METHODS: Cry signals from 2 different databases were utilized. First database contains 507 cry samples of normal (N), 340 cry samples of asphyxia (A), 879 cry samples of deaf (D), 350 cry samples of hungry (H) and 192 cry samples of pain (P). Second database contains 513 cry samples of jaundice (J), 531 samples of premature (Prem) and 45 samples of normal (N). Wavelet packet transform based energy and non-linear entropies (496 features), Linear Predictive Coding (LPC) based cepstral features (56 features), Mel-frequency Cepstral Coefficients (MFCCs) were extracted (16 features). The combined feature set consists of 568 features. To overcome the curse of dimensionality issue, improved binary dragonfly optimization algorithm (IBDFO) was proposed to select the most salient attributes or features. Finally, Extreme Learning Machine (ELM) kernel classifier was used to classify the different types of infant cry signals using all the features and highly informative features as well.
RESULTS: Several experiments of two-class and multi-class classification of cry signals were conducted. In binary or two-class experiments, maximum accuracy of 90.18% for H Vs P, 100% for A Vs N, 100% for D Vs N and 97.61% J Vs Prem was achieved using the features selected (only 204 features out of 568) by IBDFO. For the classification of multiple cry signals (multi-class problem), the selected features could differentiate between three classes (N, A & D) with the accuracy of 100% and seven classes with the accuracy of 97.62%.
CONCLUSION: The experimental results indicated that the proposed combination of feature extraction and selection method offers suitable classification accuracy and may be employed to detect the subtle changes in the cry signals.
MAIN BODY: We argue that broader consideration of lactation, incorporating evolutionary, comparative and anthropological aspects, could provide new insights into breastfeeding practices and problems, enhance research and ultimately help to develop novel approaches to improve initiation and maintenance. Our current focus on breastfeeding as a strategy to improve health outcomes must engage with the evolution of lactation as a flexible trait under selective pressure to maximise reproductive fitness. Poor understanding of the dynamic nature of breastfeeding may partly explain why some women are unwilling or unable to follow recommendations.
CONCLUSIONS: We identify three key implications for health professionals, researchers and policymakers. Firstly, breastfeeding is an adaptive process during which, as in other mammals, variability allows adaptation to ecological circumstances and reflects mothers' phenotypic variability. Since these factors vary within and between humans, the likelihood that a 'one size fits all' approach will be appropriate for all mother-infant dyads is counterintuitive; flexibility is expected. From an anthropological perspective, lactation is a period of tension between mother and offspring due to genetic 'conflicts of interest'. This may underlie common breastfeeding 'problems' including perceived milk insufficiency and problematic infant crying. Understanding this - and adopting a more flexible, individualised approach - may allow a more creative approach to solving these problems. Incorporating evolutionary concepts may enhance research investigating mother-infant signalling during breastfeeding; where possible, studies should be experimental to allow identification of causal effects and mechanisms. Finally, the importance of learned behaviour, social and cultural aspects of primate (especially human) lactation may partly explain why, in cultures where breastfeeding has lost cultural primacy, promotion starting in pregnancy may be ineffective. In such settings, educating children and young adults may be important to raise awareness and provide learning opportunities that may be essential in our species, as in other primates.