Total views : 230

GPU based Deep Learning to Detect Asphyxia in Neonates

Affiliations

  • Department of Computer Science and Engineering, R. V. College of Engineering, Mysore Road, Bangalore - 560059, Karnataka, India

Abstract


Throughout the years, some countries have not seen any reduction in the death rate of neonates. Neonate refers to a baby within its first four weeks of life. Birth asphyxia is one of the three noteworthy reasons for neonatal deaths globally. A birth injury is demonstrative of some kind of mistake that changed an ordinary delivery into a traumatic ordeal for the newborn child (and mother). Perinatal asphyxia, or neonatal asphyxia, is a birth injury where a child doesn’t inhale regularly before, during or after birth. Asphyxia is a condition that depicts a diminished or ceased level of oxygen, and the perinatal stage is the period before, during or immediately after birth. At the point when an infant has not been breathing appropriately, there is danger of cerebrum harm and acidosis (a condition when a lot of acid builds up in the blood) which may result in the death of the newborn child if undiscovered or analyzed late. Our project uses machine learning in building up a minimal effort symptomatic arrangement. This paper has composed a machine-based example framework that identifies designs in the voices of known suffocating babies (and typical newborn children) while crying. It then utilizes the created model to predict if the newborn is affected by asphyxia or not. An accuracy of 92% was achieved. It will serve as a valuable apparatus in diminishing death rate everywhere throughout the world if accuracy can be improved.

Keywords

Asphyxia, Cry, DIGITS, Machine Learning, Neural Network, Newborn.

Full Text:

 |  (PDF views: 267)

References


  • Bax MC, Flodmark O, Tydeman C. Definition and classification of cerebral palsy from syndrome toward disease. Dev Med Child Neurol Suppl. 2007; 109:39–41.
  • United Nations. We can end poverty: millennium development goals and beyond 2015. Fact Sheet; 2013.
  • Opportunities for Africa’s newborns: Practical data, policy and programmatic support for newborn care in Africa. Lawn J, Kerber K, editors. PMNCH. Cape Town; 2006.
  • Reyes-Galaviz OF, Reyes-Garcia CA. A system for the processing of infant cry to recognize pathologies in recently born babies with neural networks. SPECOM 2004, 9th Conference, Speech and Computer; St Petersburg, Russia. 2004 Sep 20-22.
  • Chakrabartty S, Singh G, Cauwenberghs G. Hybrid support vector machine I hidden markov model approach for continuous speech recognition. IEEE Midwest Symp on Circuits and Systems; Lansing MI. 2001 Aug 8-11.
  • Mercer J. Functions of positive and negative type and their connection with the theory of integral equations. Philosophical Transactions of the Royal Society A. 1909; 209(441-458):415–46. DOI: I 0.1 098/rsta.1909.0016.
  • DIGITS: The deep learning GPU training system, Luke Yeager. CA, USA: NVIDIA Corporation; 2015.
  • Onu CC. Harnessing infant cry for swift, cost-effective diagnosis of perinatal asphyxia in low-resource settings. IEEE Canada International Humanitarian Technology Conference (IHTC); 2014.
  • Moharir M, Johar D, Amulya BS, Srivastava A. An experimental review on Intel® DPDK L2 forwarding. IEEE Sponsored 3rd International Conference on Electronics and Communication Systems (ICECS 2016); Karpagam College of Engineering. 2016 Feb 25-26.

Refbacks

  • There are currently no refbacks.


Creative Commons License
This work is licensed under a Creative Commons Attribution 3.0 License.