Current AI research in pathologic myopia and detection of associated complications of myopia
| Authors and year | Title | Outcome measures | Modalities | AI models | Total sample size | Diagnostic performance |
| Hemelings et al 202191 | Pathological myopia classification with simultaneous lesion segmentation using deep learning | Detection of pathologic myopia; fovea localisation; segmentation of optic disc, retinal atrophy, and retinal detachment | Fundus images | DL-CNN | 1200 images | Detection of pathologic myopia: AUC 0.9867; foveal localisation: 58.27 pixels |
| Rauf et al 202192 | Automatic detection of pathological myopia using machine learning | Detection of pathologic myopia | Fundus images | DL-CNN | 800 images | AUC 0.9845; accuracy 95% |
| Lu et al 202193 | Development of deep learning-based detecting systems for pathologic myopia using retinal fundus images | Detection of pathologic myopia | Fundus images | DL-CNN | 16428 images | AUC 0.979; accuracy 0.963 |
| Du et al 202194 | Deep learning approach for automated detection of myopic maculopathy and pathologic myopia in fundus images | Detection of pathologic myopia and myopic maculopathy (diffuse atrophy, patchy atrophy, macular atrophy, mCNV) | Fundus images | DL-CNN | 7020 images | Diffuse atrophy AUC 0.970 sensitivity 84.44%; patchy atrophy AUC 0.978 sensitivity 87.22%; macular atrophy AUC 0.982 sensitivity 85.10%; choroidal neovascularisation AUC 0.881 sensitivity 37.07% |
| Tan et al 202195 | Retinal photograph-based deep learning algorithms for myopia and a blockchain platform to facilitate artificial intelligence medical research: a retrospective multicohort study | Detection of high myopia and MMD | Fundus images | DL-CNN | 226686 images | Detection of high myopia: AUC >0.913; detection of MMD: AUC >0.969 |
| Lu et al 202196 | AI-Model for Identifying Pathologic Myopia Based on Deep Learning Algorithms of Myopic Maculopathy Classification and ‘Plus’ Lesion Detection in Fundus Images | Detection of pathologic myopia, classification of myopic maculopathy | Fundus images | DL-CNN | 37659 images | AUC 0.995; accuracy 97.36%; sensitivity 93.92%; specificity 98.19% |
| Du et al 202197 | Automatic identification of myopic maculopathy related imaging features in optic disc region via machine learning methods | Identification of myopic maculopathy imaging features | Fundus images | ML | 457 eyes | Eight new myopic maculopathy-related image features were discovered |
| Choi et al 202198 | Deep learning models for screening of high myopia using optical coherence tomography | Detection of high myopia | OCT images | DL-CNN | 690 eyes | AUC 0.86–0.99 |
| Du et al 202199 | Validation of Soft Labels in Developing Deep Learning Algorithms for Detecting Lesions of Myopic Maculopathy From Optical Coherence Tomographic Images | Detection of myopic maculopathy | OCT images | DL | 9176 images | mCNV AUC 0.985; MTM AUC 0.946; DSM AUC 0.978 |
| Aytekin et al 2020100 | Development and validation of a deep learning system to screen vision-threatening conditions in high myopia using optical coherence tomography images | Detection of retinoschisis, macular hole, retinal detachment, mCNV | OCT images | DL-CNN | 5505 images | AUC 0.961–0.999; sensitivity & specificity >90% |
| Sogawa et al 2020101 | Accuracy of a deep convolutional neural network in the detection of myopic macular diseases using swept-source optical coherence tomography | Detection of myopic macular lesions (mCNV, retinoschisis) | Swept-source OCT images | DL-CNN | 910 images | Detection of myopic macular lesions: AUC 0.970; sensitivity 90.6%; specificity 94.2% |
| Ye et al 2021102 | Automatic Screening and Identifying Myopic Maculopathy on Optical Coherence Tomography Images Using Deep Learning | Detection of myopic maculopathy | OCT images | DL-CNN | 2342 images | AUC 0.927–0.974 |
| Sawai et al 2020106 | Usefulness of Denoising process to Depict Myopic choroidal neovascularisation Using a Single optical coherence tomography Angiography image | Novel denoising process for depicting mCNV | OCTA images | DL | 20 eyes | Use single OCTA images to provide results comparable to averaged OCTA images |
AI, artificial intelligence; AUC, area under the curve; CNN, convolution neural network; DL, deep learning; DSM, dome-shaped macula; mCNV, myopic choroidal neovascularisation; ML, machine learning; MMD, myopic macular degeneration; MTM, myopic tractional maculopathy; OCT, optical coherence tomography; OCTA, OCT angiography.