Comparative evaluation of non-contrast CAIPIRINHA-VIBE 3T-MRI and multidetector CT for detection of pulmonary nodules: In vivo evaluation of diagnostic accuracy and image quality

Abstract

Purpose

To evaluate the diagnostic accuracy, subjective image quality, and interobserver agreement of non-contrast Controlled Aliasing In Parallel Imaging Results In Higher Acceleration (CAIPIRINHA) volumetric interpolated breath-hold examination (VIBE) 3T magnetic resonance imaging (MRI) for the detection of pulmonary nodules with intra-individual comparison to computed tomography (CT).

Materials and methods

We evaluated 54 patients (27 male, 27 female; mean age, 60.8 ± 11.5 years) who prospectively underwent thoracic 3T-MRI using CAIPIRINHA-VIBE sequences and chest CT. Diagnostic accuracy for the detection of lung nodules on CAIPIRINHA-VIBE MRI by three independent observers were compared to the reference standard CT. Subjective image quality was rated using a 5-point grading scale. Diagnostic accuracy was calculated and interobserver agreement was assessed using intraclass correlation coefficient (ICC).

Results

Sensitivity of 3T-MRI for the detection of pulmonary lesions compared to CT was 88.1% (95% confidence interval [CI]: 0.81–0.93) and specifity was 79.1% (95% CI: 0.50–0.95). Sensitivity for lesions <5 mm was 77.2% (95% CI: 0.59–0.90) and for lesions from 5 to 10 mm was 87.2% (95% CI: 0.76–0.94). Sensitivity for lesions >10 mm was 100%. Observer ratings regarding subjective image quality were good to excellent for 3T-MRI (1.54) and CT (1.14) with almost perfect interobserver agreement for 3T-MRI and CT (ICC = 0.83, 95% CI: 0.78–0.89; ICC = 0.89, 95% CI: 0.85–0.94).

Conclusions

Non-contrast CAIPIRINHA-VIBE 3T-MRI allows for the reliable detection of pulmonary lesions with a diameter >5 mm in comparison with chest CT with high diagnostic accuracy, subjective image quality, and interobserver agreement.

Introduction

Computed tomography (CT) remains the gold standard for high-resolution imaging of the lung, to assess pulmonary disease, and to detect pulmonary nodules. However, several studies have demonstrated that magnetic resonance imaging (MRI) may to a certain extent represent a viable alternative to CT for lung imaging without exposure to ionizing radiation [1], [2]. Current sequences for MRI detection of pulmonary nodules are 3D T1-weighted gradient-echo sequences with volumetric interpolated breath-hold examination (VIBE), T2-weighted fast back echo (SE) sequences with half-Fourier acquisition single-shot turbo spin-echo (HASTE) and balanced steady state-free precession (b-SSFP) sequences together with T2-weighted fat-suppressed turbo spin-echo sequences [3]. While the spatial and temporal resolution of CT are currently superior to MRI, several recent technological advances have been made in MRI technique to improve image quality in lung MRI and visualization of pulmonary nodules.

Due to generally slower and repeated imaging acquisition, respiratory and cardiac motion have a stronger influence on image quality in MRI compared to CT. Furthermore, for optimal visualization of lung parenchyma on MRI, two characteristics are of major importance: the relatively low amount of soft tissue or water components of the lung tissue and the poor differentiation between parenchyma and air [4]. This usually results in a low signal-to-noise ratio (SNR) in lung MRI [4].

In recent years, several advances in MRI technology have been introduced to obtain diagnostic image quality in lung imaging [5], [6], [7], [8], [9]. Parallel MRI (pMRI) operates with a reduced amount of k-space data to generate an image. Prior studies have demonstrated improved diagnostic performance for the detection of pulmonary nodules for this technique [6], [10], [11], [12], [13], [14], [15]. While this allows for a more rapid image acquisition, this can also result in aliasing artifacts. The Controlled Aliasing In Parallel Imaging Results In Higher Acceleration (CAIPIRINHA) sequence is able to reduce these artifacts and overcome other limitations of pMRI [16]. CAIPIRINHA facilitates the application of higher acceleration factors (>2) with modified data acquisition patterns and image reconstruction to improve utilization of the sensitivity variations of the receiver coil array in 2 phase-encoding dimensions [17]. Recent reports have demonstrated that CAIPIRINHA reduces breath-hold times and simultaneously optimizes contrast in liver MRI [17].

T1-weighted gradient echo sequences with volume volumetric interpolated breath-hold examination (VIBE) have been shown to allow for motion-free imaging during a single breath-hold [18]. Generally, this technique is used in combination with the parallel imaging technique to reduce the required time and improve the spatial resolution. Restrictions of the parallel imaging technique are a decrease in the signal-to-noise ratio (SNR) and aliasing artifacts due to a high acceleration factor (AF) compared to fully sampled k-space data, ultimately leading to a reduction in image quality.

The dedicated CAIPIRINHA sampling pattern helps allocate k-space points more consistently and can improve algorithms for reconstruction in two-dimensional parallel Imaging [18], [19]. Furthermore, higher parallel acquisition technique factors involve higher SNR and higher spatial resolution which result in excellent image quality [18], [20]. These potential advantages may help increase patient compliance during the examination.

However, to our knowledge, the usefulness of the CAIPIRINHA-VIBE sequence in MRI for the detection of pulmonary nodules has not been investigated so far. Furthermore, its diagnostic performance in comparison with CT and its impact on subjective image quality and interobserver agreement remain unclear. We aimed to evaluate this sequence and its potential to provide MRI data of the lung in patients with pulmonary nodules as and assess its diagnostic performance in comparison with CT since patients with known pulmonary nodules but different oncologic disease often undergo multi-modality imaging. We hypothesized that CAIPIRINHA-VIBE may provide adequate image quality and diagnostic performance compared to CT for the assessment of pulmonary nodules.

Thus, the purpose of our study was to evaluate the diagnostic accuracy, subjective image quality, and interobserver agreement of non-contrast CAIPIRINHA-VIBE 3T-MRI for the detection of pulmonary nodules with intra-individual comparison to CT.