Image quality assessment for machine learning tasks using meta-reinforcement learning

被引：22

作者：

Saeed S.U. ^{[1
]}

Fu Y. ^{[1
,2
]}

Stavrinides V. ^{[3
,4
]}

Baum Z.M.C. ^{[1
]}

Yang Q. ^{[1
]}

Rusu M. ^{[5
]}

Fan R.E. ^{[6
]}

Sonn G.A. ^{[5
,6
]}

Noble J.A. ^{[7
]}

Barratt D.C. ^{[1
]}

Hu Y. ^{[1
,7
]}

机构：

[1] Centre for Medical Image Computing, Wellcome/EPSRC Centre for Interventional & Surgical Sciences, Department of Medical Physics & Biomedical Engineering, University College London, UK, London

[2] InstaDeep, UK, London

[3] Division of Surgery & Interventional Science, University College London, UK, London

[4] Department of Urology, University College Hospital NHS Foundation Trust, UK, London

[5] Department of Radiology, Stanford University, Stanford, CA

[6] Department of Urology, Stanford University, Stanford, CA

[7] Department of Engineering Science, University of Oxford, UK, Oxford

来源：

Medical Image Analysis | 2022年 / 78卷

基金：

英国科研创新办公室; 英国惠康基金; 加拿大自然科学与工程研究理事会; 英国医学研究理事会; 英国工程与自然科学研究理事会;

关键词：

Image quality assessment; Meta-learning; Meta-reinforcement learning; Task amenability;

D O I：

10.1016/j.media.2022.102427

中图分类号：

学科分类号：

摘要：

In this paper, we consider image quality assessment (IQA) as a measure of how images are amenable with respect to a given downstream task, or task amenability. When the task is performed using machine learning algorithms, such as a neural-network-based task predictor for image classification or segmentation, the performance of the task predictor provides an objective estimate of task amenability. In this work, we use an IQA controller to predict the task amenability which, itself being parameterised by neural networks, can be trained simultaneously with the task predictor. We further develop a meta-reinforcement learning framework to improve the adaptability for both IQA controllers and task predictors, such that they can be fine-tuned efficiently on new datasets or meta-tasks. We demonstrate the efficacy of the proposed task-specific, adaptable IQA approach, using two clinical applications for ultrasound-guided prostate intervention and pneumonia detection on X-ray images. © 2022

引用

共 62 条

[1]

Abdi A.H., Luong C., Tsang T., Allan G., Nouranian S., Jue J., Hawley D., Fleming S., Gin K., Swift J., Rohling R., Abolmaesumi P., Automatic quality assessment of echocardiograms using convolutional neural networks: feasibility on the apical four-chamber view, IEEE Trans. Med. Imag., 36, 6, pp. 1221-1230, (2017)

[2]

Baum Z., Bonmati E., Cristoni L., Walden A., Prados F., Kanber B., Barratt D., Hawkes D., Parker G., Wheeler-Kingshott C., Hu Y., Image quality assessment for closed-loop computer-assisted lung ultrasound, Medical Imaging 2021: Image-Guided Procedures, Robotic Interventions, and Modeling. International Society for Optics and Photonics, pp. 160-166, (2021)

[3]

Botvinick M., Ritter S., Wang J., Kurth-Nelson Z., Blundell C., Hassabis D., Reinforcement learning, fast and slow, Trends Cogn. Sci. (Regul. Ed.), 23, 5, pp. 408-422, (2019)

[4]

Brown A.M., Elbuluk O., Mertan F., Sankineni S., Margolis D.J., Wood B.J., Pinto P.A., Choyke P.L., Turkbey B., Recent advances in image-guided targeted prostate biopsy, Abdom. Imag., 40, 6, pp. 1788-1799, (2015)

[5]

Camps S., Houben T., Carneiro G., Edwards C., Antico M., Dunnhofer M., Martens E., Baeza J., Vanneste B., van Limbergen E., Et al., Automatic quality assessment of transperineal ultrasound images of the male pelvic region, using deep learning, Ultrasound Med. Biol., 46, 2, pp. 445-454, (2020)

[6]

Choong M.K., Logeswaran R., Bister M., Improving diagnostic quality of mr images through controlled lossy compression using spiht, J. Med. Syst., 30, 3, pp. 139-143, (2006)

[7]

Chow L., Paramesran R., Review of medical image quality assessment, Biomed. Signal Process. Control, 27, pp. 145-154, (2016)

[8]

Chow L., Rajagopal H., Comparison of difference mean opinion score (dmos) of magnetic resonance images with full-reference image quality assessment (fr-iqa), Image Processing, Image Analysis and Real-Time Imaging (IPARTI) Symposium, (2015)

[9]

Cubuk E., Zoph B., Mane D., Vasudevan V., Le Q., (2019)

[10]

Daly S.J., Visible differences predictor: an algorithm for the assessment of image fidelity, Human Vision, Visual Processing, and Digital Display III. Vol. 1666. International Society for Optics and Photonics, pp. 2-15, (1992)

← 1 2 3 4 5 6 7 →