A dual-layer flat panel x-ray detector based on an engineered amorphous chalcogenide alloy for quantifying coronary artery calcium

PROJECT ABSTRACT
Heart disease is extremely prevalent, with about one in every four deaths (in the US) being attributed to heart
disease. Early detection of cardiovascular events, especially before patients become symptomatic, has immense
impact in preventive healthcare, reducing the morbidity and mortality associated with cardiovascular disease.
Coronary artery calcification (CAC), a strong predictor for future cardiovascular events, is a component of
atherosclerotic plaque buildup in the arteries that supply blood to the heart, leading to coronary artery disease
(CAD). Identification of CAC is clinically important because it is used for cardiovascular risk and therapy decision
making. Currently, CAC is quantified by computed tomography (CT), however, CT-based population screening
is not widely utilized due to cost and radiation burden. Chest x-rays (CXR) are the most common medical imaging
procedure and have higher availability than CT in low-resource settings, lower radiation dose, and higher patient
throughput that could be used for screening purposes. Unfortunately, due to the lack of quantification in CXR,
only qualitative descriptors are possible. The objective of this proposal is therefore to bring much-needed
quantification to CXR, particularly for detecting and quantifying CAC by combining a new dual-layer x-ray
detector and artificial-intelligence based image processing. The proposed dual-layer detector utilizes alloys of
amorphous selenium (a-Se) that achieve favorable electro-optical properties (e.g., higher charge carrier
mobilities and higher gain) compared to conventional a-Se based x-ray detectors. This technology has four major
components: (1) a top layer direct convection a-Se alloys on an imaging backplane, (2) a bottom layer indirect
conversion a-Se alloy with intrinsic gain on an imaging backplane coupled to a scintillator, (3) top panel and
bottom panel integration into a dual-layer detector, and (4) a machine learning algorithm that enhances accuracy
of the quantitative information from the dual-layer detector. The detector development leverages a mature
platform from Varex Imaging, a leading manufacturer of x-ray detectors. We expect to show that the proposed
system has higher spatial resolution images and higher sensitivity to detect small, high contrast features
(calcifications) and to separate materials such as calcium from soft tissue. This approach will allow accurate
quantification of predictive factors and will have immense impact in proactive healthcare, improving the clinical
outcomes of patients, and reducing the number of deaths associated with cardiovascular disease. While our
focus is on CAC, we expect this technology to broadly improve CXR for early detection of lung cancer,
tuberculosis, and other diseases such as osteoporosis via quantification of bone mineral density.

Grant Number: 5R01EB033466-04
NIH Institute/Center: NIH
Principal Investigator: Shiva Abbaszadeh