![References
● Ma, J., Song, Y., Tian, X. et al. Survey on deep learning for pulmonary medical imaging. Front.
Med. (2019). https://doi.org/10.1007/s11684-019-0726-4
● Ozturk T, Talo M, Yildirim EA, Baloglu UB, Yildirim O, Rajendra Acharya U. Automated detection of COVID-19
cases using deep neural networks with X-ray images [published online ahead of print, 2020 Apr 28]. Comput
Biol Med. 2020;103792. https://doi.org/10.1016/j.compbiomed.2020.103792
● Q. Dou, H. Chen, L. Yu, J. Qin, and P. Heng, “Multilevel Contextual 3-D CNNs for False Positive Reduction in
Pulmonary Nodule Detection,” in IEEE Transactions on Biomedical Engineering, vol. 64, no. 7, pp. 1558–1567, July
2017, 10.1109/TBME.2016.2613502
● Magalhães Barros Netto S, Corrêa Silva A, Acatauassú Nunes R, Gattass M. Automatic segmentation of lung
nodules with growing neural gas and support vector machine. Comput Biol Med 2012; 42(11): 1110–
1121https://doi.org/10.1016/j.compbiomed.2012.09.003
● Ronneberger O, Fischer P, Brox T. U-net: convolutional networks for biomedical image segmentation. In:
International Conference on Medical image computing and computer-assisted intervention. Springer. 2015.
234–241 https://doi.org/10.1007/978-3-319-24574-4_28](https://image.slidesharecdn.com/presentation-200517190619/85/Detect-COVID-19-with-Deep-Learning-A-survey-on-Deep-Learning-for-Pulmonary-Medical-Imaging-17-320.jpg)
Detect COVID-19 with Deep Learning- A survey on Deep Learning for Pulmonary Medical Imaging
- 1. Detect COVID-19 with Deep Learning- A survey on Deep Learning for Pulmonary Medical Imaging By- Jumana Nadir Medium Article Short Story
- 2. Introduction ● Who knew Deep Learning can come so handy to us during this period of global crisis? ● There has yet been no vaccine or any effective treatment for the 2019 novel Coronavirus (COVID-19), but generative deep learning is helping in detecting and monitoring coronavirus patients by chest CT screening.
- 3. Survey on deep learning for pulmonary medical imaging https://link.springer.com/article/10.1007/s11684-019-0726-4 ● Survey Authors- Jiechao Ma, Yang Song, Xi Tian, Yiting Hua, Rongguo Zhang, and Jianlin Wu. ● Published: 16 December 2019 ● Abstract- Elaborates the state-of-the-art Deep Learning techniques used to detect various lung diseases, Lung cancers, Pneumonia, Tuberculosis, and its contribution to the classification, detection, and segmentation of Pulmonary Medical Diseases.
- 4. Contents ● Overview Of Deep Learning. ● 3 important aspects of Medical Image Analysis. ● Deep Learning in Medical Pulmonary Image (Lung Cancer). ● Pneumonia, Tuberculosis, Interstitial Lung Diseases. ● Existing Datasets with download links. ● Conclusion.
- 5. Overview ● Modern lifestyle changes, pollution, and global warming have attracted many Respiratory Diseases which are life- threatening. ● No signs during early stage. ● People miss this period of early treatment and realize when it becomes serious.
- 6. Overview Of Deep Learning ●In medical imaging accurate diagnosis depends on image collection and its computer-aided diagnosis (CAD) ●1993, first time Neural Networks were used, but were not accepted because of computation requirements. ●Detection Criteria- Finding possible lesions and tumors (region/organ suffering damage) ●Shape, size, density, textures
- 7. History
- 8. 3 important aspects of Medical Image Analysis. 1. Classification- Normal/Abnormal; Binary/Multiclass 2. Detection- Detect Region of Interest (ROI) 3. Segmentation- Segment meaningful parts- organs, substructure, lesion, extract features
- 9. Deep Learning in Medical Pulmonary Image (Lung Cancer) ● Classify Malignant/ Benign using CNN and SVM. ● Detection using RCNN and DCNN (high sensitivity + precision required). ● Segmentation label- generate accurate voxel-level nodule segmentation.
- 10. Fatal Lung Diseases and their Diagnosis with Deep Learning
- 11. Pulmonary Embolism ● About 650000 cases occur annually ● Artery in lung becomes partially or completely blocked. ● Doctors approach- Angiography ● DL- Neural Hypernetworks ● Knowledge base hybrid learning algorithms
- 12. Pneumonia ●Common among Children ●Early detection can save lot of lives ●X-ray examination is most common method ●Images are very similar ●Template matching algorithm are used in CUDA and CNN architectures
- 13. Tuberculosis ● Respiratory tract disease, caused by pathogen ‘Mycobacterium tuberculosis’. ● Common methods are X-rays, patient’s signs and symptoms, mucus exam. ● Multi-instance learning combined with RNN’s achieved good results.
- 14. Interstitial Lung Diseases ● Heterogenous, non-neoplastic, non-infectious disease ● Contains abnormal imaging patterns ● Below models achieved good results
- 15. Existing Datasets with download links ● LIDC- Lung Image Database Consortium- consists of chest medical images of 1018 research cases. ● LUNA16- it’s a subset of the above dataset that contains low-dose lung CT images. ● Pneumonia Dataset by National Institutes of Health (size- 42GB). ● Tuberculosis Dataset by Shenzhen Hospital (size- 4GB). ● Geneva Database has Interstitial Lung Disease Dataset. Performances of the two pulmonary nodule datasets LIDC- IDRI, and LUNA16-
- 16. Conclusion ● Lots of prospects with emerging Deep Learning technology. ● DL can transform Healthcare System ● Can Machines replace Doctors?
- 17. References ● Ma, J., Song, Y., Tian, X. et al. Survey on deep learning for pulmonary medical imaging. Front. Med. (2019). https://doi.org/10.1007/s11684-019-0726-4 ● Ozturk T, Talo M, Yildirim EA, Baloglu UB, Yildirim O, Rajendra Acharya U. Automated detection of COVID-19 cases using deep neural networks with X-ray images [published online ahead of print, 2020 Apr 28]. Comput Biol Med. 2020;103792. https://doi.org/10.1016/j.compbiomed.2020.103792 ● Q. Dou, H. Chen, L. Yu, J. Qin, and P. Heng, “Multilevel Contextual 3-D CNNs for False Positive Reduction in Pulmonary Nodule Detection,” in IEEE Transactions on Biomedical Engineering, vol. 64, no. 7, pp. 1558–1567, July 2017, 10.1109/TBME.2016.2613502 ● Magalhães Barros Netto S, Corrêa Silva A, Acatauassú Nunes R, Gattass M. Automatic segmentation of lung nodules with growing neural gas and support vector machine. Comput Biol Med 2012; 42(11): 1110– 1121https://doi.org/10.1016/j.compbiomed.2012.09.003 ● Ronneberger O, Fischer P, Brox T. U-net: convolutional networks for biomedical image segmentation. In: International Conference on Medical image computing and computer-assisted intervention. Springer. 2015. 234–241 https://doi.org/10.1007/978-3-319-24574-4_28