A Review of Automated Pain Assessment in Infants: Features, Classification Tasks, and Databases

Abstract

Bedside caregivers assess infants' pain at constant intervals by observing specific behavioral and physiological signs of pain. This standard has two main limitations. The first limitation is the intermittent assessment of pain, which might lead to missing pain when the infants are left unattended. Second, it is inconsistent since it depends on the observer's subjective judgment and differs between observers. Intermittent and inconsistent assessment can induce poor treatment and, therefore, cause serious long-term consequences. To mitigate these limitations, the current standard can be augmented by an automated system that monitors infants continuously and provides quantitative and consistent assessment of pain. Several automated methods have been introduced to assess infants' pain automatically based on analysis of behavioral or physiological pain indicators. This paper comprehensively reviews the automated approaches (i.e., approaches to feature extraction) for analyzing infants' pain and the current efforts in automatic pain recognition. In addition, it reviews the databases available to the research community and discusses the current limitations of the automated pain assessment.

Multi-Channel Neural Network for Assessing Neonatal Pain from Videos

“…Neonatal face detection in the NICU is a challenging task due to several external factors such as illumination variations and partial occlusions (e.g., tapes or pacifier). As discussed in [10], several facial landmark trackers faced difficulty when applied to detect and track the face of infants.…”

“…Existing automated methods for assessing pain of neonates from facial expression are broadly classified into handcrafted methods and deep learning methods [10].…”

Can Pre-Trained Convolutional Neural Networks be used as Feature Extractors for Video-based Neonatal Sleep and Wake Classification?

“…In contrast, video-based methods appear to be a promising approach, since it is more comfortable and convenient to use both in the home or in the hospitals [19] [20]. With the advancements in deep learning algorithms and clinical research on neonatal facial patterns [21] [22] a new, unobtrusive approach of monitoring sleep patterns has been proposed [23] [24]. However, evaluation of the deep learning models demand big database to train the prediction model.…”

Visual Perception of Pain in Neonatal Face Images

“…Neste contexto, diversos trabalhos têm sido desenvolvidos nosúltimos anos para automatizar a avaliação da dor do RN [1] [2] [3]. Em particular, a metodologia desenvolvida em [2] mostrou-se bem sucedida no que se refereà possibilidade de automatizar a avaliação da dor do RN por meio da captura de imagens de face, da decodificação das mesmas e da classificação do resultado, como dor presente ou ausente em intervalo de tempo quase instantâneo.…”

“…Entretanto, essa metodologia [2] e as demais existentes [1] [3], no estágio em que se encontram, não trazem informações a respeito da relevância das características intrínsecas da face do RN que definem a dor usadas pelo profissional de saúde para interpretação deste fenômeno, sendo capazes de identificar a dor como presente somente se o recém-nascido expressar-se caracteristicamente por meio de expressões faciais específicas e pré-determinadas.…”