Background: ALS is a progressive neuro-muscular disease, which is characterized by motor neuron loss in the Central Nervous System (CNS) and Peripheral Nervous System (PNS). Up to now, no accurate clinical method for diagnosis of the disease have been provided. In most cases, ALS patients are unable to walk normally due to abnormalities in the nervous system. For this reason, one of the most appropriate methods in the diagnosis of ALS from other neurological diseases or from healthy volunteers is the gait motor signal analysis.

Materials and Methods: In this study, gait signals available in Physionet database have been used. The database consists of 13 patients with ALS (ALS1, ALS2, …, ALS13) and 16 normal subjects (CO1, CO2, …, CO16). The patients participating in this study had no history of any psychiatric disorders and did not use any assistive device for walking, like wheelchair. The power spectrum of stride, swing, and stance of normal subjects and patients was computed for both left and right legs. To provide appropriate inputs for the classifier, the frequency band of the power spectrum of all signals was divided into eight equal parts. The area of all regions was computed. Three frequency band of the lower range of power spectra selected as inputs of the classifier.

Results: In this study, power spectra, as frequency attributes, were used to explore probable differences of time series in both patients and healthy subjects.

Conclusion: Artificial Neural Network was used to classify normal and ALS groups with the accuracy of 83% for the test data set. It seems that the present algorithm can be used in discriminating patients from normal subjects in the early stages of the disease.

Background: Individual differences, especially gender, have an important role on individuals responds to the emotions. In cognitive science investigations, the analysis of biological signals has been introduced as a confident way to evaluate such responses. In this paper, by adopting a comprehensive approach on biomedical signal processing techniques, a precise examination on women and men differences in affective responses has been provided into different emotional stimuli, including fear, sadness, happiness, and peacefulness.

Materials and Methods: Accordingly, signal processing methods were divided into three general categories, linear, wavelet, and non-linear based techniques. In the proposed method, different features from each of three categories and from three autonomic signals, including electrocardiogram (ECG), finger pulse, and galvanic skin response (GSR), were extracted. To induce emotions in participants, validated emotional pieces of music were broadcast in four affective classes.

Results: The results indicate the different patterns of responses into affective incentives in women and men. The differences were more noticeable in the features of pulse signal than those of the other signals. Among emotional classes, fear resulted in the highest rate of distinction between men and women emotional responses.

Conclusion: By the comprehensive evaluation of autonomic signals and different signal processing techniques, this study has tried to offer a new insight for better understanding of gender differences in emotional responses. In addition, it will help the researchers to adopt appropriate decisions in identifying efficient processing approach to deal with large amount of information achieved from signal analysis.