Deep-learning enabled traumatic brain injury analysis

Primary supervisor

Lan Du

Co-supervisors

  • Joanna Dipnall

Traumatic brain injury (TBI) is an injury to the brain caused by an external force from incidents such as motor vehicle crashes, falls, assault or sports collisions. Almost seventy million individuals globally are estimated to suffer from TBI per annum [1], deeming it a major public health concern which is estimated to cost the global economy approximately $US400 billion annually [2]. Early identification of severe TBI with proper assessment and treatment lowers the risk of secondary injury and subsequent long-term disability and subsequent costs. Missed diagnoses can lead to severe complications, consequences and increased cost.

Deep learning approaches on CT have been gaining popularity. Pretrained convolutional neural networks have been recently used to detect COVID-19 [3] and hybrid methods are evolving that combine template matching, artificial neural networks and active contours for segmentation of significant anatomical landmarks and estimation of haematoma volume on brain CT scans [5]. Improving deep learning algorithms to accurately identify and classify head CT scan abnormalities have been identified as requiring urgent attention [4], opening up the possibility to translate into automating the triage process.


 

Student cohort

Double Semester

Aim/outline

The aim of this research project is to study how deep learning algorithm can advance the study of TBI, and develop SOTA deep learning algorithm designed for the TBI analysis. We are interested in the technologies that are capable of exploiting different modes of the collected data, which include images and text associated with each patient. Meanwhile, the interpretability of the results is as important as the prediction accuracy. 

The candidate will work closely with Dr. Lan Du from the Faculty of IT and Dr. Joanna F Dipnall from the School of Public Health and Preventive Medicine.

URLs/references

  1. Dewan MC, Rattani A, Gupta S, Baticulon RE, Hung Y-C, Punchak M, et al. Estimating the global incidence of traumatic brain injury. Journal of neurosurgery. 2018;130(4):1080-97.
  2. Maas AI, Menon DK, Adelson PD, Andelic N, Bell MJ, Belli A, et al. Traumatic brain injury: integrated approaches to improve prevention, clinical care, and research. The Lancet Neurology. 2017;16(12):987-1048.
  3. Pham TD. A comprehensive study on classification of COVID-19 on computed tomography with pretrained convolutional neural networks. Scientific RepoRtS. 2020;10(1):1-8.
  4. Chilamkurthy S, Ghosh R, Tanamala S, Biviji M, Campeau NG, Venugopal VK, et al. Deep learning algorithms for detection of critical findings in head CT scans: a retrospective study. The Lancet. 2018;392(10162):2388-96.
  5. Qureshi AN. Semi-automated classification of CT scans in traumatic brain injury patients. Int J Comput Appl. 2015;113:1-8.

Required knowledge

  • the candidate should have basic knowledge about machine/deep learning, e.g., various classification models.
  • the candidate should have good python programming skill and is familiar with either Pytorch or TensorFlow.
  • Knowing advanced deep learning models for NLP and CV is a plus.