Deep learning-based quantitative analyses of spontaneous movements and their association with early neurological development in preterm infants

Hyun Iee Shin,Woo Hyung Lee,Yoo-Jin Kim,Don-Kyu Kim,Seul Gi Park,Seung Han Shin,Ee-Kyung Kim,Hye Min Ji,Moon Suk Bang,Hyung-Ik Shin,Eun Sun Lee

doi:10.1038/s41598-022-07139-x

Abstract

This study aimed to develop quantitative assessments of spontaneous movements in high-risk preterm infants based on a deep learning algorithm. Video images of spontaneous movements were recorded in very preterm infants at the term-equivalent age. The Hammersmith Infant Neurological Examination (HINE) was performed in infants at 4 months of corrected age. Joint positional data were extracted using a pretrained pose-estimation model. Complexity and similarity indices of joint angle and angular velocity in terms of sample entropy and Pearson correlation coefficient were compared between the infants with HINE < 60 and ≥ 60. Video images of spontaneous movements were recorded in 65 preterm infants at term-equivalent age. Complexity indices of joint angles and angular velocities differed between the infants with HINE < 60 and ≥ 60 and correlated positively with HINE scores in most of the joints at the upper and lower extremities (p < 0.05). Similarity indices between each joint angle or joint angular velocity did not differ between the two groups in most of the joints at the upper and lower extremities. Quantitative assessments of spontaneous movements in preterm infants are feasible using a deep learning algorithm and sample entropy. The results indicated that complexity indices of joint movements at both the upper and lower extremities can be potential candidates for detecting developmental outcomes in preterm infants.

Highlights

This study aimed to develop quantitative assessments of spontaneous movements in high-risk preterm infants based on a deep learning algorithm
The current study demonstrated that the complexity of term-equivalent spontaneous movements in preterm infants quantitatively analyzed using a deep learning algorithm is associated with early neurological development assessed by Hammersmith Infant Neurological Examination (HINE) at 4 months of corrected age[38,39]
The complexity indices of both joint angles and joint angular velocities were different between the very preterm infants with HINE < 60 and ≥ 60, and showed positive correlations with the HINE scores in most joints of the upper and lower extremities

Summary

Introduction

This study aimed to develop quantitative assessments of spontaneous movements in high-risk preterm infants based on a deep learning algorithm. The General Movement Assessment (GMA) for evaluating spontaneous infantile movements is the validated tool for the early detection of cerebral p alsy[5,8] along with neonatal magnetic resonance imaging[9] and the Hammersmith Infant Neurological Examination (HINE)[10]. The vision-based approaches using RGB cameras has been getting great interest with popularization of smartphones and recent advances in pose-estimation models based on deep learning algorithms This approach can have several advantages compared to methods using direct sensing modalities including relative easiness to understand, high spatial resolution, high context information, non-intrusiveness, not-dependence on reflective markers or inertial sensors, and high a vailability[17,18]

Objectives

Methods

Results

Conclusion