https://www.ncbi.nlm.nih.gov/pubmed/30407743
Computer vs human: Deep learning versus perceptual training for the detection of neck of femur fractures.
Author information
- 1
- Radiology Department, Royal Melbourne Hospital, Melbourne, Victoria, Australia.
- 2
- School of Psychological Sciences, University of Melbourne, Melbourne, Victoria, Australia.
- 3
- Radiology Department, University of Melbourne, Melbourne, Victoria, Australia.
Abstract
INTRODUCTION:
To evaluate the accuracy of deep convolutional neural networks (DCNNs) for detecting neck of femur (NoF) fractures on radiographs, in comparison with perceptual training in medically-naïve individuals.
METHODS:
This study extends a previous study that conducted perceptual training in medically-naïve individuals for the detection of NoF fractures on a variety of dataset sizes. The same anteroposterior hip radiograph dataset was used to train two DCNNs (AlexNet and GoogLeNet) to detect NoF fractures. For direct comparison with perceptual training results, deep learning was completed across a variety of dataset sizes (200, 320 and 640 images) with images split into training (80%) and validation (20%). An additional 160 images were used as the final test set. Multiple pre-processing and augmentation techniques were utilised.
RESULTS:
AlexNet and GoogLeNet DCNNs NoF fracture detection accuracy increased with larger training dataset sizes and mildly with augmentation. Accuracy increased from 81.9% and 88.1% to 89.4% and 94.4% for AlexNet and GoogLeNet respectively. Similarly, the test accuracy for the perceptual training in top-performing medically-naïve individuals increased from 87.6% to 90.5% when trained on 640 images compared with 200 images.
CONCLUSIONS:
Single detection tasks in radiology are commonly used in DCNN research with their results often used to make broader claims about machine learning being able to perform as well as subspecialty radiologists. This study suggests that as impressive as recognising fractures is for a DCNN, similar learning can be achieved by top-performing medically-naïve humans with less than 1 hour of perceptual training.
© 2018 The Royal Australian and New Zealand College of Radiologists.
KEYWORDS:
X-rays; femoral neck fractures; learning; radiology; supervised machine learning
- PMID:
- 30407743
- DOI:
- 10.1111/1754-9485.12828
처음 배우는 사람 정형외과 의사와, 인공지능 Deep CNN 신경망 알고리즘.
눈으로 판단하고 진단하는 것은, 이제 앞으로 사람보다 신경망 인공지능이 더 잘할듯. 시간이 갈수록 더울 그럴 것 같음.
학습시간도 짧고,
더 잘 찾아내고,
비용도 덜들고,
더 빨리 찾고,
쉬지도 않고,
업데이트도 잘되고.
일이 많다고 컴플레인도 없고.....단점이 별로 없는 듯.
(전기료만 조금 공급하고 프로그램만 돌리면..)
2018 JMIRO _Computer and Human _ femur neck fracture detection.pdf
