Analyzing how a newborn reacts to different stimuli is crucial to diagnose possible neurological conditions. In this sense, a key factor that expert clinicians assess in newborns is the presence of crying under nociceptive stimuli, since the absence of this factor under the stimulation procedure increases the chances of developing neonatal encephalopathy. To this end, this work tackles this problem by using Machine Learning techniques to detect infant cry events from audios recorded during nociceptive stimulation of 63 newborns at the “Hospital Universitario de Burgos”. Particularly, this work analyzes the performance of several well-known machine learning models (Multi-Layer Perceptron, Support Vector Machines and Long-Short Term Memory network) after preprocessing the audios with two feature extraction techniques (Mel Frequency Cepstral Coefficients and Linear Predictive Coding). Promising results are obtained in the analysis carried out achieving a F1-score of 0.93, a result that shows the potential of this approach in improving pediatric diagnosis, neonatal care, and the diagnose of other health problems through cry analysis.