Advertisement

The Importance of Data Quality in the Nascent Algorithmic Age of Radiology

Published:March 26, 2022DOI:https://doi.org/10.1016/j.acra.2022.03.003
      A decade has passed since the widely celebrated victory of AlexNet, a convolutional neural network (CNN) at the 2012 ImageNet classification challenge, which demonstrated the wide accessibility of GPU-accelerated CNN implementations, and the remarkable potential of CNNs in accurate image classification (
      • Krizhevsky A
      • Sutskever I
      • Hinton GE.
      ImageNet classification with deep convolutional neural networks.
      ). This impressive performance and the apparent adaptability to medical image classification problems led to suggestions that expert-based professions relying on image classification would be imminently entirely supplanted by machines, with bold proposals to disband radiology residency programs. Following a period of in-depth analysis of the clinical value of AI in medical imaging and the numerous efforts in supplanting human radiologists with varying degrees of success and generalizability (
      • Roberts M
      • Driggs D
      • Thorpe M
      • et al.
      Common pitfalls and recommendations for using machine learning to detect and prognosticate for COVID-19 using chest radiographs and CT scans.
      ), a new more mature paradigm emerged. In this paradigm, humans, and AI are envisioned working together to redefine and expand the role of medical imaging to support unprecedented advancements in patient care (
      • Reardon S.
      Rise of robot radiologists.
      ). Indeed, the opportunities for fruitful interactions of human and machine seem boundless, including radiomics-supported “digital biopsy” (
      • Sahiner B
      • Pezeshk A
      • Hadjiiski LM
      • et al.
      Deep learning in medical imaging and radiation therapy.
      ), image-based treatment response prediction and plan selection (
      • Jin C
      • Yu H
      • Ke J
      • et al.
      Predicting treatment response from longitudinal images using multi-task deep learning.
      ), and automated image segmentation for improved on-the-fly visualization of patient anatomy to enable precise personalized interventions (
      • Isensee F
      • Jaeger PF
      • Kohl SAA
      • et al.
      nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation.
      ) among numerous others.

      Abbreviations:

      AI (Artificial Intelligence), CNN (Convolutional Neural Network), GPU (Graphics Processing Unit), NIH (National Institutes of Health), NLP (Natural Language Processing), RADS (Reporting, and Data System)
      To read this article in full you will need to make a payment

      Purchase one-time access:

      Academic & Personal: 24 hour online accessCorporate R&D Professionals: 24 hour online access
      One-time access price info
      • For academic or personal research use, select 'Academic and Personal'
      • For corporate R&D use, select 'Corporate R&D Professionals'

      Subscribe:

      Subscribe to Academic Radiology
      Already a print subscriber? Claim online access
      Already an online subscriber? Sign in
      Institutional Access: Sign in to ScienceDirect

      References

        • Krizhevsky A
        • Sutskever I
        • Hinton GE.
        ImageNet classification with deep convolutional neural networks.
        (In: Pereira F, Burges CJC, Bottou L, Weinberger KQ, editors)Advances in neural information processing systems 25. Curran Associates, Inc., Red Hook, NY2012: 1097-1105
        • Roberts M
        • Driggs D
        • Thorpe M
        • et al.
        Common pitfalls and recommendations for using machine learning to detect and prognosticate for COVID-19 using chest radiographs and CT scans.
        Nat Mach Intell. 2021; 3: 199-217
        • Reardon S.
        Rise of robot radiologists.
        Nature. 2019; 576: S54-S58
        • Sahiner B
        • Pezeshk A
        • Hadjiiski LM
        • et al.
        Deep learning in medical imaging and radiation therapy.
        Med Phys. 2019; 46: e1-e36
        • Jin C
        • Yu H
        • Ke J
        • et al.
        Predicting treatment response from longitudinal images using multi-task deep learning.
        Nat Commun. 2021; 12: 1851
        • Isensee F
        • Jaeger PF
        • Kohl SAA
        • et al.
        nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation.
        Nat Methods. 2021; 18: 203-211
      1. Banerjee I, Bhimireddy AR, Burns JL, et al. Reading race: AI recognises patient's racial identity in medical images 2021 Accessed at: February 22, 2022. Accessed from: https://arxiv.org/abs/2107.10356

        • Wang X
        • Peng Y
        • Lu L
        • et al.
        ChestX-Ray8: hospital-scale chest X-ray database and benchmarks on weakly-supervised classification and localization of common thorax diseases.
        in: 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). IEEE Computer Society, 2017: 3462-3471 (Jul 1)
        • Oakden-Rayner L.
        Exploring large-scale public medical image datasets.
        Acad Radiol. 2020; 27: 106-112
        • Taylor AG
        • Mielke C
        • Mongan J.
        Automated detection of moderate and large pneumothorax on frontal chest X-rays using deep convolutional neural networks: a retrospective study.
        PLoS Med. 2018; 15e1002697
        • Hallinan JTPD
        • Feng M
        • Ng D
        • et al.
        Detection of pneumothorax with deep learning models: learning from radiologist labels vs natural language processing model generated labels.
        Acad Radiol. 2021; (PubMed is S1076-6332(21)00427-X)
        • Lee B
        • Whitehead MT.
        Radiology reports: what you think you're saying and what they think you're saying.
        Curr Probl Diagn Radiol. 2017; 46: 186-195
        • Watadani T
        • Sakai F
        • Johkoh T
        • et al.
        Interobserver variability in the CT assessment of honeycombing in the lungs.
        Radiology. 2013; 266: 936-944
        • Cury RC
        • Abbara S
        • Achenbach S
        • et al.
        Coronary artery disease - reporting and data system (CAD-RADS): an expert consensus document of SCCT, ACR and NASCI: endorsed by the ACC.
        JACC Cardiovasc Imaging. 2016; 9: 1099-1113
        • Kohli M
        • Alkasab T
        • Wang K
        • et al.
        Bending the artificial intelligence curve for radiology: informatics tools from ACR and RSNA.
        J Am Coll Radiol. 2019; 16: 1464-1470