Deep Learning Super Resolution Reconstruction for Fast and Motion Robust T2-weighted Prostate MRI

Prostate Cancer Clinical Trial

Official title:

Deep Learning Super Resolution Reconstruction for Fast and Motion Robust T2-weighted Prostate MRI

Clinical Trial Details — Status: Completed

Administrative data

• NCT number: NCT05820113
• Other study ID #: SR.001
• Status: Completed
• Phase: N/A
• Start date: August 1, 2022
• Est. completion date: November 30, 2022

Study information

• Verified date: April 2023
• Source: University Hospital, Bonn
• Contact: n/a
• Is FDA regulated: No
• Study type: Interventional

Clinical Trial Summary

The aim of this study was therefore to investigate a new unrolled DL super resolution reconstruction of an initially low-resolution Cartesian T2 turbo spin echo sequence (T2 TSE) and compare it qualitatively and quantitatively to standard high-resolution Cartesian and non-Cartesian T2 TSE sequences in the setting of prostate mpMRI with particular interest in image sharpness, conspicuity of lesions and acquisition time. Furthermore, the investigators assessed the agreement of assigned PI-RADS scores between deep learning super resolution and standard sequences.

Description:

Prostate cancer has been among the most prevalent cancer types in men for years, being responsible for 7.8% of all newly diagnosed cases in 2020, holding the 2th place right after lung cancer. Early and non-invasive diagnostics was improved vastly by multiparametric MRI (mpMRI) of the prostate, detecting clinically significant prostate cancer and forming the baseline for guided biopsy of the prostate, while it can prevent unnecessary biopsies in patients with elevated prostate specific antigen, but no visible lesions. With an aging population fast, efficient and highly qualitative MRI scans are needed to satisfy this increasing demand. Deep learning image reconstruction has become increasingly important solving these tasks to produce highly qualitative images while drastically reducing acquisition time. Standard acquisition protocols of prostate mpMRI include T2-weighted, diffusion-weighted and dynamically contrast-enhanced sequences to allow for the classification of prostatic lesions according to the Prostate Imaging Reporting & Data System (PI-RADS). While the assignment of the PI-RADS score in the peripheral zone of the prostate is mainly determined by the diffusion weighted imaging, the T2-weighted-sequences are mainly responsible for the assessment of the transitional zone. Furthermore, thorough assessment of the prostate necessitates acquisition of T2-weighted sequences in axial and sagittal planes, thus extending acquisition time of MRI protocols. Different approaches have been proposed to accelerate and improve the image acquisition, ranging from the implementation of shortened protocols to the improvement of diffusion weighted sequences or using compressed sensing for the reconstruction of non-Cartesian T2-weighted-sequences. Besides these methods that rely on traditional acquisition and reconstruction methods, deep learning (DL) image reconstruction has become increasingly important solving these tasks to produce highly qualitative images while drastically reducing acquisition time. Despite that, reliable DL-methods for the process of image acquisition and reconstruction of prostate mpMRI itself are sparse. While first approaches for DL denoising has been established, effectively replacing the conventional wavelet function, the remainder of the iterative reconstruction cycle is unaffected and the impact on diagnostic performance of the PI-RADS score remains unclear. Recently developed super resolution deep learning networks are promising to overcome this limitation. First results for DL denoising in different applications, e.g. in musculoskeletal MRI already show good results, leading to significant acceleration of acquisition time while maintaining high image quality. However, the application of these denoising DL-networks in combination with more advanced super resolution networks in prostate mpMRI hasn't been evaluated yet. In this prospective study, between August and November 2022, participants with suspicion for prostate cancer underwent prostate MRI with standard high-resolution Cartesian T2 (T2C) and non-Cartesian T2 (T2NC) sequences. Additionally, a low-resolution Cartesian T2 TSE (T2SR) with DL denoising and super resolution reconstruction was acquired. Artifacts, image sharpness, lesion conspicuity, capsule delineation, overall image quality and diagnostic confidence were rated on a 5-point-Likert-Scale with being non-diagnostic and 5 being excellent. Apparent signal-to-noise ratio (aSNR), contrast-to-noise ratio (aCNR) and edge rise distance (ERD) were calculated. Friedman test and One-way ANOVA were used for group comparisons. Regarding agreement of PI-RADS scores were compared with Cohen's Kappa.

Recruitment information / eligibility

• Status: Completed
• Enrollment: 109
• Est. completion date: November 30, 2022
• Est. primary completion date: November 30, 2022
• Accepts healthy volunteers: No
• Gender: Male
• Age group: 18 Years and older

Eligibility

Inclusion Criteria:

• Clinical suspicion of prostate cancer (PSA >4 ng/ml or suspicious digital rectal exam/transrectal ultrasound)

Exclusion Criteria:

• General contraindications for MRI (cardiac pacemakers, neurostimulators, ferric metal) or gadolinium based contrast agents (GFR <30 ml/min/1.73 m2, prior severe allergic reactions)
• Severe claustrophobia

Related Conditions & MeSH terms

• Body Weight
• Multiparametric MRI
• Prostate Cancer
• Prostatic Neoplasms

Intervention

Device:
• Deep learning based reconstruction of T2-TSE sequence
A newly developed deep-learning based reconstruction of a primarily low-resolved T2-TSE sequence is included in the imaging protocol for evaluation of prostate cancer.

Locations

Country	Name	City	State
Germany	University Hospital Bonn	Bonn	NRW

Sponsors (2)

Lead Sponsor	Collaborator
University Hospital, Bonn	Philips Healthcare

Country where clinical trial is conducted

Germany, 

References & Publications (24)

11. Li Y, Sixou B, Peyrin F. A Review of the Deep Learning Methods for Medical Images Super Resolution Problems. IRBM, Volume 42, Issue 2, April 2021, Pages 120-133. doi: 10.1016/j.irbm.2020.08.004

ACR, ESUR and AdMeTech Foundation. Prostate Imaging Reporting & Data System (PI-RADS). 2019. Version 2.1.

Agarwal S, Singh O.P, Nagaria D. Analysis and Comparison of Wavelet Transforms for Denoising MRI Image. Biomed Pharmacol J 2017;10(2).

Almansour H, Herrmann J, Gassenmaier S, Afat S, Jacoby J, Koerzdoerfer G, Nickel D, Mostapha M, Nadar M, Othman AE. Deep Learning Reconstruction for Accelerated Spine MRI: Prospective Analysis of Interchangeability. Radiology. 2023 Mar;306(3):e212922. doi: 10.1148/radiol.212922. Epub 2022 Nov 1. — View Citation

Bischoff LM, Katemann C, Isaak A, Mesropyan N, Wichtmann B, Kravchenko D, Endler C, Kuetting D, Pieper CC, Ellinger J, Weber O, Attenberger U, Luetkens JA. T2 Turbo Spin Echo With Compressed Sensing and Propeller Acquisition (Sampling k-Space by Utilizing Rotating Blades) for Fast and Motion Robust Prostate MRI: Comparison With Conventional Acquisition. Invest Radiol. 2023 Mar 1;58(3):209-215. doi: 10.1097/RLI.0000000000000923. Epub 2022 Sep 2. — View Citation

Chaudhari AS, Fang Z, Kogan F, Wood J, Stevens KJ, Gibbons EK, Lee JH, Gold GE, Hargreaves BA. Super-resolution musculoskeletal MRI using deep learning. Magn Reson Med. 2018 Nov;80(5):2139-2154. doi: 10.1002/mrm.27178. Epub 2018 Mar 26. — View Citation

Dong C, Loy CC, He K, Tang X. Image Super-Resolution Using Deep Convolutional Networks. IEEE Trans Pattern Anal Mach Intell. 2016 Feb;38(2):295-307. doi: 10.1109/TPAMI.2015.2439281. — View Citation

Eklund M, Jaderling F, Discacciati A, Bergman M, Annerstedt M, Aly M, Glaessgen A, Carlsson S, Gronberg H, Nordstrom T; STHLM3 consortium. MRI-Targeted or Standard Biopsy in Prostate Cancer Screening. N Engl J Med. 2021 Sep 2;385(10):908-920. doi: 10.1056/NEJMoa2100852. Epub 2021 Jul 9. — View Citation

Gassenmaier S, Warm V, Nickel D, Weiland E, Herrmann J, Almansour H, Wessling D, Afat S. Thin-Slice Prostate MRI Enabled by Deep Learning Image Reconstruction. Cancers (Basel). 2023 Jan 18;15(3):578. doi: 10.3390/cancers15030578. — View Citation

Harder FN, Weiss K, Amiel T, Peeters JM, Tauber R, Ziegelmayer S, Burian E, Makowski MR, Sauter AP, Gschwend JE, Karampinos DC, Braren RF. Prospectively Accelerated T2-Weighted Imaging of the Prostate by Combining Compressed SENSE and Deep Learning in Patients with Histologically Proven Prostate Cancer. Cancers (Basel). 2022 Nov 22;14(23):5741. doi: 10.3390/cancers14235741. — View Citation

Hosny A, Parmar C, Quackenbush J, Schwartz LH, Aerts HJWL. Artificial intelligence in radiology. Nat Rev Cancer. 2018 Aug;18(8):500-510. doi: 10.1038/s41568-018-0016-5. — View Citation

Hugosson J, Mansson M, Wallstrom J, Axcrona U, Carlsson SV, Egevad L, Geterud K, Khatami A, Kohestani K, Pihl CG, Socratous A, Stranne J, Godtman RA, Hellstrom M; GOTEBORG-2 Trial Investigators. Prostate Cancer Screening with PSA and MRI Followed by Targeted Biopsy Only. N Engl J Med. 2022 Dec 8;387(23):2126-2137. doi: 10.1056/NEJMoa2209454. — View Citation

Johnson PM, Tong A, Donthireddy A, Melamud K, Petrocelli R, Smereka P, Qian K, Keerthivasan MB, Chandarana H, Knoll F. Deep Learning Reconstruction Enables Highly Accelerated Biparametric MR Imaging of the Prostate. J Magn Reson Imaging. 2022 Jul;56(1):184-195. doi: 10.1002/jmri.28024. Epub 2021 Dec 7. — View Citation

Kim J, Lee JK, Lee KM. Accurate Image Super-Resolution Using Very Deep Convolutional Networks. arXiv:1511.04587

Kim M, Lee SM, Park C, Lee D, Kim KS, Jeong HS, Kim S, Choi MH, Nickel D. Deep Learning-Enhanced Parallel Imaging and Simultaneous Multislice Acceleration Reconstruction in Knee MRI. Invest Radiol. 2022 Dec 1;57(12):826-833. doi: 10.1097/RLI.0000000000000900. Epub 2022 Jul 1. — View Citation

Langbein BJ, Szczepankiewicz F, Westin CF, Bay C, Maier SE, Kibel AS, Tempany CM, Fennessy FM. A Pilot Study of Multidimensional Diffusion MRI for Assessment of Tissue Heterogeneity in Prostate Cancer. Invest Radiol. 2021 Dec 1;56(12):845-853. doi: 10.1097/RLI.0000000000000796. — View Citation

Pezzotti N, de Weerdt E, Yousefi S, et al. Adaptive-CS-Net: FastMRI with Adaptive Intelligence. arxiv:1912.12259. 2019;(NeurIPS).

Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, Bray F. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin. 2021 May;71(3):209-249. doi: 10.3322/caac.21660. Epub 2021 Feb 4. — View Citation

Suzuki S, Machida H, Tanaka I, Ueno E. Measurement of vascular wall attenuation: comparison of CT angiography using model-based iterative reconstruction with standard filtered back-projection algorithm CT in vitro. Eur J Radiol. 2012 Nov;81(11):3348-53. doi: 10.1016/j.ejrad.2012.02.009. Epub 2012 Mar 19. — View Citation

Ueda T, Ohno Y, Yamamoto K, Murayama K, Ikedo M, Yui M, Hanamatsu S, Tanaka Y, Obama Y, Ikeda H, Toyama H. Deep Learning Reconstruction of Diffusion-weighted MRI Improves Image Quality for Prostatic Imaging. Radiology. 2022 May;303(2):373-381. doi: 10.1148/radiol.204097. Epub 2022 Feb 1. — View Citation

Weiss J, Martirosian P, Notohamiprodjo M, Kaufmann S, Othman AE, Grosse U, Nikolaou K, Gatidis S. Implementation of a 5-Minute Magnetic Resonance Imaging Screening Protocol for Prostate Cancer in Men With Elevated Prostate-Specific Antigen Before Biopsy. Invest Radiol. 2018 Mar;53(3):186-190. doi: 10.1097/RLI.0000000000000427. — View Citation

Wright KL, Hamilton JI, Griswold MA, Gulani V, Seiberlich N. Non-Cartesian parallel imaging reconstruction. J Magn Reson Imaging. 2014 Nov;40(5):1022-40. doi: 10.1002/jmri.24521. Epub 2014 Jan 10. — View Citation

Zaitsev M, Maclaren J, Herbst M. Motion artifacts in MRI: A complex problem with many partial solutions. J Magn Reson Imaging. 2015 Oct;42(4):887-901. doi: 10.1002/jmri.24850. Epub 2015 Jan 28. — View Citation

Zhang J and Ghanem B. ISTA-Net: Interpretable optimization-inspired deep network for image compressive sensing. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2018, pp. 1828-1837

* Note: There are 24 references in all — Click here to view all references

Outcome

Type	Measure	Description	Time frame	Safety issue
Primary	Qualitative assessment of image quality (Artifacts, image sharpness, lesion conspicuity, capsule delineation, overall image quality and diagnostic confidence)	Artifacts, image sharpness, lesion conspicuity, capsule delineation, overall image quality and diagnostic confidence were rated on a 5-point-Likert-Scale with 1 being non-diagnostic and 5 being excellent. Friedman test was used for significance testing with p<0.05 considered as indicative of a significant difference.	4 months
Primary	Acquisition time	Measurement of acquisition time of T2-weighted sequences	4 months
Primary	Degree of agreement on PI-RADS ratings	To assess the PI-RADS score, all MRIs were read blinded by a radiologist with 11 years expertise at two different time points in random order. The MRI sequences for PI-RADS assessment included either T2NC (reference standard at our institution) or T2SR as the T2-weighted sequence in the reading protocol. The remainder of sequences were the same (axial T1-weighted TSE pre and post contrast administration, axial dynamically contrast enhanced T1, sagittal T2 TSE and axial diffusion weighted sequences with apparent diffusion coefficient map). Cohen's Kappa was used for correlation of readings with inclusion of either T2SR or T2NC.	4 months
Primary	Quantitative assessment of image quality (apparent signal-to-noise and contrast-to-noise ratio)	Apparent signal-to-noise ratio (aSNR: signal intensity of peripheral zone/standard deviation of muscle) and contrast-to-noise ratio (aCNR: signal intensity of peripheral zone - signal intensity of muscle)/standard deviation of muscle) was calculated to quantify the image sharpness. One-way ANOVA was used for significance testing with p<0.05 considered as indicative of a significant difference.	4 months
Primary	Quantitative assessment of image quality (edge rise distance)	Edge rise distance (ERD) was calculated to quantify the image sharpness. The ERD was determined as a measure of image sharpness. For this purpose, a line was drawn perpendicularly crossing the dorsal border of the prostate capsule. The edge rise distance was then determined as the distance (in mm) between the 10% and 90% signal intensity levels relative to the low and high signal intensity areas. One-way ANOVA was used for significance testing with p<0.05 considered as indicative of a significant difference.	4 months