Myocardial Motion Analysis for Determination of Tei-Index of Human Heart

Shengyong Chen, Jianhua Zhang, Houxiang Zhang, Qiu Guan, Yahui Du, Chunyan Yao, Jianwei Zhang2010 · DOI

TL;DR

This paper introduces a novel method for directly calculating the Tei-index, a crucial indicator of heart function, by accurately segmenting the cardiac cycle into distinct physical phases. This approach, validated on real human heart images, offers a more precise and direct measurement than previous indirect methods, paving the way for improved diagnosis of cardiovascular diseases.

What Problem Does This Paper Address?

Cardiovascular diseases are a leading cause of death, and early detection is critical. The Tei-index is a valuable tool for assessing both systolic and diastolic heart function, providing a comprehensive view of heart health. However, calculating the Tei-index has been challenging because it relies on measuring the isovolumic contraction time (ICT) and isovolumic relaxation time (IRT), which are difficult to determine directly from medical images. Existing methods are indirect, relying on Doppler or acoustic quantification, which can introduce errors. This paper aims to bridge this gap by proposing a direct method for Tei-index calculation through accurate cardiac cycle phase division.

Methods

The core of this paper lies in two main contributions:

  • Cardiac Physical Phase (CPP) Division: The authors propose a method to divide the cardiac cycle into seven distinct physical phases (Isovolumic Systolic, Rapid Ejection Period, Slow Down Ejection Period, Isovolumic Relaxation Period, Rapid Filling Period, Slow Down Filling Period, and Atrial Systole). This division is based on analyzing the dynamics of the left ventricle (LV) shape and volume, specifically using B-spline models to represent the LV's inside-wall surface. They track changes in LV Volume (LVV) and LV Long-Axis Length (LVLL) and their derivatives to identify the boundaries of these phases.
  • Direct Tei-Index Calculation: Once the cardiac cycle is accurately partitioned into CPPs, the authors can directly calculate the Tei-index. They equate the isovolumic contraction time (ICT) to the Isovolumic Systolic (IS) phase and the isovolumic relaxation time (IRT) to the Isovolumic Relaxation Period (IRP). The ejection time (ET) is then calculated as the sum of the Rapid Ejection Period (REP) and Slow Down Ejection Period (SDEP). This allows for a direct computation of the Tei-index using the durations of these identified phases.
  • Left Ventricular Wall Motion Amplitude (LVMA) Calculation: In addition to the Tei-index, the paper also presents a method for calculating LVMA. This involves tracking a large number of points on the LV wall and fitting their trajectories with B-spline curves to represent the displacement of the LV wall over time. LVMA is then computed for each CPP.

Datasets

The experiments were conducted using real human heart medical images. The paper does not specify the exact source or nature of these images beyond them being SPECT heart images. No new benchmark datasets were introduced.

Key Insights and Ideas

  • The Cardiac Cycle is More Than Just Systole and Diastole: The paper highlights that a more granular division of the cardiac cycle into seven distinct physical phases, based on LV deformation, is crucial for accurate functional analysis.
  • Direct Tei-Index Calculation is Feasible and Accurate: By accurately identifying the CPPs, the authors demonstrate that the Tei-index can be directly calculated, overcoming the limitations of previous indirect methods.
  • LVMA Provides Complementary Diagnostic Information: The proposed method for calculating LVMA offers a way to visualize and quantify regional wall motion abnormalities, which can be indicative of heart disease.
  • B-spline Models are Powerful for Representing Complex Shapes: The use of B-spline models for representing the dynamic 3D shape of the left ventricle is a key enabler for both CPP division and LVMA analysis.

Key Results

  • The proposed CPP partition method resulted in phase durations that were very close to medical statistical average data, validating the accuracy of their segmentation.
  • The calculated Tei-index was 0.3617, which falls within the common range of 0.39 ± 0.05 as per medical statistics, suggesting the direct calculation method is reliable.
  • The distribution of LVMA across different CPPs showed patterns consistent with medical knowledge, with apical LVMA being larger than basal LVMA and concentrated in specific regions during systole and diastole.

Limitations and Future Work

  • Image Modality: While SPECT images were used, the paper could benefit from validation on other imaging modalities like MRI or echocardiography to assess broader applicability.
  • AS Phase Deformation: The paper notes that LV deformation is inconspicuous during the Atrial Systole (AS) phase, which might limit the diagnostic utility of LVMA during this specific phase.
  • Future Work: The authors suggest that their method can be used for analyzing myocardial motion and that the color figures of LVMA in different CPPs show rules of myocardial motion that are very close to the results of other researchers, implying potential for further research and clinical application.

Summary updated

BibTeX
@misc{chen2010myocardial,
  author = {Shengyong Chen and Jianhua Zhang and Houxiang Zhang and Qiu Guan and Yahui Du and Chunyan Yao and Jianwei Zhang},
  title = {{Myocardial Motion Analysis for Determination of Tei-Index of Human Heart}},
  year = {2010},
  doi = {10.3390/s101211428},
}
Figure