Share:
Share this content in WeChat
X
Technical Article
Study on acceleration efficiency and image quality of artificial intelligence compressed sensing and compressed sensing in knee MRI
PAN Ke  LIU Qianqian  TANG Lingling  HU Yuntao  HUANG Xiaohua 

Cite this article as: Pan K, Liu QQ, Tang LL, et al. Study on acceleration efficiency and image quality of artificial intelligence compressed sensing and compressed sensing in knee MRI[J]. Chin J Magn Reson Imaging, 2022, 13(5): 94-98. DOI:10.12015/issn.1674-8034.2022.05.017.


[Abstract] Objective To evaluate the utility of compressed sensing (CS) and artificial intelligence compressed sensing (ACS) in knee magnetic resonance imaging (MRI) and the effect on image quality.Materials and Methods Sixty-seven subjects were scanned with a 3.0 T scanner, undergoing proton density weighted imaging (PDWI) using three CS factors (CS 2.0, CS 2.5, CS 3.0), four ACS factors (ACS 2.5, ACS 3.0, ACS 3.5, ACS 4.0), with parallel imaging (PI) 2.0 as the reference. Two independent readers rated the overall image quality with a 4-point subjective scale. ROIs were placed on media femoral condyle, the medial head of gastrocnemius muscle, the intercondylar fossa effusion, the infrapatellar fat pad and lateral condylar cartilage of femur to measure signal intensity and noise, then signal-to-noise (SNR) were calculated as objective index. For the subjective scoring inter-rater reliability of the two readers was calculated using Cohen kappa statistics and the subjective scores and objective index were evaluated using Kruskal-Walli's test and Wilcoxon signed rank test with SPSS 23.0 software.Results Scan time decreased with increasing acceleration factors (PI 2.0: 152 s; CS 2.0: 128 s, CS 2.5: 104 s, CS 3.0: 86 s; ACS 2.5: 76 s, ACS 3.0: 65 s, ACS 3.0: 57 s, ACS 4.0: 51 s). The results of subjective scores by two readers showed strong agreement and almost perfect agreement (0.735≤κ≤0.869). There was significant difference in subjective scores and SNR among the 8 groups (P<0.05). There was no significant difference in subjective scores and 5 SNRs between CS 2.0, ACS 3.0 and PI 2.0 respectively (P>0.05). There was no significant difference in subjective scores and 4 SNRs between ACS 2.5 and PI 2.0, while a SNR of ACS 2.5 was higher than that of PI 2.0 significantly (P<0.05). The subjective scores of CS 2.5, CS 3.0, ACS 3.5, ACS 4.0 were lower than that of PI 2.0 significantly, and there was 1, 5, 2, 2 SNRs lower than those of PI 2.0 respectively(P<0.05).Conclusions ACS and CS could accelerate shorten MR acquisition time. Compared with CS, ACS was more effectively, and ACS could reduce the acquisition time by 57% on the premise of ensuring image quality in keen PDWI.
[Keywords] artificial intelligence compressed sensing;compressed sensing;magnetic resonance imaging;knee;acceleration

PAN Ke   LIU Qianqian   TANG Lingling   HU Yuntao   HUANG Xiaohua*  

Department of Radiology, Affiliated Hospital of North Sichuan Medical College, Nanchong 637000, China

Huang XH, E-mail: 15082797553@163.com

Conflicts of interest   None.

Received  2021-10-22
Accepted  2022-04-28
DOI: 10.12015/issn.1674-8034.2022.05.017
Cite this article as: Pan K, Liu QQ, Tang LL, et al. Study on acceleration efficiency and image quality of artificial intelligence compressed sensing and compressed sensing in knee MRI[J]. Chin J Magn Reson Imaging, 2022, 13(5): 94-98.DOI:10.12015/issn.1674-8034.2022.05.017

[1]
Liu Q, Niu JB, Li H, et al. Knee symptomatic osteoarthritis, walking disability, NSAIDs use and all-cause mortality: population-based Wuchuan osteoarthritis study[J]. Sci Rep, 2017, 7(1): 3309. DOI: 10.1038/s41598-017-03110-3.
[2]
Silverwood V, Blagojevic-Bucknall M, Jinks C, et al. Current evidence on risk factors for knee osteoarthritis in older adults: a systematic review and meta-analysis[J]. Osteoarthritis Cartilage, 2015, 23(4): 507-515. DOI: 10.1016/j.joca.2014.11.019.
[3]
Liu Y, Du GY, Liu J. Meniscal anterior and posterior horn heights are associated with MRI-defined knee structural abnormalities in middle-aged and elderly patients with symptomatic knee osteoarthritis[J]. BMC Musculoskelet Disord, 2022, 23: 218. DOI: 10.1186/s12891-022-05143-w.
[4]
Li XS, Hou Q, Zhan XH, et al. The accuracy of MRI in diagnosing and classifying acute traumatic multiple ligament knee injuries[J]. BMC Musculoskelet Disord, 2022, 23(1): 43. DOI: 10.1186/s12891-021-04976-1.
[5]
Phelan N, Rowland P, Galvin R, et al. A systematic review and meta-analysis of the diagnostic accuracy of MRI for suspected ACL and meniscal tears of the knee[J]. Knee Surg Sports Traumatol Arthrosc, 2016, 24(5): 1525-1539. DOI: 10.1007/s00167-015-3861-8.
[6]
The Joint Surgery Branch of the Chinese Orthopaedic Association, The Subspecialty Group of Osteoarthritis, Chinese Association of Orthopaedic Surgeons, The National Clinical Research Center for Geriatric Disorders (Xiangya Hospital). Chinese guideline for diagnosis and treatment of osteoarthritis (2021 edition)[J]. Chin J Orthop, 2021, 41(18): 1291-1314. DOI: 10.3760/cma.j.cn121113-20210624-00424.
[7]
Kloppenburg M, Berenbaum F. Osteoarthritis year in review 2019: epidemiology and therapy[J]. Osteoarthritis and cartilage, 2020, 28(3): 242-248. DOI: 10.1016/j.joca.2020.01.002.
[8]
Katz JN, Arant KR, Loeser RF. Diagnosis and treatment of hip and knee osteoarthritis: a review[J]. JAMA, 2021, 325(6): 568-578. DOI: 10.1001/jama.2020.22171.
[9]
Ariyachaipanich A, Bae WC, Statum S, et al. Update on MRI pulse sequences for the knee: imaging of cartilage, meniscus, tendon, and hardware[J]. Semin Musculoskelet Radiol, 2017, 21(2): 45-62. DOI: 10.1055/s-0037-1599209.
[10]
Deshmane A, Gulani V, Griswold MA, et al. Parallel MR imaging[J]. J Magn Reson Imaging, 2012, 36(1): 55-72. DOI: 10.1002/jmri.23639.
[11]
Lustig M, Donoho D, Pauly JM. Sparse MRI: the application of compressed sensing for rapid MR imaging[J]. Magn Reson Med, 2007, 58(6): 1182-1195. DOI: 10.1002/mrm.21391.
[12]
Delattre BMA, Boudabbous S, Hansen C, et al. Compressed sensing MRI of different organs: ready for clinical daily practice? [J]. Eur Radiol, 2020, 30(1): 308-319. DOI: 10.1007/s00330-019-06319-0.
[13]
Johnson PM, Recht MP, Knoll F. Improving the speed of MRI with artificial intelligence[J]. Semin Musculoskelet Radiol, 2020, 24(1): 12-20. DOI: 10.1055/s-0039-3400265.
[14]
Wang YD, Song Y, Xie HB, et al. Reconstruction of under-sampled magnetic resonance image based on convolutional neural network[J]. Chin J Magn Reson Imaging, 2018, 9(6): 453-459. DOI: 10.12015/issn.1674-8034.2018.06.010.
[15]
Bustin A, Fuin N, Botnar RM, et al. From compressed-sensing to artificial intelligence-based cardiac MRI reconstruction[J]. Front Cardiovasc Med, 2020, 7: 17. DOI: 10.3389/fcvm.2020.00017.
[16]
Lee D, Lee JG, Ko J, et al. Deep learning in MR image processing[J]. Investig Magn Reson Imaging, 2019, 23(2): 81. DOI: 10.13104/imri.2019.23.2.81.
[17]
Wang XZ, Ma JF, Bhosale P, et al. Novel deep learning-based noise reduction technique for prostate magnetic resonance imaging[J]. Abdom Radiol (NY), 2021, 46(7): 3378-3386. DOI: 10.1007/s00261-021-02964-6.
[18]
Wang K, Tamir JI, de Goyeneche A, et al. High fidelity deep learning-based MRI reconstruction with instance-wise discriminative feature matching loss[J]. Magn Reson Med, 2022, 88(1): 476-491. DOI: 10.1002/mrm.29227.
[19]
Terpstra ML, Maspero M, D'Agata F, et al. Deep learning-based image reconstruction and motion estimation from undersampled radial k-space for real-time MRI-guided radiotherapy[J]. Phys Med Biol, 2020, 65(15): 155015. DOI: 10.1088/1361-6560/ab9358.
[20]
Kim M, Kim HS, Kim HJ, et al. Thin-slice pituitary MRI with deep learning-based reconstruction: diagnostic performance in a postoperative setting[J]. Radiology, 2021, 298(1): 114-122. DOI: 10.1148/radiol.2020200723.
[21]
Naganawa S, Nakamichi R, Ichikawa K, et al. MR imaging of endolymphatic hydrops: utility of iHYDROPS-Mi2 combined with deep learning reconstruction denoising[J]. Magn Reson Med Sci, 2021, 20(3): 272-279. DOI: 10.2463/mrms.mp.2020-0082.
[22]
Shanbhogue K, Tong A, Smereka P, et al. Accelerated single-shot T2-weighted fat-suppressed (FS) MRI of the liver with deep learning-based image reconstruction: qualitative and quantitative comparison of image quality with conventional T2-weighted FS sequence[J]. Eur Radiol, 2021, 31(11): 8447-8457. DOI: 10.1007/s00330-021-08008-3.
[23]
Peterfy CG, Guermazi A, Zaim S, et al. Whole-organ magnetic resonance imaging score (WORMS) of the knee in osteoarthritis[J]. Osteoarthritis Cartilage, 2004, 12(3): 177-190. DOI: 10.1016/j.joca.2003.11.003.
[24]
Iuga AI, Abdullayev N, Weiss K, et al. Accelerated MRI of the knee. Quality and efficiency of compressed sensing[J]. Eur J Radiol, 2020, 132: 109273. DOI: 10.1016/j.ejrad.2020.109273.
[25]
Bratke G, Rau R, Weiss K, et al. Accelerated MRI of the lumbar spine using compressed sensing: quality and efficiency[J]. J Magn Reson Imaging, 2019, 49(7): e164-e175. DOI: 10.1002/jmri.26526.
[26]
Wang SY, Liu AL, Wang JZ, et al. Study on the influence of compressed sense technique with different acceleration factors on the image quality of 3D modified-Dixon sequence of liver[J]. Radiol Pract, 2020, 35(12): 1605-1609. DOI: 10.13609/j.cnki.1000-0313.2020.12.021.
[27]
Ueda T, Ohno Y, Yamamoto K, et al. Compressed sensing and deep learning reconstruction for women's pelvic MRI denoising: utility for improving image quality and examination time in routine clinical practice[J]. Eur J Radiol, 2021, 134: 109430. DOI: 10.1016/j.ejrad.2020.109430.
[28]
Gassenmaier S, Afat S, Nickel D, et al. Deep learning-accelerated T2-weighted imaging of the prostate: reduction of acquisition time and improvement of image quality[J]. Eur J Radiol, 2021, 137: 109600. DOI: 10.1016/j.ejrad.2021.109600.
[29]
Kim EH, Choi MH, Lee YJ, et al. Deep learning-accelerated T2-weighted imaging of the prostate: impact of further acceleration with lower spatial resolution on image quality[J]. Eur J Radiol, 2021, 145: 110012. DOI: 10.1016/j.ejrad.2021.110012.
[30]
Herrmann J, Nickel D, Mugler JP, et al. Development and evaluation of deep learning-accelerated single-breath-hold abdominal HASTE at 3 T using variable refocusing flip angles[J]. Invest Radiol, 2021, 56(10): 645-652. DOI: 10.1097/RLI.0000000000000785.
[31]
Recht MP, Zbontar J, Sodickson DK, et al. Using deep learning to accelerate knee MRI at 3 T: results of an interchangeability study[J]. AJR Am J Roentgenol, 2020, 215(6): 1421-1429. DOI: 10.2214/AJR.20.23313.

PREV The effect of different compressed sensing factors on analysis of "swallow tail sign" in cerebral SWI
NEXT The neuromechanism of life satisfaction: Evidence from MRI studies
  



Tel & Fax: +8610-67113815    E-mail: editor@cjmri.cn