A deep learning methodology for the automated detection of end-diastolic frames in intravascular ultrasound images.
Intravascular ultrasound
Machine learning
Near-infrared spectroscopy
Journal
The international journal of cardiovascular imaging
ISSN: 1875-8312
Titre abrégé: Int J Cardiovasc Imaging
Pays: United States
ID NLM: 100969716
Informations de publication
Date de publication:
Jun 2021
Jun 2021
Historique:
received:
06
11
2020
accepted:
07
01
2021
pubmed:
17
2
2021
medline:
16
10
2021
entrez:
16
2
2021
Statut:
ppublish
Résumé
Coronary luminal dimensions change during the cardiac cycle. However, contemporary volumetric intravascular ultrasound (IVUS) analysis is performed in non-gated images as existing methods to acquire gated or to retrospectively gate IVUS images have failed to dominate in research. We developed a novel deep learning (DL)-methodology for end-diastolic frame detection in IVUS and compared its efficacy against expert analysts and a previously established methodology using electrocardiographic (ECG)-estimations as reference standard. Near-infrared spectroscopy-IVUS (NIRS-IVUS) data were prospectively acquired from 20 coronary arteries and co-registered with the concurrent ECG-signal to identify end-diastolic frames. A DL-methodology which takes advantage of changes in intensity of corresponding pixels in consecutive NIRS-IVUS frames and consists of a network model designed in a bidirectional gated-recurrent-unit (Bi-GRU) structure was trained to detect end-diastolic frames. The efficacy of the DL-methodology in identifying end-diastolic frames was compared with two expert analysts and a conventional image-based (CIB)-methodology that relies on detecting vessel movement to estimate phases of the cardiac cycle. A window of ± 100 ms from the ECG estimations was used to define accurate end-diastolic frames detection. The ECG-signal identified 3,167 end-diastolic frames. The mean difference between DL and ECG estimations was 3 ± 112 ms while the mean differences between the 1st-analyst and ECG, 2nd-analyst and ECG and CIB-methodology and ECG were 86 ± 192 ms, 78 ± 183 ms and 59 ± 207 ms, respectively. The DL-methodology was able to accurately detect 80.4%, while the two analysts and the CIB-methodology detected 39.0%, 43.4% and 42.8% of end-diastolic frames, respectively (P < 0.05). The DL-methodology can identify NIRS-IVUS end-diastolic frames accurately and should be preferred over expert analysts and CIB-methodologies, which have limited efficacy.
Identifiants
pubmed: 33590430
doi: 10.1007/s10554-021-02162-x
pii: 10.1007/s10554-021-02162-x
pmc: PMC8255253
doi:
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
1825-1837Subventions
Organisme : UCLH Biomedical Research Centre
ID : BRC492B
Organisme : Rosetrees Trust
ID : A1773
Organisme : British Heart Foundation
ID : PG/17/18/32883
Pays : United Kingdom
Références
Böse D, von Birgelen C, Erbel R (2007) Intravascular ultrasound for the evaluation of therapies targeting coronary atherosclerosis. J Am Coll Cardiol 49:925–932. https://doi.org/10.1016/j.jacc.2006.08.067
doi: 10.1016/j.jacc.2006.08.067
pubmed: 17336714
Mintz G, Garcia-Garcia H, Nicholls S et al (2011) Clinical expert consensus document on standards for acquisition, measurement and reporting of intravascular ultrasound regression/progression studies. EuroIntervention 6:1123–1130. https://doi.org/10.4244/EIJV6I9A195
doi: 10.4244/EIJV6I9A195
pubmed: 21518687
Arbab-Zadeh A, DeMaria AN, Penny WF et al (1999) Axial movement of the intravascular ultrasound probe during the cardiac cycle: implications for three-dimensional reconstruction and measurements of coronary dimensions. Am Heart J 138:865–872. https://doi.org/10.1016/S0002-8703(99)70011-6
doi: 10.1016/S0002-8703(99)70011-6
pubmed: 10539817
Weissman NJ, Palacios IF, Weyman AE (1995) Dynamic expansion of the coronary arteries: implications for intravascular ultrasound measurements. Am Heart J 130:46–51. https://doi.org/10.1016/0002-8703(95)90234-1
doi: 10.1016/0002-8703(95)90234-1
pubmed: 7611122
Ge J, Erbel R, Gerber T et al (1994) Intravascular ultrasound imaging of angiographically normal coronary arteries: a prospective study in vivo. Heart 71:572–578. https://doi.org/10.1136/hrt.71.6.572
doi: 10.1136/hrt.71.6.572
Bruining N, von Birgelen C, de Feyter PJ et al (1998) ECG-gated versus nongated three-dimensional intracoronary ultrasound analysis: implications for volumetric measurements. Cathet Cardiovasc Diagn 43:254–260. https://doi.org/10.1002/(SICI)1097-0304(199803)43:3%3c254::AID-CCD3%3e3.0.CO;2-8
doi: 10.1002/(SICI)1097-0304(199803)43:3<254::AID-CCD3>3.0.CO;2-8
pubmed: 9535359
de Winter S, Hamers R, Roelandt J et al (2009) Quantitative gated intravascular ultrasound largely reduces the population size for atherosclerosis progression-regression trials: a computer simulation study. Comput Cardiol 36:829–832
von Birgelen C, de Vrey EA, Mintz GS et al (1997) ECG-gated three-dimensional intravascular ultrasound: feasibility and reproducibility of the automated analysis of coronary lumen and atherosclerotic plaque dimensions in humans. Circulation 96:2944–2952. https://doi.org/10.1161/01.CIR.96.9.2944
doi: 10.1161/01.CIR.96.9.2944
de Winter SA, Hamers R, Degertekin M et al (2004) Retrospective image-based gating of intracoronary ultrasound images for improved quantitative analysis: the intelligate method. Cathet Cardiovasc Interv 61:84–94. https://doi.org/10.1002/ccd.10693
doi: 10.1002/ccd.10693
O’Malley SM, Carlier SG, Naghavi M, Kakadiaris IA (2007) Image-based frame gating of IVUS pullbacks: a surrogate for ECG. In: 2007 IEEE International Conference on Acoustics, Speech and Signal Processing—ICASSP ’07. IEEE, Honolulu, HI, USA, pp. I-433-I–436. https://doi.org/10.1109/ICASSP.2007.366709
Gatta C, Pujol O, Leor OR et al (2008) Robust image-based IVUS pullbacks gating. In: Metaxas D, Axel L, Fichtinger G, Székely G (eds) Medical image computing and computer-assisted intervention—MICCAI 2008. Springer, Berlin, Heidelberg, pp 518–525
doi: 10.1007/978-3-540-85990-1_62
Hutchison D, Kanade T, Kittler J, et al (2010) Manifold learning for image-based gating of intravascular ultrasound(IVUS) pullback sequences. In: Liao H, Edwards PJ “Eddie,” Pan X, et al (eds) Medical imaging and augmented reality. Springer, Berlin, Heidelberg. pp. 139–148. https://doi.org/10.1007/978-3-642-15699-1_15
Maso Talou GD, Larrabide I, Blanco PJ et al (2015) Improving cardiac phase extraction in IVUS studies by integration of gating methods. IEEE Trans Biomed Eng 62:2867–2877. https://doi.org/10.1109/TBME.2015.2449232
doi: 10.1109/TBME.2015.2449232
pubmed: 26111388
Nadkarni SK, Boughner D, Fenster A (2005) Image-based cardiac gating for three-dimensional intravascular ultrasound imaging. Ultrasound Med Biol 31:53–63. https://doi.org/10.1016/j.ultrasmedbio.2004.08.025
doi: 10.1016/j.ultrasmedbio.2004.08.025
pubmed: 15653231
Torbati N, Ayatollahi A, Sadeghipour P (2019) Image-based gating of intravascular ultrasound sequences using the phase information of dual-tree complex wavelet transform coefficients. IEEE Trans Med Imaging 38:2785–2795. https://doi.org/10.1109/TMI.2019.2914074
doi: 10.1109/TMI.2019.2914074
pubmed: 31056492
Hernàndez-Sabaté A, Gil D, Garcia-Barnés J, Martí E (2011) Image-based cardiac phase retrieval in intravascular ultrasound sequences. IEEE Trans Ultrason Ferroelectr Freq Control 58:60–72. https://doi.org/10.1109/TUFFC.2011.1774
doi: 10.1109/TUFFC.2011.1774
pubmed: 21244975
Guo D, Richardson P (2006) Detection of cardiac cycle from intracoronary ultrasound. Ultrasound Med Biol 32:345–356. https://doi.org/10.1016/j.ultrasmedbio.2005.12.011
doi: 10.1016/j.ultrasmedbio.2005.12.011
pubmed: 16530093
Al’Aref SJ, Anchouche K, Singh G et al (2019) Clinical applications of machine learning in cardiovascular disease and its relevance to cardiac imaging. Eur Heart J 40:1975–1986. https://doi.org/10.1093/eurheartj/ehy404
doi: 10.1093/eurheartj/ehy404
pubmed: 30060039
Bae Y, Kang S-J, Kim G et al (2019) Prediction of coronary thin-cap fibroatheroma by intravascular ultrasound-based machine learning. Atherosclerosis 288:168–174. https://doi.org/10.1016/j.atherosclerosis.2019.04.228
doi: 10.1016/j.atherosclerosis.2019.04.228
pubmed: 31130215
Min H-S, Yoo JH, Kang S-J et al (2019) Detection of optical coherence tomography-defined thin-cap fibroatheroma in the coronary artery using deep learning. EuroIntervention. https://doi.org/10.4244/EIJ-D-19-00487
doi: 10.4244/EIJ-D-19-00487
Xia M, Yan W, Huang Y, et al (2020) Extracting membrane borders in IVUS images using a multi-scale feature aggregated U-Net. In: 2020 42nd Annual international conference of the IEEE Engineering in Medicine & Biology Society (EMBC). IEEE, Montreal, QC, Canada, pp. 1650–1653. https://doi.org/10.1109/EMBC44109.2020.9175970
Mishra D, Chaudhury S, Sarkar M, Soin AS (2019) Ultrasound image segmentation: a deeply supervised network with attention to boundaries. IEEE Trans Biomed Eng 66:1637–1648. https://doi.org/10.1109/TBME.2018.2877577
doi: 10.1109/TBME.2018.2877577
pubmed: 30346279
Gao Z, Chung J, Abdelrazek M et al (2020) Privileged modality distillation for vessel border detection in intracoronary imaging. IEEE Trans Med Imaging 39:1524–1534. https://doi.org/10.1109/TMI.2019.2952939
doi: 10.1109/TMI.2019.2952939
pubmed: 31715563
Lo Vercio L, del Fresno M, Larrabide I (2019) Lumen-intima and media-adventitia segmentation in IVUS images using supervised classifications of arterial layers and morphological structures. Comput Methods Programs Biomed 177:113–121. https://doi.org/10.1016/j.cmpb.2019.05.021
doi: 10.1016/j.cmpb.2019.05.021
pubmed: 31319939
Ramasamy A, Safi H, Moon JC et al (2020) Evaluation of the efficacy of computed tomographic coronary angiography in assessing coronary artery morphology and physiology: rationale and study design. Cardiology 145:285–293. https://doi.org/10.1159/000506537
doi: 10.1159/000506537
pubmed: 32289784
Schuster M, Paliwal KK (1997) Bidirectional recurrent neural networks. IEEE Trans Signal Process 45:2673–2681. https://doi.org/10.1109/78.650093
doi: 10.1109/78.650093
Talou GDM, Blanco PJ, Larrabide I et al (2017) Registration methods for IVUS: transversal and longitudinal transducer motion compensation. IEEE Trans Biomed Eng 64:890–903. https://doi.org/10.1109/TBME.2016.2581583
doi: 10.1109/TBME.2016.2581583
pubmed: 27323357
Jensen LO, Thayssen P (2007) Accuracy of electrocardiographic-gated versus nongated volumetric intravascular ultrasound measurements of coronary arterial narrowing. Am J Cardiol 99:279–283. https://doi.org/10.1016/j.amjcard.2006.08.024
doi: 10.1016/j.amjcard.2006.08.024
pubmed: 17223434
Bourantas CV, Jaffer FA, Gijsen FJ et al (2017) Hybrid intravascular imaging: recent advances, technical considerations, and current applications in the study of plaque pathophysiology. Eur Heart J 38:400–412. https://doi.org/10.1093/eurheartj/ehw097
doi: 10.1093/eurheartj/ehw097
pubmed: 27118197
Bourantas CV, Räber L, Sakellarios A et al (2020) Utility of multimodality intravascular imaging and the local hemodynamic forces to predict atherosclerotic disease progression. JACC Cardiovasc Imaging 13:1021–1032. https://doi.org/10.1016/j.jcmg.2019.02.026
doi: 10.1016/j.jcmg.2019.02.026
pubmed: 31202749
Stone PH, Saito S, Takahashi S et al (2012) Prediction of progression of coronary artery disease and clinical outcomes using vascular profiling of endothelial shear stress and arterial plaque characteristics: the prediction study. Circulation 126:172–181. https://doi.org/10.1161/CIRCULATIONAHA.112.096438
doi: 10.1161/CIRCULATIONAHA.112.096438
pubmed: 22723305
Guo X, Giddens DP, Molony D et al (2018) Combining IVUS and optical coherence tomography for more accurate coronary cap thickness quantification and stress/strain calculations: a patient-specific three-dimensional fluid-structure interaction modeling approach. J Biomech Eng 140:041005. https://doi.org/10.1115/1.4038263
doi: 10.1115/1.4038263
Browne MW (2000) Cross-validation methods. J Math Psychol 44:108–132. https://doi.org/10.1006/jmps.1999.1279
doi: 10.1006/jmps.1999.1279
pubmed: 10733860