Abstract

Hip fractures are a major cause of death and disability for older people and are one of the costliest treatments for the NHS (£1billion annually). Given the high mortality rates ( 9% within 30 days; 30% within 1 year), any improvement in classification, and hence treatment, will have significant benefits. Hip fractures are classified using a standard classification system. However, there is wide varia- tion between hospitals regarding who performs the classification. As classification influences the chosen treatment, this can affect patient outcomes. We are develop- ing a machine learning based method to automatically classify hip fractures using X-rays, with the eventual aim of standardising classification across the NHS. There are two stages in our proposed classification process: automatically recognise the hip joints in the X-ray image; and then classify the hip fracture given this selected region. Using a GPU and caffe, we have trained a fully convolutional network to automatically locate the hip joint, scoring an intersection-over-union above 0.8 (>0.5 is correct) for 93% of the test set. We are currently training a convolutional neural network to automatically classify the type of fracture, us- ing a dataset of X-ray images that have been hand-labelled by a panel of experts. Details of our methodology and results of our study will be given.
Original languageEnglish
Publication statusPublished - 2018
EventHPC Symposium 2018 - University of Bath, Bath, UK United Kingdom
Duration: 6 Jun 20186 Jun 2018

Conference

ConferenceHPC Symposium 2018
CountryUK United Kingdom
CityBath
Period6/06/186/06/18

Fingerprint

X rays
Learning systems
Neural networks
Graphics processing unit

Cite this

Murphy, E., Gregson, C., Von Arx, O., Whitehouse, M., Budd, C., & Gill, H. (2018). ARCHi: Automatic Recognition and Classification of Hip Fractures. Abstract from HPC Symposium 2018, Bath, UK United Kingdom.

ARCHi: Automatic Recognition and Classification of Hip Fractures. / Murphy, Ellen; Gregson, Celia; Von Arx, Otto; Whitehouse, Michael; Budd, Christopher; Gill, Harinderjit.

2018. Abstract from HPC Symposium 2018, Bath, UK United Kingdom.

Research output: Contribution to conferenceAbstract

Murphy, E, Gregson, C, Von Arx, O, Whitehouse, M, Budd, C & Gill, H 2018, 'ARCHi: Automatic Recognition and Classification of Hip Fractures' HPC Symposium 2018, Bath, UK United Kingdom, 6/06/18 - 6/06/18, .
Murphy E, Gregson C, Von Arx O, Whitehouse M, Budd C, Gill H. ARCHi: Automatic Recognition and Classification of Hip Fractures. 2018. Abstract from HPC Symposium 2018, Bath, UK United Kingdom.
Murphy, Ellen ; Gregson, Celia ; Von Arx, Otto ; Whitehouse, Michael ; Budd, Christopher ; Gill, Harinderjit. / ARCHi: Automatic Recognition and Classification of Hip Fractures. Abstract from HPC Symposium 2018, Bath, UK United Kingdom.
@conference{8bf2034df2db4ee7a6a42c97c5dd078b,
title = "ARCHi: Automatic Recognition and Classification of Hip Fractures",
abstract = "Hip fractures are a major cause of death and disability for older people and are one of the costliest treatments for the NHS (£1billion annually). Given the high mortality rates ( 9{\%} within 30 days; 30{\%} within 1 year), any improvement in classification, and hence treatment, will have significant benefits. Hip fractures are classified using a standard classification system. However, there is wide varia- tion between hospitals regarding who performs the classification. As classification influences the chosen treatment, this can affect patient outcomes. We are develop- ing a machine learning based method to automatically classify hip fractures using X-rays, with the eventual aim of standardising classification across the NHS. There are two stages in our proposed classification process: automatically recognise the hip joints in the X-ray image; and then classify the hip fracture given this selected region. Using a GPU and caffe, we have trained a fully convolutional network to automatically locate the hip joint, scoring an intersection-over-union above 0.8 (>0.5 is correct) for 93{\%} of the test set. We are currently training a convolutional neural network to automatically classify the type of fracture, us- ing a dataset of X-ray images that have been hand-labelled by a panel of experts. Details of our methodology and results of our study will be given.",
author = "Ellen Murphy and Celia Gregson and {Von Arx}, Otto and Michael Whitehouse and Christopher Budd and Harinderjit Gill",
year = "2018",
language = "English",
note = "HPC Symposium 2018 ; Conference date: 06-06-2018 Through 06-06-2018",

}

TY - CONF

T1 - ARCHi: Automatic Recognition and Classification of Hip Fractures

AU - Murphy, Ellen

AU - Gregson, Celia

AU - Von Arx, Otto

AU - Whitehouse, Michael

AU - Budd, Christopher

AU - Gill, Harinderjit

PY - 2018

Y1 - 2018

N2 - Hip fractures are a major cause of death and disability for older people and are one of the costliest treatments for the NHS (£1billion annually). Given the high mortality rates ( 9% within 30 days; 30% within 1 year), any improvement in classification, and hence treatment, will have significant benefits. Hip fractures are classified using a standard classification system. However, there is wide varia- tion between hospitals regarding who performs the classification. As classification influences the chosen treatment, this can affect patient outcomes. We are develop- ing a machine learning based method to automatically classify hip fractures using X-rays, with the eventual aim of standardising classification across the NHS. There are two stages in our proposed classification process: automatically recognise the hip joints in the X-ray image; and then classify the hip fracture given this selected region. Using a GPU and caffe, we have trained a fully convolutional network to automatically locate the hip joint, scoring an intersection-over-union above 0.8 (>0.5 is correct) for 93% of the test set. We are currently training a convolutional neural network to automatically classify the type of fracture, us- ing a dataset of X-ray images that have been hand-labelled by a panel of experts. Details of our methodology and results of our study will be given.

AB - Hip fractures are a major cause of death and disability for older people and are one of the costliest treatments for the NHS (£1billion annually). Given the high mortality rates ( 9% within 30 days; 30% within 1 year), any improvement in classification, and hence treatment, will have significant benefits. Hip fractures are classified using a standard classification system. However, there is wide varia- tion between hospitals regarding who performs the classification. As classification influences the chosen treatment, this can affect patient outcomes. We are develop- ing a machine learning based method to automatically classify hip fractures using X-rays, with the eventual aim of standardising classification across the NHS. There are two stages in our proposed classification process: automatically recognise the hip joints in the X-ray image; and then classify the hip fracture given this selected region. Using a GPU and caffe, we have trained a fully convolutional network to automatically locate the hip joint, scoring an intersection-over-union above 0.8 (>0.5 is correct) for 93% of the test set. We are currently training a convolutional neural network to automatically classify the type of fracture, us- ing a dataset of X-ray images that have been hand-labelled by a panel of experts. Details of our methodology and results of our study will be given.

M3 - Abstract

ER -