A pilot study using low-dose Spectral CT and ASIR (Adaptive Statistical Iterative Reconstruction) algorithm to diagnose solitary pulmonary nodules

Lung cancer is the most common cancer which has the highest mortality rate. In the past decades, the incidence of lung cancer has gradually increased in China [1, 2]. With the development of computed tomography (CT) techniques, the case detection rates of solitary pulmonary nodules (SPN) has constantly increased and the diagnosis accuracy of SPN has remained a hot topic in clinical and imaging diagnosis. Contrast-enhanced CT of the chest still remains the standard imaging test for the initial assessment of patients with suspected lung cancer. Using standard contrast-enhanced CT the characterization of pulmonary nodules is based on simple morphological criteria e.g., irregular or spiculated margins as a sign for malignancy or calcifications as a sign of benignity. However, in a clinical context these simple morphologic criteria are unreliable for an accurate differentiation between benign and malignant lung nodules.

X-ray computed tomography (CT) is a medical imaging modality that allows reconstruction of the internal stucture of the human body from a large number of x-ray attenuation measurements. The spectral CT in which the energy dependence of the x-ray attenuation coefficient is utilized Multiple parameters can be acquired by means of spectral CT techniques, such as monochromatic imaging, material decomposition images, spectral HU curve and effective atomic number, etc. [3, 4] ASIR (Adaptive Statistical Iterative Reconstruction) algorithm, is a compromise that relies on the accurate modeling of the noise distribution of the acquired data, rather than modeling the system optics. The result is an algorithm that is computationally fast and is effective at reducing noise, enabling radiation dose reductions that would not be possible [5‐12]. The ASIR reconstruction algorithm is a promising technique for providing diagnostic quality CT images at significantly reduced radiation doses. ASIR is also helpful in improving CT image quality for obese patients.

In this study, patients with solitary pulmonary nodules (SPN) underwent dual-phase scanning by low dose spectral CT. The iodine and water concentrations were derived and the spectral HU curves were also acquired. By calculating and comparing the normalized concentration of iodine and water, the slopes of spectral curves in the benign and malignant SPN, the practical value of multiple parameters which was acquired by low dose spectral CT in SPN diagnosis are discussed. In this paper we propose an improved method to detect SPN. By analysis of different comparison parameters between benign and malignant pulmonary nodules and provide a reference for clinical diagnosis and treatment.

The methodology to diagnose SPN has been developing over the years from a traditional morphological examination to a dynamic functional examination. Spectral CT imaging has the potential to provide multiple techniques, such as polychromatic and monochromatic imaging, material decomposition imaging, etc. Therefore, spectral CT has been widely used in the detection of early stage cancer, qualitative diagnosis of neoplastic disease, staging diagnosis and reducing the metal artifacts and so on [13‐17].

It was reported that tumor cells can produce angiogenic factors that stimulate and generate a large number of new blood vessels. Since the wall of these new blood vessels are immature there is a lack of hemangiopericyte and smooth muscle tissue and there is increased permeability of the blood vessels [18‐20]. Lung cancer cells by themselves are highly aggressive; they can invade the corresponding artery making the terminal blood vessels get thicker and circuitous. Since there is a lack of venous and lymphatic drainage system in the cancer lesion, the contrast agent gets diffused into the extravascular space. But in the case of benign SPN, most of them do not have abundant blood supply and the vascular basement membrane is not damaged, so the permeability of blood vessels is not increased. Tateishi et al. analyzed the correlation of tumor enhancement with MVD and VEGF expression in 130 patients with histological proven lung cancer [21]. They found a significantly higher peak enhancement of VEGF-positive tumors than in VEGF-negative tumors. In this study, the ICAP, ICVP, NIC and ICD in malignant SPN were all significantly higher than that of benign SPN. In benign SPN, the active inflammatory nodule also contains more blood vessels due to the stimulation of inflammatory substances. With the conversion of an active nodule to a chronic inflammatory nodule, the fabric content in the lesion increases while the vascular content decreases. As a consequence, the iodine concentration in active inflammatory nodule may also be high, and comparable to a malignant nodule. In the chronic inflammatory and other benign nodule, due to the lack of blood vessel, the contrast agent diffuse slowly, therefore, in some case, the ICVP is higher than ICAP. It is similar to the study of Schmid-Bindert et al., who investigated the correlation between maximum standardized uptake value (SUV(max)) of (18) FDG PET-CT and iodine-related attenuation (IRA) of dual energy CT (DECT) of primary tumours and (18) FDG PET-CT positive thoracic lymph nodes (LN) in patients with lung cancer and a moderate correlation was found between SUV(max) and maximum IRA in all tumours [22].

In the iodine/water based material decomposition image, the water content was measured. Based on the calculation and statistical analysis, the results showed that WC, NWC and WCD in both benign and malignant SPN have no significant differences. In this study, the inflammatory nodule cases were relatively high in the benign SPN group. Since the blood flow of inflammatory nodule increases in the congestive stage, the water content in both intra- and extra-cellular is high. As for the malignant SPN, the water content in intra- and extra-cellular is also high due to the relatively higher vascular capacitance and large amount of tumor cells. The center of tuberculoma consists of caseous necrotic tissue surrounded by fibrous tissue, and the caseous necrotic tissue is a pink amorphous granular mass, which exhibits more severe necrosis: coagulative necrosis. Coagulative necrosis is enriched with lipid and less water content. In this study, the water content is high in inflammatory nodule cases, which led to the premise that water concentration related parameters (WC, NWC, WCD) are not significantly different between benign and malignant groups.

The iodine concentration in the ROI directly reflects the blood supply situation in the nodule. Research showed that the IC in malignant nodules is higher than in benign nodules [23]. Several studies have compared the CT numbers of pulmonary nodules on iodine-enhanced image with that on enhanced weighted average images The results of both the CT number on iodine-enhanced images and the degree of enhancement showed that malignant nodules showed a significantly higher enhancement than benign nodules (P = 0.001), and iodine-enhanced images had a higher sensitivity and accuracy than the degree of enhancement [24‐28]. The spectra curve is a reflection of different lesions and different tissues or organs in the human body absorb X-rays at different rates. It shows the variation of CT values in different regions of different keV [29, 30].The difference of spectral curve needs to be correlated to the iodine concentration in the lesions when the contrast agent is applied. In this study, the slopes of spectral HU curve were decreased with the increase in keV in both benign and malignant nodules. However in all the 3 curve sections, the slopes of malignant SPN are all higher than the corresponding slopes of benign SPN.

The low dose spectral scan mode was applied in this research. The tube current was set to 260 mA, CTDIVol was 4.17 mGy per phase, which is significantly lower compared to the dose of the first generation technology. In addition, the ASIR algorithm reduced the noise and improved the imaging quality, and makes it possible to acquire good quality image in much lower dosage. The percentage of ASIR (10–100 %) is operator selectable at the console. It reflects a linear combination of the original FBP image (0 % ASIR) and an essentially noise-free image created by full compliance with the mathematic model (100 % ASIR). A choice of 40 % ASIR implies that 40 % of the ASIR image was blended with the FBP image. A preliminary phantom analysis and clinical feasibility study of low-dose body CT using 40 % ASIR determined that it provided quantitative and qualitative image noise and quality nearly identical to those of routine-dose CT [31].

In this study, the ultralow concentration iso-osmolar contrast agent (270mgI/mL) was given to the patients; consequently, the incidence of adverse reaction was reduced, which is a better choice of contrast agent when the patients have potential renal damage or cardiac insufficiency.

Although the results in this research are concrete and convincing, this work still has several aspects that need to be explored thoroughly in further studies, such as, expanding the cases; detailed study based on different pathological types of lung cancer; distinguishing the active inflammatory nodules from malignant nodules; comparing the water content related parameters in different pathological nodules, etc. In conclusion, the combination of low dose spectral CT and ASIR algorithm can help acquire multiple parameters under the low dose mode and acquiring these parameters is highly practicable in the clinic during diagnosis of benign and malignant SPN.