Edit this page

NA-MIC Project Weeks

Back to Projects List

DICOM series classification and visualization of parameters

Key Investigators

Presenter location: In-person

Project Description

To use and develop AI methods, significant data curation is required. In some cases like prostate cancer segmentation, clinicians often use multiple MRI sequences for diagnosis such as T2, diffusion-weighted series, and derived maps.

Unfortunately, the information describing the sequences is often missing or incorrect, as it’s prone to errors from technicians. The proper sequence could be analyzed visually, but this is cumbersome if thousands of scans need to be analyzed. Therefore, automatic methods for determining the right series are of interest.

We propose methods to aid in the curation of DICOM data, as well as aids to help in vizualization of DICOM parameters.

Objective

We would like to develop approaches for aiding in the curation of data. The first will be the development of visualization tools to understand DICOM parameters used in scanning. Secondly, we will develop AI methods for classifying MRI scans, with a focus on prostate cancer.

Approach and Plan

  1. Use packages such as hiplot to visualize DICOM scanning parameters across different collections and modalities in IDC.
  2. Develop approaches for data curation using AI - e.g. determine the scan sequence, or if endorectal coil is present, etc.

Progress and Next Steps

  1. Started repo here for initial hiplot exploration of DICOM tags of T2 weighted axial series of prostate imaging collections from IDC
  2. Had some very helpful discussions with David, Maria and Chris about understanding of parameters and previous work done in this area
  3. Created similar interactive plots for DWI and ADC across different prostate collections
  4. Developed a hierarchical approach for classification of prostate scans, starting with ProstateX collection – for T2 axial, DWI, ADC, and DCE classification.
  5. We’ll later try this out on other prostate collections.

Illustrations

Hiplot visualization of T2 weighted axial parameters from 5 different prostate cancer imaging collections in IDC

Same hiplot but with rendering in the browser!

** Workflow **

PRODICOM

Background and References

GitHub repo

Some earlier work with parallel coordinates plots in Slicer:

Some earlier work on sequence classification:

  1. Ranjbar S, Singleton KW, Jackson PR, Rickertsen CR, Whitmire SA, Clark-Swanson KR, et al. A Deep Convolutional Neural Network for Annotation of Magnetic Resonance Imaging Sequence Type. J Digit Imaging. 2020 Apr;33(2):439–46. doi:10.1007/s10278-019-00282-4
  2. Noguchi T, Higa D, Asada T, Kawata Y, Machitori A, Shida Y, et al. Artificial intelligence using neural network architecture for radiology (AINNAR): classification of MR imaging sequences. Jpn J Radiol. 2018 Dec;36(12):691–7. doi:10.1007/s11604-018-0779-3
  3. Cluceru J, Interian Y, Lupo JM, Bove R, Butte A, Crane J. Automatic Classification of MR Image Contrast. In: ISMRM. 2020. Available from: https://archive.ismrm.org/2020/1804.html
  4. Remedios S, Roy S, Pham DL, Butman JA. Classifying magnetic resonance image modalities with convolutional neural networks. In: Mori K, Petrick N, editors. Medical Imaging 2018: Computer-Aided Diagnosis. Houston, United States: SPIE; 2018. p. 89. Available from: https://www.spiedigitallibrary.org/conference-proceedings-of-spie/10575/2293943/Classifying-magnetic-resonance-image-modalities-with-convolutional-neural-networks/10.1117/12.2293943.full doi:10.1117/12.2293943
  5. Braeker N, Schmitz C, Wagner N, Stanicki BJ, Schröder C, Ehret F, et al. Classifying the Acquisition Sequence for Brain MRIs Using Neural Networks on Single Slices. Cureus. 2022 Feb;14(2):e22435. doi:10.7759/cureus.22435
  6. Vieira de Mello JP, Paixão TM, Berriel R, Reyes M, Badue C, De Souza AF, et al. Deep Learning-based Type Identification of Volumetric MRI Sequences. In: 2020 25th International Conference on Pattern Recognition (ICPR). Milan, Italy: IEEE; 2021. p. 1–8. Available from: https://ieeexplore.ieee.org/document/9413120 doi:10.1109/ICPR48806.2021.9413120
  7. Mahmutoglu MA, Preetha CJ, Meredig H, Tonn J-C, Weller M, Wick W, et al. Deep Learning–based Identification of Brain MRI Sequences Using a Model Trained on Large Multicentric Study Cohorts. Radiology: Artificial Intelligence. 2024 Jan;6(1):e230095. doi:10.1148/ryai.230095
  8. Kasmanoff N, Lee MD, Razavian N, Lui YW. Deep multi-task learning and random forest for series classification by pulse sequence type and orientation. Neuroradiology. 2023 Jan 1;65(1):77–87. doi:10.1007/s00234-022-03023-7
  9. Svdvoort. DeepDicomSort. 2022. Available from: https://github.com/Svdvoort/DeepDicomSort
  10. van der Voort SR, Smits M, Klein S, for the Alzheimer’s Disease Neuroimaging Initiative. DeepDicomSort: An Automatic Sorting Algorithm for Brain Magnetic Resonance Imaging Data. Neuroinform. 2021 Jan 1;19(1):159–84. doi:10.1007/s12021-020-09475-7
  11. HD-SEQ-ID. www.neuroAI-HD.org; 2023. Available from: https://github.com/NeuroAI-HD/HD-SEQ-ID
  12. HeuDiConv. NIPY developers; 2022. Available from: https://github.com/nipy/heudiconv
  13. Gai ND. Highly Efficient and Accurate Deep Learning–Based Classification of MRI Contrast on a CPU and GPU. J Digit Imaging. 2022 Jun 1;35(3):482–95. doi:10.1007/s10278-022-00583-1
  14. Cluceru J, Lupo JM, Interian Y, Bove R, Crane JC. Improving the Automatic Classification of Brain MRI Acquisition Contrast with Machine Learning. J Digit Imaging. 2023 Feb;36(1):289–305. doi:10.1007/s10278-022-00690-z
  15. Mello JPV de. Jpvmello/type-identification-mri-sequences. 2023. Available from: https://github.com/Jpvmello/type-identification-mri-sequences
  16. Kasmanoff N. MRI Content Detection. 2022. Available from: https://github.com/nkasmanoff/mri-content-detection
  17. MRI Sequence Classification - No overlapping. Available from: https://docs.google.com/document/d/1UmE7jFfWaAxsS6wXodPRkdKeyDAcVdEWXfMKGzGiiKk/edit?usp=sharing&usp=embed_facebook
  18. T1 vs T2 MRI T1and T2 MRI image comparison. mrimaster. Available from: https://mrimaster.com/t1-vs-t2-mri/
  19. Pizarro R, Assemlal H-E, De Nigris D, Elliott C, Antel S, Arnold D, et al. Using Deep Learning Algorithms to Automatically Identify the Brain MRI Contrast: Implications for Managing Large Databases. Neuroinform. 2019;17(1):115–30. doi:10.1007/s12021-018-9387-8
  20. Gauriau R, Bridge C, Chen L, Kitamura F, Tenenholtz NA, Kirsch JE, et al. Using DICOM Metadata for Radiological Image Series Categorization: a Feasibility Study on Large Clinical Brain MRI Datasets. J Digit Imaging. 2020 Jun;33(3):747–62. doi:10.1007/s10278-019-00308-x