MRI-based diagnosis of brain tumours using a deep neural network framework

Milan Acharya, Abeer Alsadoon, Shahd Al-Janabi, P. W.C. Prasad, Ahmed Dawoud, Ghossoon Alsadoon, Manoranjan Paul

Research output: Book chapter/Published conference paperConference paperpeer-review

2 Citations (Scopus)

Abstract

The median survival time of patients with high grade glioma, a form of brain tumour, is 1-3 years. The current best practice adopts Convolutional Neural Network (CNN) for image classification and tumour detection. This method provides a significant improvement in brain tumour segmentation of Magnetic Resonance Imaging (MRI) images in comparison to other frameworks, but it is nonetheless slow and lacks precision. We sought to build upon the current best practice model by utilising a Deep Neural Network (DNN) model, which entailed modification of the segmentation and feature-extraction stages in order to improve the accuracy of those stages and the resulting segmentation. We contrasted the accuracy and efficiency of our model to the current best practice model using 10 brain tumour patient MRI datasets. First, the segmentation accuracy of our proposed model (M= 90%) outperformed that of the current best practice (M=78%). Second, the tumour detection processing time of our proposed model (M=34 ms) also outperformed that of the current best practice (M=73 ms). We, therefore, replicated previous studies by showing that automatic segmentation can aid in brain tumour detection. Importantly, we extended previous studies by proposing a model that classifies a brain tumour with greater accuracy and within lower processing times. Validation of the model with a larger dataset is recommended.

Original languageEnglish
Title of host publication2020 5th International Conference on Innovative Technologies in Intelligent Systems and Industrial Applications (CITISIA)
Place of PublicationUnited States
PublisherIEEE
Pages1-5
Number of pages5
ISBN (Electronic)9781728194370
DOIs
Publication statusPublished - 25 Nov 2020
EventCITISIA 2020 Conference: Conference on Innovative Technologies in Intelligent System & Industrial Applications - Charles Sturt University Study Centre, Sydney Campus, Sydney, Australia
Duration: 25 Nov 202027 Nov 2020
https://web.archive.org/web/20201128085551/https://ieee-citisia.org/ (Conference website)
https://web.archive.org/web/20210124015105/https://ieee-citisia.org/wp-content/uploads/2020/11/Conference-Program-new1.pdf (Conference program)
https://ieeexplore.ieee.org/xpl/conhome/9371766/proceeding?pageNumber=4 (Full paper proceedings)

Publication series

NameCITISIA 2020 - IEEE Conference on Innovative Technologies in Intelligent Systems and Industrial Applications, Proceedings

Conference

ConferenceCITISIA 2020 Conference
Country/TerritoryAustralia
CitySydney
Period25/11/2027/11/20
OtherThe “Conference on Innovative Technologies in Intelligent Systems & Industrial Applications” (CITISIA) is a student conference that aims to provide students of higher learning institutions with a platform for presenting their own projects. It is also a measure of recognition of students’ professional and technical achievements – by industries and international organizations such as IEEE. This conference is designed to facilitate exchanges of ideas through communication, networking and learning from others, for students and IEEE Chapters in terms of greater collaboration.
The conference provides a unique platform for students and researchers to share their experience and views through their latest research and to promote research and development activities among students and researchers. CITISIA 2020 provides an international forum for those actively involved in research to report on latest innovations and developments, to summarize state-of-the-art works, and to share ideas and advances from all aspects of engineering, where advances play an increasing role in providing enriching experiences and improving the quality of lives.
Internet address

Fingerprint

Dive into the research topics of 'MRI-based diagnosis of brain tumours using a deep neural network framework'. Together they form a unique fingerprint.

Cite this