Development of an automated region-of-interest-setting method based on a deep neural network for brain perfusion single photon emission computed tomography quantification methods

Document Type : Original Article

Authors

1 Graduate School of Health Sciences, Kumamoto University, Kumamoto, Japan

2 Department of Central Radiology Kumamoto University Hospital, Kumamoto, Japan

3 Department of Diagnostic Radiology, Faculty of Life Sciences, Kumamoto University, Kumamoto, Japan

4 Department of Information and Communication Technology, Faculty of Engineering, University of Miyazaki

5 Department of Medical Imaging, Faculty of Life Sciences, Kumamoto, Kumamoto , Japan

10.22038/aojnmb.2024.75375.1528

Abstract

Objective(s): A simple noninvasive microsphere (SIMS) method using 123I-IMP and an improved brain uptake ratio (IBUR) method using 99mTc-ECD for the quantitative measurement of regional cerebral blood flow have been recently reported. The input functions of these methods were determined using the administered dose, which was obtained by analyzing the time activity curve of the pulmonary artery (PA) for SIMS and the ascending aorta (AAo) for the IBUR methods for dynamic chest images. If the PA and AAo regions of interest (ROIs) can be determined using deep convolutional neural networks (DCNN) for segmentation, the accuracy of these ROI-setting methods can be improved through simple analytical operations to ensure repeatability and reproducibility. The purpose of this study was to develop new PA and AAo-ROI setting methods using a DCNN (DCNN-ROI method).
Methods: A U-Net architecture based on convolutional neural networks was used to determine the PA and AAo candidate regions. Images of 290 patients who underwent 123I-IMP RI-angiography and 108 patients who underwent 99mTc-ECD RI-angiography were used. The PA and AAo-ROI results for the DCNN-ROI method were compared to those obtained using manual methods. The counts for the input function on the PA and AAo-ROI were determined by integrating the area under the curve (AUC) counts of the time-activity curve of PA and AAo-ROI, respectively. The effectiveness of the DCNN-ROI method was elucidated through a comparison with the integrated AUC counts of the DCNN-ROI and the manual ROI.
Results: The coincidence ratio for the locations of the PA and AAo-ROI obtained using the DCNN method and that for the manual method was 100%. Strong correlations were observed between the AUC counts using the DCNN and manual methods.
Conclusion: New ROI- setting programs were developed using a deep convolution neural network DCNN to determine the input functions for the SIMS and IBUR methods. The accuracy of these methods is comparable to that of the manual method.

Keywords

Main Subjects

References

  1. Wong CH, Mohamed A, Larcos G, McCredie R, Somerville E, Bleasel A. Brain activation patterns of versive, hypermotor, and bilateral asymmetric tonic seizures. Epilepsia. 2010; 51(10):2131-2139.
  2. Newberg AB, Wintering N, Khalsa DS, Roggenkamp H, Waldman MR. Meditation effects on cognitive function and cerebral blood flow in subjects with memory loss: a preliminary study. J Alzheimers Dis. 2010; 20(2):517-526.
  3. Inui Y, Toyama H, Manabe Y, Sato T, Sarai M, Kosaka K, et al. Evaluation of probable or possible dementia with Lewy bodies using 123I-IMP brain perfusion SPECT, 123I-MIBG, and 99mTc-MIBI myocardial SPECT. J Nucl Med. 2007; 48(10):1641-1650.
  4. Kanai Y, Hasegawa S, Kimura Y, Oku N, Ito H, Fukuda H, et al. N-isopropyl-4- [123I] iodoamphetamine (123I-IMP) products: a difference in radiochemical purity, un-metabolized fraction, and octanol extraction function in arterial blood and regional brain uptake in rats. Ann Nucl Med. 2007; 21(7):387-391.
  5. Matsuda H, Oba H, Terada H, Tsuj S, Sumi H, Shiba K, et al. Quantitative assessment of cerebral blood flow using technetium-99m-hexamethyl-propyleneamine oxime: Part I, Design of a mathematical model. Ann Nucl Med. 1988; 2(1):13-19.
  6. Van Laere K, Dumont F, Koole M, Dierckx R. Noninvasive methods for absolute cerebral blood flow measurement using 99mTc-ECD: a study in healthy volunteers. Eur J Nucl Med. 2001; 28(7):862-872.
  7. Odano I, Ohkubo M, Yokoi T. Noninvasive quantification of cerebral blood flow using 99mTc-ECD and SPECT. J Nucl Med. 1999; 40(10):1737-1744.
  8. Kuhl DE, Barrio JR, Huang SC, Selin C, Ackermann RF, Lear JL, et al. Quantifying local cerebral blood flow by N-isopropyl-p-[123I] iodoamphetamine (IMP) tomography. J Nucl Med. 1982; 23(3):196-203.
  9. Iida H, Itoh H, Nakazawa M, Hatazawa J, Nishimura H, Onishi Y, et al. Quantitative Mapping of Regional Cerebral Blood Flow Using Iodine-123-IMP and SPECT. J Nucl Med. 1994; 35(12):2019-2030.
  10. Kyeong MK, Watanabe H, Hayashi T, Hayashida K, Katafuchi T, Enomoto N, et al. Quantiative mapping of basal and vasareactive cerebral blood flow using split-dose 123I-iodoamphetamine and single photon emission computed tomography. NeuroImage. 2006; 33(4):1126-1135.
  11. Iida H, Narita Y, Kado H, Kashikura A, Sugawara S, Shoji, Y. Effects of scatter and attenuation correction on quantitative assessment of regional cerebral blood flow with SPECT. J Nucl Med. 1998; 39(1):181-189.
  12. Nishizawa S, Yonekura Y, Tanaka F, Fujita T, Tsuchimochi S, Ishizu, K, et al. Evaluation of a double-injection method for sequential measurement of cerebral blood flow with iodine-123-iodoamphetamine. J Nucl Med. 1995; 36(7):1339-1345.
  13. Matsuda H, Tsuji S, Shuke N, Sumiya H, Tonami N, Hisada K. Noninvasive measurements of regional cerebral blood flow using technetium-99m hexamethyl-propylene amine oxime. Eur J Nucl Med. 1993; 20(5):391-401.
  14. Tomiguchi S, Tashiro K, Shiraishi S, Yoshida M, Kawanaka K, Takahashi Y. Estimation of 123I-IMP arterial blood activity from dynamic planar imaging of the chest using a graph plot method for the quantification of regional cerebral blood flow. Ann Nucl Med. 2010; 24(5):387-393.
  15. Iseya O, Mihara T, Suzuki K, Miyamae T, Matsuda H. Evaluation of the 123I-IMP Patlak plot method using the pulmonary differential curve as an input function: a comparison with cerebral blood flow (CBF) determined by the noninvasive micro-sphere (NIMS) method. Kaku Igaku. 2003; 40(2):163-174.
  16. Miyazaki Y, Kinuya S, Hashimoto M, Satake R, Inoue H, Shiozaki J, et al. Brain uptake ratio as an index of cerebral blood flow obtained with 99mTc-ECD. Kaku Igaku. 1997; 34(1):49-52.
  17. Takeuchi R, Matsuda H, Yonekura Y, Sakahara H, Konishi J. Noninvasive quantitative measurements of regional cerebral blood flow using Technetium-99m-L, L-ECD SPECT activated with acetazolamide: quantification analysis by equal-volume-split 99mTc-ECD consecutive SPECT method. J Cereb Blood Flow Metab. 1997; 17(10): 1020-1032.
  18. Takeuchi R, Yonekura Y, Matsuda H, Konishi J. Usefulness of a three-dimensional stereotaxic ROI template on anatomically standardised 99mTc-ECD SPET. Eur J Nucl Med Mol Imaging. 2002; 29(3):331-341.
  19. Ofuji A, Mimura H, Yamashita K, Takaki A, Sone T, Ito S. Development of a simple non-invasive microsphere quantification method for cerebral blood flow using I-123-IMP. Ann Nucl Med. 2016; 30(3):242-249.
  20. Ito S, Takaki A, Inoue S, Tomiguchi S, Shiraishi S, Akiyama Y, et al. Improvement of the 99mTc-ECD brain uptake ratio (BUR) method for measurement of cerebral blood flow. Ann Nucl Med. 2012; 26(4):351-358.
  21. Nagaoka R, Ofuji A, Yamashita K, Tomimatsu T, Orita S, Takaki A, et al. Usefulness of an automated quantitative method for measuring regional cerebral blood flow using 99mTc ethyl cysteinate dimer brain uptake ratio. Asia Ocean J Nucl Med Biol. 2015; 3(2):77-82.
  22. Masunaga S, Uchiyama Y, Ofuji A, Nagaoka R, Tomimatsu T, Iwata A, et al. Development of an automated ROI-setting program for input function determination in 99mTc-ECD non-invasive cerebral blood flow quantification. Phys Med. 2014; 30:513-520.
  23. Ofuji A, Nagaoka R, Yamashita K. Takaki A, Ito S. A simple noninvasive I-123-IMP autoradiography method developed by modifying the simple non-invasive I-123-IMP microsphere method. Asia Ocean J Nucl Med Biol. 2018; 6(1):50-56.
  24. Yamashita K, Uchiyama Y, Ofuji A, Mimura H, Okumiya S, Takaki A, et al. Fully automated input function determination program for simple noninvasive 123I-IMP microsphere cerebral blood flow quantification method. Phys Med. 2016; 32(9):1180-1185.
  25. LeCun Y, Bengio Y, Hinton G. Deep learning. Nature. 2015; 521:436–444.
  26. Yamashita R, Nishio M, Do RKG, Togashi K. Convolutional neural networks: an overview and application in radiology. Insights into Imaging. 2018; 9(4):611–629.
  27. Ronneberger O, Fischer P, Brox T. U-Net: Convolutional networks for biomedical image segmentation. Medical image computing and computer-assisted intervention - MICCAI 2015. 2015.
  28. Hassanzadeh T, Essam D, Sarker R. 2D to 3D Evolutionary Deep Convolutional Neural Networks for Medical Image Segmentation.IEEE Trans Med Imaging. 2021; 40(2):712-721.
  29. Zhang S, Zhou Y, Tang D, Ni M, Zheng J, Xu G, et al. A deep learning-based segmentation system for rapid onsite cytologic pathology evaluation of pancreatic masses: A retrospective, multicenter, diagnostic study. EBioMedicine. 2022; 80:104022.
  30. Jeon U, Kim H, Hong H, Wang J. Automatic Meniscus Segmentation Using Adversarial Learning-Based Segmentation Network with Object-Aware Map in Knee MR Images. Diagnostics. 2021; 11(9):1612.
  31. Chen X, Zhou B, Xie H, Shi L, Liu H, Holler W, et al. Direct and indirect strategies of deep-learning-based attenuation correction for general purpose and dedicated cardiac SPECT. Eur J Nucl Med Mol Imaging. 2022; 49(9):3046-3060.
  32. Hashimoto F, Ito M, Ote K, Isobe T, Okada H, Ouchi, Y. Deep learning-based attenuation correction for brain PET with various radiotracers. Ann Nucl Med. 2021; 35(6): 691-701.
  33. Armanious K, Hepp T, Küstner T, Dittmann H, Nikolaou K, La Fougère C, et al. Independent attenuation correction of whole-body [18F]FDG-PET using a deep learning approach with generative adversarial networks. EJNMMI Research. 2020; 10(1):53.
  34. Yin XX, Sun L, Fu Y, Lu R, Zhang Y. U-Net-Based medical image segmentation. J Healthcare Eng. 2022; 2022:4189781.
  35. Sun H, Jiang Y, Yuan J, Wang H, Liang D, Fan W, et al. High-quality PET image synthesis from ultra-low-dose PET/MRI using bi-task deep learning. Quant Imaging Med Surg. 2022; 12(12):5326-5342.
Asia Oceania Journal of Nuclear Medicine and Biology
Volume 12, Issue 2
July 2024
Pages 120-130

Files

Share

How to cite

Statistics