Lung cancer is recognised as the most severe disease that affects humans and frequently results in mortality when compared to other cancer conditions. Lung cancer cannot be detected early because it exhibits no symptoms. However, early identification of lung cancer contributes to people's continued survival rate. Computer technology has recently been employed to solve these diagnostic issues. In this research, we propose a hybrid deep-learning method for predicting lung cancer. An enhanced MobileNetV3 (EMobileNetV3) is proposed to predict the probability of lung cancer. The DenseNet-169 model is used to extract the features. An effective osprey optimisation algorithm (Os-OA) is also presented to adjust the proposed classification model's parameters to improve the classification performance. Compared to other existing models, the proposed model performed better, obtaining an accuracy of 98% and an AUC of 96% for dataset 1 and an accuracy of 99.30% and an AUC of 96.6% for dataset 2.

We focused on enhancing speech improvement algorithms by addressing the challenge of extracting high-quality LPC parameters from noisy speech. The Kalman filter is a widely used algorithm in speech enhancement, and we aim to improve it by modifying the power spectrum parameters through a new approach called the modified power spectrum method within the LPC model for the Kalman Filter algorithm (MPS-LPC-KF). We evaluated our method using the NOIZEUS corpus and found it outperformed other existing methods. We are excited to see that our research has the potential to advance speech enhancement algorithms and ultimately improve communication in noisy environments.

In response to the current challenges of low diagnostic accuracy for Alzheimer's disease (AD) and the weak ability of single-modal imaging to extract lesion feature information, a deep learning-based multimodal image fusion method for AD classification has been proposed. First, a novel residual network architecture is used to extract lesion features from three-dimensional images. Then, the improved residual network is employed as a feature extractor to separately extract image features from magnetic resonance imaging (MRI) and positron emission tomography (PET) scans. Afterward, the features from both modalities are fused and subsequently classified. Finally, to enable the fusion network to better capture the spatial relationships between dimensions and channels in three-dimensional medical images, a coordinate attention mechanism is introduced into the network structure. Experimental results show that the improved fusion network achieved an accuracy of 91.07% in the AD/mild cognitive impairment (MCI)/normal cognitive (CN) classification task. This represents a 7.14% improvement over the basic residual network, a 21.43% improvement over single-modal MRI-based methods, and a 12.5% improvement over single-modal PET-based methods. The improved fusion network demonstrates superior classification performance compared to the basic residual network in AD/CN, AD/MCI, and CN/MCI classification tasks, proving its effectiveness.

Although the dental X-rays are useful for detecting and treating oral health problems, the low-resolution images they produce often make it hard for dental professionals to see fine details. This limitation occasionally leads to diagnostic challenges and even results in missed problems. As a means of addressing this issue, our study explored the application of deep learning approaches to sharpen and improve the quality of dental X-ray pictures, making them clearer and easier to interpret. We applied several deep learning methods, known for their success in enhancing image quality, to a dataset of dental X-rays. The results show significant improvements in clarity, with higher image quality scores – measured by metrics like PSNR and SSIM – that indicate a more detailed view of dental structures. These improvements could help dental professionals to catch issues earlier and make more accurate diagnoses. Our research demonstrates the potential of deep learning to change dental X-ray imaging, supporting better outcomes for patients and providing a useful tool for dental care providers.

A serious mental illness, schizophrenia (SZ) affects 1% of people worldwide and is characterised by delusions, hallucinations and disorganised thought patterns. Diagnosis mostly is based on subjective interviews by a psychiatrist in which there is a high chance of human errors and bias. In this work, we have conducted a comprehensive analysis of electroencephalogram (EEG) data using empirical mode decomposition (EMD) algorithm which can analyse non-stationary and nonlinear signals and separates them into components at different resolutions called intrinsic mode functions (IMFs). In this work, our primary goal is to introduce a hybrid approach for IMF selection that combines four distinct parameters namely correlation, energy, statistical significance and power spectral density (PSD) distance. From the selected IMFs nine statistical features are computed and performance is evaluated using various classifiers. Among all the classifiers k-nearest neighbour (KNN) showed the best accuracy of 90.29% using the second IMF. These results suggest that EEG signals can effectively distinguish between healthy control and SZ patients and have a potential to help psychiatrists for diagnosis of SZ.

Despite the growing use of compression garments (CGs) in running, their effects on performance remain uncertain due to varied applications, diverse methodologies, and different types of CGs. This systematic review aims to synthesise existing literature, evaluating CGs' impact on running performance through biomechanical, physiological, perceptual, and fatigue-related outcomes. According to the PRISMA-P guideline for systematic reviews, a comprehensive search was conducted on the four online databases for the article published up to the December 2024, involving Google Scholar, Web of Science, PubMed, and SPORTD. The systematic review included 15 articles, with 11 focusing on professional runners and four on recreational runners, comprising 241 male and 45 female participants, of whom 157 males and 36 females were well-trained runners. This systematic review of 15 studies assessed the effects of CGs on running-related outcomes, including physiological responses, biomechanical factors, performance metrics, perceived sensations, and muscle fatigue. Physiological responses, examined in ten studies, showed inconsistent evidence of CGs benefits in long-distance or short-term running. Biomechanical factors and performance remained largely unchanged across CGs types, pressures, and runner experience levels. While CGs reduced delayed onset muscle soreness, these garments had limited impact on running fatigue.

Heat stress results in significant cardiovascular adjustments for temperature regulation through heat dissipation. The body employs autonomic and behavioural responses mediated by autonomic nervous system (ANS) to prevent hyperthermia. Therefore, the present study aimed to examine the ANS oscillations to significant heat stress stimuli. Digital lead-II electrocardiogram (ECG) and pulse plethysmogram (PPG) were recorded from control and heat-stressed groups under anesthetised conditions. Tachograms were generated from both ECG and PPG signals. The fast Fourier transform (FFT) was used to analyse the bands using Kubios 3.5.0 software. An unpaired t-test and Bland-Altman test were employed to compare the differences between the two groups. The results demonstrated a significant shift of oscillations toward sympathetic dominance with the withdrawal of parasympathetic activity under heat exposure. The LF oscillations increased and exhibited strong associations between HRV and PRV parameters under the heat stress group. PRV and HRV assess the ANS oscillations and aided with valuable insights into cardiac autonomic function.