|
| Challenges | Description and possible solutions |
|
| Preprocessing | Medical images suffer from different types of noise/artifacts and missing data. Noise reduction, artifact removal, missing data handling, contrast adjusting, and so forth could enhance the quality of images and increase the performance of processing methods. Employing multimodal data could be beneficial for this purpose [63–65]. |
|
| Compression | Reducing the volume of data while maintaining important data such as anatomically relevant data [55, 61, 66]. |
|
| Parallelization/real-time realization | Developing scalable/parallel methods and frameworks to speed up the analysis/processing [61]. |
|
| Registration/mapping | Aligning consecutive slices/frames from one scan or corresponding images from different modalities [67, 68]. |
|
| Sharing/security/anonymization | Integrity, privacy, and confidentiality of data must be protected [55, 69–71]. |
|
| Segmentation | Delineation of anatomical structure such as vessels and bones [50, 68, 72]. |
|
| Data integration/mining | Finding dependencies/patterns among multimodal data and/or the data captured at different time points in order to increase the accuracy of diagnosis, prediction, and overall performance of the system [47, 49, 52, 73]. |
|
| Validation | Assessing the performance or accuracy of the system/method. Validation can be objective or subjective. For the former, annotated data is usually required [74–76]. |
|
