtering is employed for preprocessing. Subsequently, an improved U-Net architecture is employed for segmentation, and then the features including Improved Shape Index Histogram (ISIH), Multi Binary Pattern (MBP), and Multi Texton are extracted from the segmented image. Then, an En-LeCILSTM is proposed, which combines the LeNet, CNN and improved LSTM models. Finally, the resultant output is obtained by averaging the intermediate output of each model, leading to improved detection accuracy. Finally, the proposed model's efficiency is assessed through various analyses, including classifier comparison, and performance metric evaluation. As a result, the En-LeCILSTM model scored a higher accuracy of 0.963 and an F-measure of 0.908, which surpasses the result of traditional methods. The outcomes demonstrate that the proposed model is notably more effective in detecting Alzheimer's disease.

Medical image fusion reduces the time required for medical diagnosis by creating a composite image from a set of images belonging to different modalities. This paper introduces a deep learning framework for medical image fusion, optimising the number of convolutional layers and selecting an appropriate activation function. The conducted experiments demonstrate that employing three convolution layers with a swish activation function for the intermediate layers is sufficient to extract the salient features of the input images. The tuned features are fused using element-wise fusion rules to prevent the loss of minute details crucial for medical images. The comprehensive fused image is then reconstructed from these features using another set of three convolutional layers. Experimental results confirm that the proposed methodology outperforms other conventional medical image fusion methods in terms of various metrics and the quality of the fused image.

Anterior glenoid reconstruction using bone blocks is increasingly recognised as treatment option after critical bone loss. In this study, a biomechanical test setup is used to assess micromotion after bone block augmentation at the glenoid, comparing bone block augmentation with a spina-scapula block to the standard coracoid bone block (Latarjet). Twenty-four synthetic shoulder specimens were tested. Two surgical techniques (coracoid and spina-scapula bone block augmentation) were used on two different types of synthetic bone (Synbone and Sawbone). The specimens were cyclically loaded according to the 'rocking horse' setup defined in ASTM F2028. A mediolateral force of 170 N was applied on the bone block and a complete test comprised 5000 cycles. The Micromotion between bone block and glenoid was measured using a 3D Digital Image Correlation system. The measured micromotion divided into irreversible and reversible displacement of the augmented block. Medial irreversible displacement was the dominant component of the micromotion. The spina-scapula bone block showed a significantly higher irreversible displacement in medial direction compared to the coracoid block, when aggregating both types of synthetic bone (spina: 1.00 ± 0.39 mm, coracoid: 0.56 ± 0.39 mm, p = 0.01). The dominant irreversible medial displacement can be interpreted as initial settling behaviour.

Respiratory diseases are increasingly prevalent worldwide, often leading to critical conditions that require mechanical ventilation for life support. Proper management of these cases demands that clinicians be highly trained to respond effectively to various ventilatory manoeuvres during the recovery process. In this context, training tools for medical staff in mechanical ventilation become essential. Countries with emerging economies, such as Colombia, frequently face technological and economic limitations that restrict access to advanced medical training resources. As a result, the development of physical and virtual patient simulators presents a viable solution, as they can be designed using accessible technologies to support training in low-resource settings. This study presents SimVep, a patient simulator designed to emulate the physiological behaviour of obstructive and restrictive pulmonary conditions. The primary objective of SimVep is to enhance clinician training in mechanical ventilation, enabling healthcare professionals to acquire critical skills and improve patient outcomes in real-world clinical environments.

Attention deficit hyperactivity disorder (ADHD) is one of the children's most common neurodevelopmental conditions. ADHD diagnosis is based on evaluating inattention, hyperactivity, and impulsivity symptoms that interfere with or reduce daily functioning. Although electroencephalography (EEG) tests are used for ADHD diagnosis, they are generally considered a complement to clinical evaluation. This paper proposes an approach to classify EEG records of children with ADHD and control cases. We identified and extracted relevant features from EEG signals of 47 children (22 diagnosed with ADHD and 25 controls) and evaluated machine learning techniques for classification. We used the 2-tone oddball paradigm to elicit the subjects' auditory event-related potentials (ERP), and we recorded EEG signals with a portable headset for approximately five minutes. In the feature extraction stage, we included measures from cognitive evoked potentials, frequency bands power, chaos quantification, and bispectral analysis, in addition to the age of the children and the number of high-pitched tones the children counted during the test. The SVM and Trees algorithms obtained the best performance for 86.36% accuracy and 95.45% sensitivity. These findings demonstrate the potential of portable EEG-based systems to complement standard clinical assessments, offering an objective, time-efficient, and accessible approach to support early ADHD diagnosis. Achieving high accuracy and sensitivity in classification is critical to reducing the risk of misdiagnosis and ensuring timely intervention, ultimately improving patient outcomes.