1
|
Sabouri M, Hajianfar G, Hosseini Z, Amini M, Mohebi M, Ghaedian T, Madadi S, Rastgou F, Oveisi M, Bitarafan Rajabi A, Shiri I, Zaidi H. Myocardial Perfusion SPECT Imaging Radiomic Features and Machine Learning Algorithms for Cardiac Contractile Pattern Recognition. J Digit Imaging 2023; 36:497-509. [PMID: 36376780 PMCID: PMC10039187 DOI: 10.1007/s10278-022-00705-9] [Citation(s) in RCA: 13] [Impact Index Per Article: 13.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/21/2022] [Revised: 08/31/2022] [Accepted: 09/15/2022] [Indexed: 11/16/2022] Open
Abstract
A U-shaped contraction pattern was shown to be associated with a better Cardiac resynchronization therapy (CRT) response. The main goal of this study is to automatically recognize left ventricular contractile patterns using machine learning algorithms trained on conventional quantitative features (ConQuaFea) and radiomic features extracted from Gated single-photon emission computed tomography myocardial perfusion imaging (GSPECT MPI). Among 98 patients with standard resting GSPECT MPI included in this study, 29 received CRT therapy and 69 did not (also had CRT inclusion criteria but did not receive treatment yet at the time of data collection, or refused treatment). A total of 69 non-CRT patients were employed for training, and the 29 were employed for testing. The models were built utilizing features from three distinct feature sets (ConQuaFea, radiomics, and ConQuaFea + radiomics (combined)), which were chosen using Recursive feature elimination (RFE) feature selection (FS), and then trained using seven different machine learning (ML) classifiers. In addition, CRT outcome prediction was assessed by different treatment inclusion criteria as the study's final phase. The MLP classifier had the highest performance among ConQuaFea models (AUC, SEN, SPE = 0.80, 0.85, 0.76). RF achieved the best performance in terms of AUC, SEN, and SPE with values of 0.65, 0.62, and 0.68, respectively, among radiomic models. GB and RF approaches achieved the best AUC, SEN, and SPE values of 0.78, 0.92, and 0.63 and 0.74, 0.93, and 0.56, respectively, among the combined models. A promising outcome was obtained when using radiomic and ConQuaFea from GSPECT MPI to detect left ventricular contractile patterns by machine learning.
Collapse
Affiliation(s)
- Maziar Sabouri
- Department of Medical Physics, School of Medicine, Iran University of Medical Science, Tehran, Iran
- Rajaie Cardiovascular Medical and Research Center, Iran University of Medical Science, Tehran, Iran
| | - Ghasem Hajianfar
- Rajaie Cardiovascular Medical and Research Center, Iran University of Medical Science, Tehran, Iran
| | - Zahra Hosseini
- Rajaie Cardiovascular Medical and Research Center, Iran University of Medical Science, Tehran, Iran
| | - Mehdi Amini
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, CH-1211, Geneva 4, Switzerland
| | - Mobin Mohebi
- Department of Biomedical Engineering, Tarbiat Modares University, Tehran, Iran
| | - Tahereh Ghaedian
- Nuclear Medicine and Molecular Imaging Research Center, School of Medicine, Namazi Teaching Hospital, Shiraz University of Medical Sciences, Shiraz, Iran
| | - Shabnam Madadi
- Rajaie Cardiovascular Medical and Research Center, Iran University of Medical Science, Tehran, Iran
| | - Fereydoon Rastgou
- Rajaie Cardiovascular Medical and Research Center, Iran University of Medical Science, Tehran, Iran
| | - Mehrdad Oveisi
- Comprehensive Cancer Centre, School of Cancer & Pharmaceutical Sciences, Faculty of Life Sciences & Medicine, King's College London, London, UK
- Department of Computer Science, University of British Columbia, Vancouver BC, Canada
| | - Ahmad Bitarafan Rajabi
- Echocardiography Research Center, Rajaie Cardiovascular Medical and Research Center, Iran University of Medical Sciences, Tehran, Iran.
- Cardiovascular Interventional Research Center, Rajaie Cardiovascular Medical and Research Center, Iran University of Medical Sciences, Tehran, Iran.
| | - Isaac Shiri
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, CH-1211, Geneva 4, Switzerland.
| | - Habib Zaidi
- Division of Nuclear Medicine and Molecular Imaging, Geneva University Hospital, CH-1211, Geneva 4, Switzerland.
- Geneva University Neurocenter, Geneva University, Geneva, Switzerland.
- Department of Nuclear Medicine and Molecular Imaging, University of Groningen, University Medical Center Groningen, Groningen, Netherlands.
- Department of Nuclear Medicine, University of Southern Denmark, Odense, Denmark.
| |
Collapse
|
2
|
Sillanmäki S, Gimelli A, Ahmad S, Samir S, Laitinen T, Soman P. Mechanisms of left ventricular dyssynchrony: A multinational SPECT study of patients with bundle branch block. J Nucl Cardiol 2021; 28:1140-1150. [PMID: 32060855 DOI: 10.1007/s12350-020-02054-y] [Citation(s) in RCA: 3] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/03/2019] [Accepted: 01/16/2020] [Indexed: 10/25/2022]
Abstract
BACKGROUND To better understand the mechanisms of left ventricular (LV) mechanical dyssynchrony (LVMD), we explored the relative contributions of QRS duration (QRSd), LV ejection fraction (EF), volumes and scar to LVMD measured by gated single-photon emission tomography in a population of consecutive patients with left bundle branch block (LBBB) and right bundle branch block (RBBB) compared to controls. METHODS Myocardial perfusion imaging studies of 275 LBBB and 83 RBBB patients from three centers were analyzed. LVMD was defined as an abnormal phase bandwidth or phase standard deviation. Hospital and gender-specific normal values were obtained from 172 controls. RESULTS The prevalence of LVMD was 85 and 40% in LBBB and RBBB, respectively. Ejection fraction, scar severity, and LBBB morphology independently explained 70% of variance seen in PhaseBW. Ejection fraction had the highest area under the curve (AUC 0.918) in the receiver operating characteristics analysis of LVMD with an optimal cut-off of 47% (sensitivity 73% and specificity 98%). Notably, QRSd was not predictive. CONCLUSION LV mechanical dysfunction plays a greater role than conduction abnormality in the genesis of LVMD, a finding that is intriguing in the context of contemporary literature which suggests that QRSd is the parameter that is most predictive of CRT response.
Collapse
Affiliation(s)
- Saara Sillanmäki
- Institute of Clinical Medicine, University of Eastern Finland, Joensuu, Finland
- Department of Nuclear Medicine and Clinical Physiology, Kuopio University Hospital, PL 100, 70029 KYS, Kuopio, Finland
| | | | - Shahzad Ahmad
- Division of Cardiology and The Heart and Vascular Institute, University of Pittsburgh Medical Center, Pittsburgh, PA, USA
| | - Saba Samir
- Division of Cardiology and The Heart and Vascular Institute, University of Pittsburgh Medical Center, Pittsburgh, PA, USA
| | - Tomi Laitinen
- Institute of Clinical Medicine, University of Eastern Finland, Joensuu, Finland
- Department of Nuclear Medicine and Clinical Physiology, Kuopio University Hospital, PL 100, 70029 KYS, Kuopio, Finland
| | - Prem Soman
- Division of Cardiology and The Heart and Vascular Institute, University of Pittsburgh Medical Center, Pittsburgh, PA, USA.
- University of Pittsburgh Medical Center, A429 Scaife Hall, 200 Lothrop Street, Pittsburgh, PA, 15213, USA.
| |
Collapse
|
3
|
Sassone B, Nucifora G, Mele D, Valzania C, Bisignani G, Boriani G. Role of cardiovascular imaging in cardiac resynchronization therapy: a literature review. J Cardiovasc Med (Hagerstown) 2018; 19:211-222. [PMID: 29470248 DOI: 10.2459/jcm.0000000000000635] [Citation(s) in RCA: 12] [Impact Index Per Article: 2.0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 02/06/2023]
Abstract
: Cardiac resynchronization therapy (CRT) is an established treatment in patients with symptomatic drug-refractory heart failure and broad QRS complex on the surface ECG. Despite the presence of either mechanical dyssynchrony or viable myocardium at the site where delivering left ventricular pacing being necessary conditions for a successful CRT, their direct assessment by techniques of cardiovascular imaging, though feasible, is not recommended in clinical practice by the current guidelines. Indeed, even though there is growing body of data providing evidence of the additional value of an image-based approach as compared with routine approach in improving response to CRT, these results should be confirmed in prospective and large multicentre trials before their impact on CRT guidelines is considered.
Collapse
Affiliation(s)
- Biagio Sassone
- Department of Cardiology, SS.ma Annunziata Hospital.,Department of Cardiology, Delta Hospital, Azienda Unità Sanitaria Locale Ferrara, Ferrara, Italy
| | - Gaetano Nucifora
- Cardiology Department, Wythenshawe Hospital, Manchester University NHS Foundation Trust, Manchester, UK.,Flinders University, Adelaide, Australia
| | - Donato Mele
- Noninvasive Cardiology Unit, University Hospital of Ferrara, Ferrara
| | - Cinzia Valzania
- Institute of Cardiology, University of Bologna, Policlinico S. Orsola-Malpighi, Bologna
| | | | - Giuseppe Boriani
- Cardiology Division, Department of Diagnostics, Clinical and Public Health Medicine, University of Modena and Reggio Emilia, Policlinico of Modena, Modena, Italy
| | | |
Collapse
|
4
|
Tang H, Tang S, Zhou W. A Review of Image-guided Approaches for Cardiac Resynchronisation Therapy. Arrhythm Electrophysiol Rev 2017; 6:69-74. [PMID: 28845234 DOI: 10.15420/aer.2016.32.2] [Citation(s) in RCA: 3] [Impact Index Per Article: 0.4] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Indexed: 11/04/2022] Open
Abstract
Cardiac resynchronisation therapy (CRT) is a standard treatment for patients with heart failure; however, the low response rate significantly reduces its cost-effectiveness. A favourable CRT response primarily depends on whether implanters can identify the optimal left ventricular (LV) lead position and accurately place the lead at the recommended site. Myocardial imaging techniques, including echocardiography, cardiac magnetic resonance imaging and nuclear imaging, have been used to assess LV myocardial viability and mechanical dyssynchrony, and deduce the optimal LV lead position. The optimal position, presented as a segment of the myocardial wall, is then overlaid with images of the coronary veins from fluoroscopy to aid navigation of the LV lead to the target venous site. Once validated by large clinical trials, these image-guided techniques for CRT lead placement may have an impact on current clinical practice.
Collapse
Affiliation(s)
- Haipeng Tang
- School of Computing, University of Southern Mississippi, Long Beach, MS, USA
| | - Shaojie Tang
- School of Automation, Xi'an University of Posts and Telecommunications, Xi'an, Shaanxi, China
| | - Weihua Zhou
- School of Computing, University of Southern Mississippi, Long Beach, MS, USA
| |
Collapse
|
5
|
Abstract
Cardiac resynchronization therapy (CRT) is a standard treatment for patients with heart failure. However, 30-40 % of the patients having CRT do not respond to CRT with improved clinical symptom and cardiac functions. It is important for CRT response that left ventricular (LV) lead is placed away from scar and at or near the site of the latest mechanical activation. Nuclear image-guided approaches for CRT have shown significant clinical value to assess LV myocardial viability and mechanical dyssynchrony, recommend the optimal LV lead position, and navigate the LV lead to the target coronary venous site. All these techniques, once validated and implemented, should impact the current clinical practice.
Collapse
Affiliation(s)
- Weihua Zhou
- School of Computing, University of Southern Mississippi, 730 East Beach Blvd, Long Beach, MS, 39560, USA.
| | - Ernest V Garcia
- Department of Radiology and Imaging Sciences, Emory University School of Medicine, 1364 Clifton Rd NE, Atlanta, GA, USA.
| |
Collapse
|