Deep learning in structural and functional lung image analysis

Abstract

The recent resurgence of deep learning (DL) has dramatically influenced the medical imaging field. Medical image analysis applications have been at the forefront of DL research efforts applied to multiple diseases and organs, including those of the lungs. The aims of this review are twofold: (i) to briefly overview DL theory as it relates to lung image analysis; (ii) to systematically review the DL research literature relating to the lung image analysis applications of segmentation, reconstruction, registration and synthesis. The review was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. 479 studies were initially identified from the literature search with 82 studies meeting the eligibility criteria. Segmentation was the most common lung image analysis DL application (65.9% of papers reviewed). DL has shown impressive results when applied to segmentation of the whole lung and other pulmonary structures. DL has also shown great potential for applications in image registration, reconstruction and synthesis. However, the majority of published studies have been limited to structural lung imaging with only 12.9% of reviewed studies employing functional lung imaging modalities, thus highlighting significant opportunities for further research in this field. Although the field of DL in lung image analysis is rapidly expanding, concerns over inconsistent validation and evaluation strategies, intersite generalisability, transparency of methodological detail and interpretability need to be addressed before widespread adoption in clinical lung imaging workflow.

Large-scale investigation of deep learning approaches for ventilated lung segmentation using multi-nuclear hyperpolarized gas MRI

“…ventilation defects. Several pre-processing techniques have previously been used in the literature for lung image segmentation 42 . The work of Tustison, et al 21 utilizes a novel template-based data augmentation strategy with N4 bias correction and denoising, which are computationally expensive and time-consuming; however, the impact of such techniques is not assessed in their work.…”

Computer-assisted image-based risk analysis and planning in lung surgery - a review

“…Proper implementation of software for building and training such networks is emphasized. A review on deep learning based structural and functional analysis across a variety of lung imaging modalities is given in ( 45 ). The authors give an overview of the deep learning research literature with regard to lung image analysis applications.…”

Investigating the use of machine learning to generate ventilation images from CT scans

“…A recent systematic review has found that deep learning applied to functional lung imaging is a relatively small field with good opportunities for further research. 20 4DCT perfusion images have been synthesized using a deep learning approach with MAA-SPECT as the nuclear medicine ground-truth. 21 Deep learning has been implemented to generate CTVIs from 4DCT with DIR-based CTVIs as the reference ground-truth.…”

“…One type of machine learning is deep learning and is characterized by hidden layers and feature learning. A recent systematic review has found that deep learning applied to functional lung imaging is a relatively small field with good opportunities for further research 20 . 4DCT perfusion images have been synthesized using a deep learning approach with MAA‐SPECT as the nuclear medicine ground‐truth 21 .…”