- Cardiac CT Perfusion
- Myocardial mass at risk
- Application of artificial intelligence in cardiovascular imaging
- Coronary artery calcification
- Pulmonary CT perfusion
- Airway quantification for chronic obstruction pulmonary disease (COPD)
- Spectral Breast CT
1. Cardiac CT Perfusion
Coronary artery disease (CAD) is the leading cause of morbidity and mortality worldwide. As a risk factor, CAD and its resultant ischemic cardiomyopathy are strongly predicative of future cardiac events. While coronary computed tomography (CT) angiography is a powerful tool for assessing CAD risk, it is fundamentally limited in that it can only assess the morphological severity of segmental CAD but cannot define the physiological severity of concurrent mutli-vessel, diffuse, and microvascular disease. Hence, guidelines recommend additional physiological assessment of CAD, in conjunction with CT angiography, for more objective indication of patient risk. The primary modalities used for physiological assessment are single-photon emission computed tomography (SPECT), stress echocardiography, cardiac magnetic resonance (CMR), static positron emission tomography (PET), and static CT. However, such modalities only provide metrics of relative perfusion; hence, they still cannot appreciate the true physiological severity of multiform CAD. Fortunately, absolute perfusion measurement with dynamic CT can overcome these limitations, where the spatial distribution of absolute rest and stress perfusion in mL/min/g combined with physiological cutoff thresholds can be used to reliably stratify patient risk and properly guide intervention. Nevertheless, current dynamic CT perfusion techniques are known to be quantitatively inaccurate and deliver unacceptably high effective radiation doses per imaging exam, precluding their widespread clinical use. As such, there is a major unmet clinical need for an accurate, low-dose CT technique for combined morphological and physiological assessment of multiform CAD.
Our research focus has been to address that unmet clinical need through the development, validation, and preliminary clinical translation of an accurate, low-dose, comprehensive cardiac CT technique based on first-pass analysis (FPA). The comprehensive technique can accurately assess vessel-specific stress and rest perfusion, while simultaneously providing cardiac functional analysis (CFA), CT angiography, and coronary flow reserve (CFR), respectively. Thus, morphological and physiological assessment of CAD is feasible using a single low-dose exam, making comprehensive CT-based assessment of multiform CAD more accurate, accessible, and impactful to patients in need.
Example visualization of the low-dose cardiac CT technique. Low-dose CFR and stress perfusion in the absence and presence of a significant left anterior descending (LAD) coronary artery balloon stenosis (FFR = 0.70), with co-registered CTA displayed. A LAD perfusion deficit is shown (red arrows). The color bars indicate low-dose CFR and stress perfusion measurement in mL/min/g.
2. Myocardial mass at risk
Subtended myocardium, the amount of heart tissue supplied by a particular coronary artery past a lesion, is an important factor in assessing coronary artery disease. Despite its importance, subtended myocardium is not commonly calculated by physicians. Rather, physicians estimate the amount of subtended myocardium using cardiac imaging methods such as invasive coronary angiography or coronary CT angiography (CCTA).
We have developed the minimum-cost path technique, a post-processing technique that accurately calculates subtended myocardium, using a single CCTA image. After validating this technique in pig models1, 2, we have transitioned to exploring its clinical application. We are now using the minimum-cost path technique to calculate subtended myocardium in patient datasets. With over 800 patient images, we are exploring the relationship between subtended myocardium and coronary artery disease. Building on over a decade of research in allometric scaling laws of the heart, conducted by our group and collaborators3-5, subtended myocardium may link coronary “morphology” and “physiology”, improving the diagnosis of coronary artery disease.
1. Malkasian S, Hubbard L, Abbona P, Dertli B, Kwon J, Molloi S. Vessel-specific coronary perfusion territories using a CT angiogram with a minimum cost path technique and its direct comparison to the American Heart Association 17-segment model. Eur Radiol. Jun 2020;30(6):3334-3345. doi:10.1007/s00330-020-06697-w
2. Malkasian S, Hubbard L, Dertli B, Kwon J, Molloi S. Quantification of vessel-specific coronary perfusion territories using minimum-cost path assignment and computed tomography angiography: Validation in a swine model. J Cardiovasc Comput Tomogr. Sep-Oct 2018;12(5):425-435. doi:10.1016/j.jcct.2018.06.006
3. Choy JS, Kassab GS. Scaling of myocardial mass to flow and morphometry of coronary arteries. J Appl Physiol (1985). May 2008;104(5):1281-6. doi:10.1152/japplphysiol.01261.2007
4. Le HQ, Wong JT, Molloi S. Allometric scaling in the coronary arterial system. Int J Cardiovasc Imaging. Oct 2008;24(7):771-81. doi:10.1007/s10554-008-9303-7
5. Le H, Wong JT, Molloi S. Estimation of regional myocardial mass at risk based on distal arterial lumen volume and length using 3D micro-CT images. Comput Med Imaging Graph. Sep 2008;32(6):488-501. doi:10.1016/j.compmedimag.2008.05.002
Myocardial mass at risk distal to a stenosis. Subtended myocardium at risk distal to a potential stenosis.
3. Application of artificial intelligence in cardiovascular imaging
Many important preprocessing steps are involved in our CT perfusion technique as it relates to the heart and other organs, like vessel centerline extraction, myocardium segmentation, etc. These steps are time consuming when done manually or semi-automatically and therefore it becomes impractical for use in the clinic. Artificial intelligence, and more specifically deep learning, can help solve the issue of time constraints while also allowing for more consistent and reproducible results. My research is focused on automating the entire preprocessing pipeline for our cardiac CT perfusion research, along with other organs in the future.
4. Coronary artery calcification
5. Pulmonary CT perfusion
The advent of quantitative computed tomography (CT) imaging techniques has enabled better evaluation of pulmonary disease such as pulmonary embolism, pulmonary hypertension and lung cancer, etc. Despite existing dynamic CT managed to provide non-invasive quantification of pulmonary perfusion allowing for the assessment of pulmonary diseases, such techniques are known to be inaccurate and lead to high radiation doses per exam. Hence, our aim is to provide an accurate, low-dose dynamic CT perfusion technique for the improved assessment of different pulmonary diseases. In addition, we have proposed a patient-specific contrast timing protocol that can be applied in both CT perfusion technique and CT pulmonary angiography (CTPA). Such protocol significantly improves the diagnostic image quality of CTPA as compared to the existing standard protocol. As a result, a comprehensive, low-dose technique with combined morphological and functional assessment of pulmonary disease is being developed in our group.
6. Airway quantification for chronic obstruction pulmonary disease (COPD)
The alteration of airway function is associated with the airflow obstruction and inflammation in patients with COPD and asthma. Advanced computed tomography (CT) technique has enabled noninvasive quantitative structural measurements of airway dimensions for improved management of airway diseases. The quantification of small airway dimensions (inner diameter<2mm) is of great importance to characterize the severity of airway diseases, leading to better clinical management. Due to the spatial resolution limitations and partial volume effects of current CT techniques in clinical settings, a barrier remains for accurate measurement of the small airway dimensions at higher generations. Our group proposed a new integrated intensity-based (IIB) technique for an accurate quantification of the small airway dimensions. A deep learning algorithm based on convolutional neural network (CNN) is also under investigated on the measurements airway wall thickening.
7. Spectral breast CT