Introduction: Drowsy driving is a significant factor causing dire road crashes and casualties around the world. Detecting it earlier and more effectively can significantly reduce the lethal aftereffects and increase road safety. As physiological conditions originate from the human brain, so neurophysiological signatures in drowsy and alert states may be investigated for this purpose. In this preface, A passive brain-computer interface (pBCI) scheme using multichannel electroencephalography (EEG) brain signals is developed for spatially localized and accurate detection of human drowsiness during driving tasks.Methods: This pBCI modality acquired electrophysiological patterns of 12 healthy subjects from the prefrontal (PFC), frontal (FC), and occipital cortices (OC) of the brain. Neurological states are recorded using six EEG channels spread over the right and left hemispheres in the PFC, FC, and OC of the sleep-deprived subjects during simulated driving tasks. In post-hoc analysis, spectral signatures of the δ, θ, α, and β rhythms are extracted in terms of spectral band powers and their ratios with a temporal correlation over the complete span of the experiment. Minimum redundancy maximum relevance, Chi-square, and ReliefF feature selection methods are used and aggregated with a Z-score based approach for global feature ranking. The extracted drowsiness attributes are classified using decision trees, discriminant analysis, logistic regression, naïve Bayes, support vector machines, k-nearest neighbors, and ensemble classifiers. The binary classification results are reported with confusion matrix-based performance assessment metrics.Results: In inter-classifier comparison, the optimized ensemble model achieved the best results of drowsiness classification with 85.6% accuracy and precision, 89.7% recall, 87.6% F₁-score, 80% specificity, 70.3% Matthews correlation coefficient, 70.2% Cohen’s kappa score, and 91% area under the receiver operating characteristic curve with 76-ms execution time. In inter-channel comparison, the best results were obtained at the F8 electrode position in the right FC of the brain. The significance of all the results was validated with a p-value of less than 0.05 using statistical hypothesis testing methods.Conclusions: The proposed scheme has achieved better results for driving drowsiness detection with the accomplishment of multiple objectives. The predictor importance approach has reduced the feature extraction cost and computational complexity is minimized with the use of conventional machine learning classifiers resulting in low-cost hardware and software requirements. The channel selection approach has spatially localized the most promising brain region for drowsiness detection with only a single EEG channel (F8) which reduces the physical intrusiveness in normal driving operation. This pBCI scheme has a good potential for practical applications requiring earlier, more accurate, and less disruptive drowsiness detection using the spectral information of EEG biosignals.

Human Behavior Recognition (HBR) or Human Action Recognition (HAR) is an important area of research having numerous applications in the field of computer vision and machine learning. In this article, we proposed a new method for HBR based on multi-types features fusion and irrelevant features reduction named MtFR. The Proposed MtFR approach initially selects a luminance channel and calculates motion estimation using optical flow. Afterwards, the moving regions are extracted through background subtraction approach. In the features extraction step, shape, color, and Gabor wavelet features are extracted and fused based on serial method. Thereafter, reduced irrelevant and redundant features are removed by Von Neuman entropy approach. The selected reduced features are finally recognized by One-Against-All (OAA) Multi-class SVM classifier. Extensive experiments are performed using three famous datasets such as Muhavi, WVU and YouTube, and achieved the recognition rate of 99.2%, 99.3%, and 100%, respectively.

Recently, medical imaging and machine learning gained significant attention in the early detection of brain tumor. Compound structure and tumor variations, such as change of size, make brain tumor segmentation and classification a challenging task. In this review, we survey existing work on brain tumor, their stages, survival rate of patients after each stage, and computerized diagnosis methods. We discuss existing image processing techniques with a special focus on preprocessing techniques and their importance for tumor enhancement, tumor segmentation, feature extraction and features reduction techniques. We also provide the corresponding mathematical modeling, classification, performance matrices, and finally important datasets. Last but not least, a detailed analysis of existing techniques is provided which is followed by future directions in this domain.

Malignant melanoma is considered as one of the most deadly cancers, which has broadly increased worldwide since the last decade. In 2018, around 91,270 cases of melanoma were reported and 9,320 people died in the US. However, diagnosis at the initial stage indicates a high survival rate. The conventional diagnostic methods are expensive, inconvenient and subject to the dermatologist’s expertise as well as a highly equipped environment. Recent achievements in computerized based systems are highly promising with improved accuracy and efficiency. Several measures such as irregularity, contrast stretching, change in origin, feature extraction and feature selection are considered for accurate melanoma detection and classification. Typically, digital dermoscopy comprises four fundamental image processing steps including preprocessing, segmentation, feature extraction and reduction, and lesion classification. Our survey is compared with the existing surveys in terms of preprocessing techniques (hair removal, contrast stretching) and their challenges, lesion segmentation methods, feature extraction methods with their challenges, features selection techniques, datasets for the validation of the digital system, classification methods and performance measure. Also, a brief summary of each step is presented in the tables. The challenges for each step are also described in detail, which clearly indicate why the digital systems are not performing well. Future directions are also given in this survey.