Advertisement

Prospective evaluation of thoracic diseases using a compact flat-panel detector spiral computed tomographic scanner

  • Da Som Kim
    Affiliations
    Department of Radiology, Inje University Busan Paik Hospital, Inje University College of Medicine, Busan, South Korea
    Search for articles by this author
  • Seung-Jin Yoo
    Affiliations
    Department of Radiology, Hanyang University Medical Center, Hanyang University College of Medicine, Seoul, South Korea
    Search for articles by this author
  • Jung Hee Hong
    Affiliations
    Department of Radiology, Keimyung University School of Medicine, Daegu, South Korea
    Search for articles by this author
  • Nakwon Kwak
    Affiliations
    Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, Seoul National University Hospital, Seoul, South Korea
    Search for articles by this author
  • Jae-Joon Yim
    Affiliations
    Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, Seoul National University Hospital, Seoul, South Korea

    Department of Internal Medicine, Seoul National University College of Medicine, Seoul, South Korea
    Search for articles by this author
  • Soon Ho Yoon
    Correspondence
    Correspondence to: Department of Radiology, Seoul National University Hospital, Seoul National University College of Medicine, 101, Daehak-ro, Jongno-gu, Seoul 03080, South Korea.
    Affiliations
    Department of Radiology, Seoul National University Hospital, Seoul National University College of Medicine, Seoul, South Korea
    Search for articles by this author
Open AccessPublished:November 16, 2022DOI:https://doi.org/10.1016/j.ejro.2022.100452

      Abstract

      Objective

      To prospectively evaluate the image quality and diagnostic performance of a compact flat-panel detector (FD) scanner for thoracic diseases compared to a clinical CT scanner.

      Materials and methods

      The institutional review board approved this single-center prospective study, and all participants provided informed consent. From December 2020 to May 2021, 30 patients (mean age, 67.1 ± 8.3 years) underwent two same-day low-dose chest CT scans using clinical state-of-art and compact FDCT scanners. Image quality was assessed visually and quantitatively. Two readers evaluated the diagnostic performance for nodules, parenchymal opacifications, bronchiectasis, linear opacities, and pleural abnormalities in 40 paired CT scans. The other 20 paired CT scans were used to examine the agreement of semi-quantitative CT scoring regarding bronchiectasis, bronchiolitis, nodules, airspace consolidations, and cavities.

      Results

      FDCT images had significantly lower visual image quality than clinical CT images (all p < 0.001). The two CT image sets showed no significant differences in signal-to-noise and contrast-to-noise ratios (56.8 ± 12.5 vs. 57.3 ± 15.2; p = 0.985 and 62.9 ± 11.7 vs. 60.7 ± 16.9; p = 0.615). The pooled sensitivity was comparable for nodules, parenchymal opacifications, linear opacities, and pleural abnormalities (p = 0.065–0.625), whereas the sensitivity was significantly lower in FDCT images than in clinical CT images for micronodules (p = 0.007) and bronchiectasis (p = 0.004). The specificity was mostly 1.0. Semi-quantitative CT scores were similar between the CT image sets (p > 0.05), and intraclass correlation coefficients were around 0.950 or higher, except for bronchiectasis (0.869).

      Conclusion

      Compact FDCT images provided lower image quality but comparable diagnostic performance to clinical CT images for nodules, parenchymal opacifications, linear opacities, and pleural abnormalities.

      Keywords

      Abbreviations:

      CT (computed tomography), COVID-19 (coronavirus disease 2019), LDCT (low-dose CT), FD (flat-panel detector), SNR (signal-to-noise ratio), CNR (contrast-to-noise ratio), ROI (region of interest), POC (point-of-care), NTM (nontuberculous mycobacterial), ICC (intraclass correlation coefficient)

      1. Introduction

      Chest CT is an essential imaging modality to evaluate pulmonary diseases in modern medicine and plays a salient role in managing the COVID-19 pandemic [
      • Rubin G.D.
      • Ryerson C.J.
      • Haramati L.B.
      • Sverzellati N.
      • Kanne J.P.
      • Raoof S.
      • et al.
      The role of chest imaging in patient management during the covid-19 pandemic: a multinational consensus statement from the fleischner society.
      ,
      • Kim H.
      • Hong H.
      • Yoon S.H.
      Diagnostic performance of ct and reverse transcriptase polymerase chain reaction for coronavirus disease 2019: a meta-analysis.
      ]. Nevertheless, extant clinical CT machines have limited mobility and require substantial efforts and complex sanitizing measures to scan both patients with and without infectious pulmonary diseases such as COVID-19 [
      • Chen Q.
      • Zu Z.Y.
      • Jiang M.D.
      • Lu L.
      • Lu G.M.
      • Zhang L.J.
      Infection control and management strategy for covid-19 in the radiology department: focusing on experiences from china.
      ]. Compact flat-panel detector (FD) CT machines are smaller than clinical CT machines and are mobile, as they are typically equipped with a C-arm [
      • Kalender W.A.
      • Kyriakou Y.
      Flat-detector computed tomography (fd-ct).
      ]. FD CT machines may be deployed rapidly to respond to outbreaks of infectious diseases for exclusive point-of-care (POC) diagnostic use in patients with contagious diseases (e.g., in small, isolated spaces in the intensive care unit) [
      • Rho J.Y.
      • Yoon K.H.
      • Jeong S.
      • Lee J.H.
      • Park C.
      • Kim H.W.
      Usefulness of mobile computed tomography in patients with coronavirus disease 2019 pneumonia: a case series.
      ]. However, FD CT has a few disadvantages, such as lower dose efficiency, a smaller field of view, and lower temporal resolution for pulmonary applications [
      • Kalender W.A.
      • Kyriakou Y.
      Flat-detector computed tomography (fd-ct).
      ]. An FD CT machine with a relatively large-area detector might overcome the current limitations, but it remains underexplored whether FD CT images provide sufficient diagnostic information comparable to that obtained using clinical CT machines.
      This study aimed to prospectively evaluate the image quality and diagnostic performance of a compact POC FD CT scanner by performing two consecutive low-dose CT scans using FD and clinical state-of-art scanners in patients with thoracic diseases.

      2. Materials and methods

      2.1 Patients

      This single-arm prospective trial was approved by our institutional review board, and informed consent was obtained from all participants. The study protocol was registered in the Korean Clinical Information Service (Registration Number: KCT0005633).

      2.2 Patients

      We applied the following inclusion criteria to determine eligibility from December 2020 to May 2021: a) adult patients 20 years or older who were suspected of having pulmonary disease; b) patients who planned to undergo chest CT scans electively; and c) patients who fully understood the content of this trial and provided informed consent. We excluded women who were pregnant or might become pregnant. A study coordinator collected baseline characteristics of the enrolled patients, including age, sex, body mass index (BMI), and smoking history, from the electronic medical record system.

      2.3 CT acquisition

      2.3.1 Conventional low-dose CT scans

      All low-dose chest CT scans were performed using an up-to-date flagship multidetector CT scanner (Somatom Force, Siemens Healthineers, Erlangen, Germany) with gantry dimensions of 1.99 × 1.18 × 2.40 m3. The CT scans were obtained at full inspiration and served as reference images. The scanner’s size and weight were 3.27 × 1.99 × 1.18 m3 and 3150 kg, respectively. The minimum space for CT deployment was 11.6 × 4.5 m2. The following CT parameters were used in clinical practice: tube voltage, Sn100 kVp without CARE kVp; quality reference tube current, 130 mAs with CAREDose4D; scan fields of view, 50 × 50 cm; matrix, 512 × 512; rotation time, 0.25 s/rotation; detector collimation, 192 × 0.6 mm; and pitch, 3, indicating that a whole lung was typically scanned within one second. All CT scans were reconstructed using advanced modeled iterative reconstruction algorithms into 3-mm axial images with a soft tissue kernel (Br40–3) and 1-mm axial images with a sharp kernel (Br59–3). The craniocaudal scan covered the area from the lower neck to the costophrenic angle in a single breath-hold.

      2.3.2 Compact POC FD low-dose CT scans

      After conventional CT scanning for clinical purposes, additional ultra-low CT scans for research purposes were subsequently taken on the same day using a compact FD spiral CT scanner (Smart M, Vatech, Gyeonggi-do, South Korea) with gantry dimensions of 1.24 × 0.73 × 1.57 m3. The scanner’s size and weight were 1.24 × 2.38 × 1.57 m3 and 600 kg, respectively. The minimum space for CT deployment was 3.5 × 2 m2. The CT machine was located close to the clinical CT machine, within 1 min on foot. The fixed CT parameters were used as follows: tube voltage, 120 kVp; tube current, 10 mAs; scan field of view, 42 × 30 cm; matrix, 500 × 217; rotation time, 1 s/rotation; detector collimation, 5.3 mm; and pitch, 1.0, indicating that a whole lung was scanned in around 30 s in total. All CT scans were reconstructed using iterative reconstruction algorithms into 1-mm axial images with a soft tissue kernel. The craniocaudal scan covered the area from the lung apex to the costophrenic angle level in two breath-holds, as the CT scanning in a single breath-hold could cover approximately 16 cm.

      2.4 Qualitative image quality evaluation

      Two thoracic radiologists (J.H.H. and S.J.Y., with 6–7 years of experience in chest CT interpretation) independently evaluated the image quality of the low-dose CT images. The readers reviewed a total of 60 CT scans (30 clinical and 30 FD CT scans, each) in a random order to grade overall image quality, image noise, presence of artifact, and depiction of normal structure. They were blinded to clinical information and CT scanners. Detailed information on the scale used for image quality assessment is provided in Supplementary Table 1.

      2.5 Quantitative image quality assessment

      One thoracic radiologist (S.H.Y., with 16 years of experience in body CT interpretation) analyzed objective image quality by measuring the signal-to-noise ratio (SNR) and the contrast-to-noise ratio (CNR). For SNR analysis, the radiologist manually drew a total of six round regions of interest (ROIs) in the right and left lung parenchyma per CT scan at the level of the aortic arch, hilum, and inferior pulmonary vein, while avoiding pulmonary lesions, bronchi, and vessels. The reader reviewed the clinical and FD low-dose CT scans side-by-side on a picture archiving and communication system to place the ROIs identically on both CT scans. The ROIs were maintained between 50 and 100 mm2. For CNR analysis, the reader placed round ROIs in the main pulmonary trunk and the largest pulmonary lesion if a 10-mm or larger lesion existed.

      2.6 Quantitative assessment of lesion detectability

      Two thoracic radiologists (J.H.H. and S.J.Y) independently identified parenchymal abnormalities in 40 paired CT scans from 20 patients, with the exclusion of 20 scans from 10 patients with confirmed nontuberculous mycobacterial (NTM) lung disease. A total of 40 paired CT scans were assigned randomly, regardless of the pair. After reviewing half of the CT scans, readers reviewed the rest in two weeks to reduce potential recall bias.
      The reviewers detected and recorded lesions characteristics and locations: nodules (micronodules [3 mm or smaller] and nodules), parenchymal opacifications (ground-glass opacities and consolidations), bronchiectasis, linear opacities (fibrotic scars, mucus plugs, atelectasis, and reticular opacities), and pleural lesions (pleural effusion or thickening). The other two thoracic radiologists (D.S.K. with a 6-year experience of CT interpretation and S.H.Y.) reviewed the CT images in consensus and prepared reference judgments. Then, the identified lesion records were matched with the reference data.
      In total, 100 pulmonary lobes in 20 patients (five lobes per patient) were evaluated for lesion detectability. Sensitivity was defined as the proportion of lesions that matched between the identified lesions and the reference. Multiple lesions in the same lobe were individually counted. Specificity was defined as the proportion of lobes without abnormalities that matched between the identified lesions and the reference.

      2.7 Quantitative assessment of lesion extent

      Ten of 30 patients had NTM lung disease, which tended to have an extensive distribution, hampering the assessment of lesion detectability. Their CT scans were used to determine whether compact FD low-dose CT scans could be suitable for quantifying the extent of various parenchymal abnormalities. The two radiologist readers independently graded the extent of parenchymal abnormalities in a semi-quantitative manner regarding bronchiectasis, cellular or inflammatory bronchiolitis, nodules of 10–30 mm in diameter, airspace consolidation, and cavities based on the NTM CT scoring system proposed by Kim et al. [
      • Kim H.S.
      • Lee K.S.
      • Koh W.J.
      • Jeon K.
      • Lee E.J.
      • Kang H.
      • et al.
      Serial ct findings of mycobacterium massiliense pulmonary disease compared with mycobacterium abscessus disease after treatment with antibiotic therapy.
      ]. In the system, the CT scores reflect the severity of lung involvement in NTM pulmonary diseases. The mean CT score in each patient was calculated by averaging the CT score in the pulmonary lobes. A total of 20 paired scans were randomly assigned regardless of the pair. Readers first reviewed 10 CT scans and reviewed the rest in two weeks to reduce potential recall bias.

      3. Statistical analysis

      The normality of the distribution of continuous variables was tested using the Kolmogorov-Smirnov test. The paired t-test and Wilcoxon signed-rank test were used for continuous variables after normality testing to compare the qualitative and quantitative image quality parameters between CT machines. Interobserver agreement was evaluated using intraclass correlation coefficients (ICCs) with a two-way model for each qualitative analysis item: 0.76–1.0, excellent agreement; 0.40–0.75, fair to good agreement; and < 0.40, poor agreement. Bland-Altman analyses were used to assess the agreement among the CT machines. The McNemar test was used to compare the sensitivity and specificity of clinical and compact FD low-dose CT scans. All statistical analyses were performed using the IBM SPSS Statistics 26.0 software package (IBM Corp., Armonk, NY, USA). P-values < 0.05 were considered statistically significant.

      4. Results

      4.1 Patient characteristics

      Patient characteristics are shown in Table 1. The 30 patients agreed to undergo additional FD CT scans to clinical CT scans and included 12 men (mean age, 67.1 ± 8.3 years; age range, 51–79 years) and 18 women (mean age, 62.8 ± 10.5 years; age range, 43–83 years). The mean BMI was 20.8 kg/m2 (range, 12.8–29.8 kg/m2). Of these 30 patients, five had a history of smoking and 10 had received a diagnosis of NTM lung disease. Four participants were suspected of having NTM lung disease in a limited extent of the lung parenchyma, but without pathological confirmation.
      Table 1Baseline patient characteristics.
      Patient characteristicsValue
      Age64.5 (43–79)
      Sex
      Men12 (40 %)
      Women18 (60 %)
      Height162.1 (145–183)
      Weight55.1 (32–88)
      Body mass index (kg/m2)20.8 (12.8–29.8)
      Smoking
      Non-smoker25 (83.7 %)
      Current or former smoker5 (16.7 %)
      Diagnosis
      NTM lung disease10
      M. avium5
      M. intracellulare3
      M. abscessus2
      M. iranicum1
      Suspected NTM lung disease4
      Bronchiectasis with bronchiolitis2
      Healed tuberculosis4
      Suspected interstitial lung disease3
      Chronic pulmonary aspergillosis1
      Data in parentheses indicate data range or proportion.
      NTM=nontuberculous mycobacterial.

      4.2 Qualitative comparison of image quality

      Both readers gave better scores to the clinical CT images than to the compact FD CT images in terms of all image quality parameters (Table 2): overall image quality (4.34 ± 0.81 vs. 2.37 ± 0.60; p < 0.001), image noise (4.40 ± 0.76 vs. 2.93 ± 0.80; p < 0.001), motion and streak artifacts (4.45 ± 0.79 vs. 2.67 ± 0.87; p < 0.001), depiction of intrapulmonary vessels and the lobar fissure (4.79 ± 0.69 vs. 3.28 ± 0.82; p < 0.001), depiction of mediastinal structures (4.54 ± 0.79 vs. 2.76 ± 0.56; p < 0.001), and depiction of bronchial structures (4.65 ± 0.74 vs. 3.34 ± 0.99; p < 0.001). The CT scores did not significantly differ according to sex and BMI (Supplementary Table 2).
      Table 2Qualitative comparison of image quality between clinical and compact flat-panel detector low-dose CT scans.
      Clinical LDCTCompact FD LDCTP-valueICC [95 % CI]
      Overall image quality4.34 ± 0.812.37 ± 0.60< 0.0010.88 [0.80, 0.93]
      Image noise4.40 ± 0.762.93 ± 0.80< 0.0010.76 [0.59, 0.85]
      Motion and streak artifacts4.45 ± 0.792.67 ± 0.87< 0.0010.86 [0.77, 0.92]
      Depiction of intrapulmonary vessels and lobar fissure4.79 ± 0.693.28 ± 0.82< 0.0010.90 [0.83, 0.94]
      Depiction of mediastinal structures4.54 ± 0.792.76 ± 0.56< 0.0010.85 [0.75, 0.91]
      Depiction of bronchial structures4.65 ± 0.743.34 ± 0.99< 0.0010.83 [0.72, 0.90]
      LDCT=low-dose CT; FD=flat-panel detector; ICC=intraclass correlation coefficient; CI=confidence interval.

      4.3 Quantitative comparison of image quality

      There was no significant difference in the SNR and CNR between clinical and compact FD low-dose CT scans (mean SNR: 56.8 ± 12.5 vs. 57.3 ± 15.2, p = 0.985; lesion CNR: 62.9 ± 11.7 vs. 60.7 ± 16.9, p = 0.615) (Table 3).
      Table 3Quantitative comparison of image quality between clinical and compact flat-panel detector low-dose CT scans.
      Clinical LDCTCompact FD LDCTP-value
      Contrast-to-noise ratio
      Lesion62.9 ± 11.760.7 ± 16.90.615
      Main pulmonary artery59.5 ± 12.958.3 ± 14.40.239
      Signal-to-noise ratio
      Right upper lobe59.3 ± 17.553.1 ± 23.50.150
      Left upper lobe58.6 ± 16.566.2 ± 24.00.112
      Right middle lobe62.6 ± 19.762.2 ± 22.70.940
      Left lingula62.6 ± 16.471.7 ± 24.90.063
      Right lower lobe54.1 ± 19.359.0 ± 24.00.286
      Left lower lobe59.9 ± 20.061.9 ± 27.40.695
      Mean56.8 ± 12.857.3 ± 15.20.985
      LDCT=low-dose CT; FD=flat-panel detector.

      4.4 Sensitivity and specificity of lesion detection

      In 20 patients, a total of 165 lesions were identified: 27 micronodules, 40 nodules, 30 parenchymal opacifications (23 consolidations and 7 ground-glass opacities), 46 cases of bronchiectasis, 16 linear opacities, and six pleural abnormalities (Table 4). The detection rate of parenchymal opacifications was higher than the detection rates of other findings for both clinical and compact FD low-dose CT images (Fig. 1). The pooled sensitivity of the clinical CT images was significantly higher than that of the compact FD CT images in terms of micronodules (0.83 [45/54] vs. 0.61 [33/54]; p = 0.007) (Fig. 2) and bronchiectasis (0.91 [84/92] vs. 0.77 [71/92]; p = 0.004) (Fig. 3). There were two false-positive micronodules on compact FD CT images in one reader, in which the pulmonary vessels were mistaken due to a motion artifact and poor image quality, respectively.
      Table 4Sensitivity and specificity for lesion detectability between clinical and compact flat panel detector low dose CT scans.
      Clinical LDCTCompact FD LDCT
      Reader 1Reader 2Pooled readerReader 1Reader 2Pooled readerP-value
      Sensitivity
      Micronodules (≤3 mm)0.89 (24/27)0.78 (21/27)0.83 (45/54)0.67 (18/27)0.56 (15/27)0.61 (33/54)0.007
      Nodule0.73 (29/40)0.63 (25/40)0.68 (54/80)0.60 (24/40)0.63 (25/40)0.61 (49/80)0.648
      Parenchymal opacification0.97 (29/30)0.87 (26/30)0.92 (55/60)0.90 (27/30)0.77 (23/30)0.83 (50/60)0.125
      Bronchiectasis0.93 (43/46)0.89 (41/46)0.91 (84/92)0.85 (39/46)0.70 (32/46)0.77 (61/92)0.004
      Linear opacity0.69 (11/16)0.69 (11/16)0.69 (22/32)0.50 (8/16)0.44 (7/16)0.47 (15/32)0.065
      Pleural effusion or thickening0.83 (5/6)0.83 (5/6)0.83 (10/12)1.00 (6/6)0.33 (2/6)0.67 (8/12)0.625
      Specificity
      Micronodules (≤3 mm)1.00 (74/74)1.00 (74/74)1.00 (148/148)0.97 (72/74)1.00 (74/74)0.99 (146/148)0.125
      Nodule1.00 (71/71)1.00 (71/71)1.00 (142/142)1.00 (71/71)1.00 (71/71)1.00 (142/142)1.000
      Parenchymal opacification1.00 (71/71)1.00 (71/71)1.00 (142/142)1.00 (71/71)1.00 (71/71)1.00 (142/142)1.000
      Bronchiectasis1.00 (54/54)1.00 (54/54)1.00 (108/108)1.00 (54/54)1.00 (54/54)1.00 (108/108)1.000
      Linear opacity1.00 (84/84)1.00 (84/84)1.00 (168/168)1.00 (84/84)1.00 (84/84)1.00 (168/168)1.000
      Pleural effusion or thickening1.00 (94/94)1.00 (94/94)1.00 (188/188)1.00 (94/94)1.00 (94/94)1.00 (188/188)1.000
      LDCT=low-dose CT; FD=flat-panel detector.
      Fig. 1
      Fig. 1Clinical and compact flat-panel detector low-dose CT images in a 75-year old woman. Clinical (A,C) and compact flat-panel detector low-dose CT (B,D) images show nodular consolidation (arrows) and branching opacities (arrowheads) in the right upper and lower lobes. The clinical CT images had better quality than the flat-panel detector CT images, but both sets of CT images depicted all parenchymal abnormalities.
      Fig. 2
      Fig. 2Clinical and compact flat-panel detector low-dose CT images in a 64-year old woman with nontuberculous mycobacterial lung disease. A clinical low-dose CT image (A) shows multiple clustered tiny nodules in the right middle lobe and bronchiectasis in the left lingula. The multiple tiny nodules (arrows) on the clinical CT image were poorly visualized on a compact flat-panel CT image (B) due to the limited spatial resolution.
      Fig. 3
      Fig. 3Clinical and compact flat-panel detector low-dose CT images in a 71-year old man with interstitial lung disease. A clinical low-dose CT image (A) shows reticular opacities, traction bronchiectasis (arrow), and ground-glass opacities in the bilateral peripheral lungs. A compact flat-panel detector low-dose CT image (B) suffered from respiratory motion and streak artifacts. The two readers inconsistently assessed the presence of traction bronchiectasis in multiple lobes on flat-panel detector CT images.

      4.5 Correlation and reproducibility of NTM CT scoring

      There was no significant difference in NTM scores between the clinical and compact FD low-dose CT images (p > 0.05) regarding the total score and each item of bronchiectasis, cellular bronchiolitis, nodules, consolidations, and cavities (Table 5, Fig. 4). The ICCs between the CT images were roughly 0.950 or higher for the total score and each item, except bronchiectasis (0.869). The 95 % limits of agreement in CT scores between the CT images were − 0.95–2.52 in assessing bronchiectasis, and the 95 % limits of agreement for other items were smaller than ± 1. The 95 % limits of agreement of the total CT score between clinical CT and compact FD CT images were − 0.92–2.83 (Supplementary Figure 1), and bronchiectasis preferentially contributed to the degree of difference in the total CT score between the CT image sets.
      Table 5Correlation and reproducibility for NTM CT scoring between clinical and compact flat panel detector low dose CT scans.
      Mean CT scoreCorrelationMeasurement difference
      Clinical

      LDCT
      Compact

      FD LDCT
      P valueICC95 % CIP-valueMean ± SD95 % limits of agreement
      Bronchiectasis5.2 ± 1.34.4 ± 1.3< 0.0010.8690.670, 0.948< 0.0010.79 ± 0.89-0.95–2.52
      Cellular bronchiolitis4.6 ± 1.74.4 ± 1.70.3070.9870.968, 0.995< 0.0010.12 ± 0.39-0.64–0.88
      Cavity2.7 ± 3.62.7 ± 3.61.0001.0001.000, 1.000NA0.00 ± 0.000.00–0.00
      Nodule0.4 ± 0.50.4 ± 0.50.3170.9460.865, 0.979< 0.0010.05 ± 0.22-0.39–0.49
      Consolidation1.1 ± 1.01.1 ± 1.01.0001.0001.000, 1.000NA0.00 ± 0.000.00–0.00
      Total score14.0 ± 6.313.0 ± 6.5< 0.0010.9940.986, 0.998< 0.0010.96 ± 0.96-0.92–2.83
      Total score except bronchiectasis8.8 ± 5.58.6 ± 5.50.1710.9980.996, 0.999< 0.0010.17 ± 0.43-0.68–1.02
      LDCT=low-dose CT; FD=flat-panel detector; ICC=intraclass correlation coefficient; CI=confidence interval; SD=standard deviation; NA=Not applicable.
      Fig. 4
      Fig. 4Clinical and compact flat-panel detector low-dose CT images in a 76-year old man with nontuberculous mycobacterial lung disease. Clinical (A,C) and compact flat-panel detector CT (B,D) images show bronchiolitis (arrows) and bronchiectasis (arrowheads) in the right upper, middle, and left lower lobes. The CT score was lower in the flat-panel detector CT images than in the clinical CT images for the extent of bronchiectasis, but the CT scores for cellular bronchiolitis extent were the same on both scans.

      4.6 CT radiation dose

      The mean volume CT dose index and dose-length products were 0.5 mGy and 18.2 ± 1.6 mGy·cm in the clinical low-dose CT scans, and the corresponding values were 2.5 mGy and 57.7 mGy·cm in the compact FD low-dose CT scans (p < 0.001).

      5. Discussion

      This prospective study showed that compact FD low-dose CT images had visually lower but quantitatively comparable image quality to clinical state-of-the-art low-dose CT images. Regarding lesion detectability, the FD CT images were significantly inferior for visualizing micronodules and bronchiectasis, whereas nodules, parenchymal opacifications, linear opacities, and pleural lesions were equally to non-significantly less frequently depicted. In addition, the evaluation of the extent of various NTM abnormalities on the FD CT images tended to be lower for bronchiectasis and comparable for cellular bronchiolitis, cavities, nodules, and consolidations. Those results indicate that the FD low-dose CT images provided sufficient diagnostic capability for most parenchymal abnormalities with increased density, but were insufficient to depict minute nodules and bronchiectasis due to their lower image quality and the presence of artifacts.
      FD CT images have the potential to be deployed promptly as a POC diagnostic tool when the next pandemic occurs. Viral pneumonia is responsible for recent major pandemics and will likely account for the next pandemic [
      • Ruuskanen O.
      • Lahti E.
      • Jennings L.C.
      • Murdoch D.R.
      Viral pneumonia.
      ,
      • Zhou P.
      • Yang X.L.
      • Wang X.G.
      • Hu B.
      • Zhang L.
      • Zhang W.
      • et al.
      A pneumonia outbreak associated with a new coronavirus of probable bat origin.
      ]. Viral pneumonia usually shows diffuse lung involvement, and its CT findings mainly comprise ground-glass opacities, consolidation, and nodules [
      • Koo H.J.
      • Lim S.
      • Choe J.
      • Choi S.H.
      • Sung H.
      • Do K.H.
      Radiographic and ct features of viral pneumonia.
      ]. Those findings generally could be depicted on the FD low-dose CT scans. Furthermore, the weight and the minimum space of FD CT scanner are one-fifth or less of clinical CT scanners. FD CT scanners use a standard household voltage, whereas clinical scanners typically require a power generator. In particular, the smaller size and weight are advantageous in intensive care units where radiologic examinations are indicated in patients at risk for deterioration [
      • Rubin G.D.
      • Ryerson C.J.
      • Haramati L.B.
      • Sverzellati N.
      • Kanne J.P.
      • Raoof S.
      • et al.
      The role of chest imaging in patient management during the covid-19 pandemic: a multinational consensus statement from the fleischner society.
      ], but are less accessible in the pandemic setting.
      The mean CT radiation dose of the FD CT images was less than 3 mGy, but five times higher than that of the clinical CT images. The tube voltage and the tube current of the FD CT scans were 120 kVp and 10 mAs, respectively, which were in the range of ultra-low-dose CT scanning (around 0.3 mGy) in clinical CT scans [
      • Svahn T.M.
      • Sjoberg T.
      • Ast J.C.
      Dose estimation of ultra-low-dose chest ct to different sized adult patients.
      ]. The higher CT radiation despite the low-exposure parameters in the FD CT images seems to be associated with longer radiation exposure. The FD CT scanning required two sequential scans of the upper and lower thoraces, which took around 30 s in total due to lower image efficiency and temporal resolution. Nevertheless, these two factors have room for improvement, and an updated version of the FD CT scanner enabled improved efficiency by using better reconstruction and faster rotation of FD (not shown).
      This study has several limitations. The number of patients was relatively small, and it was challenging to enroll more patients even without COVID-19 in the middle of the COVID-19 pandemic. We could not apply FD CT scanning in patients with COVID-19 as it required complex sanitation procedures, and departmental policy did not allow FD CT scanning for research purposes during the pandemic. Besides, our study population did not include patients with acute pneumonia. Before applying FD CT scanning to such patients, it was required to first examine if FD CT scans were feasible for patients with non-acute pulmonary diseases. Our results warrant further extensive validation, and we are about to begin the next prospective investigation in patients with acute pneumonia with the upgrade FD CT scanner. We could not evaluate parenchymal abnormalities with decreased density, such as emphysema or air-trapping. Moreover, we did not evaluate the diagnostic capability of the FD CT images for COVID-19 or viral pneumonia or whether their use would affect patient management in practice. Lastly, clinical low-dose CT scans used as reference images might have a limitation in evaluating ground-glass opacity or subtle abnormalities compared to standard-dose CT images. It was somewhat inevitable to minimize radiation exposure of sequential CT scanning to the enrolled patients under the ‘As Low As Reasonably Achievable’ principle, and our result warrants further validation.
      In conclusion, the FD low-dose CT images provided visually lower image quality, but comparable detectability to clinical low-dose CT images for nodules and parenchymal opacifications. The current limitations of the FD CT images include a relatively higher CT radiation dose and a limited ability to detect bronchiectasis and micronodules. Nevertheless, considering the portability of FD CT scanners, FD low-dose CT scans may be a potential POC imaging tool in future pandemics.

      Ethical statement

      This single-arm prospective trial was approved by our institutional review board, and informed consent was obtained from all participants. The study protocol was registered in the Korean Clinical Information Service (Registration Number: KCT0005633).

      Ethics approval

      This single-arm prospective trial was approved by the institutional review board of Seoul National University Hospital (IRB No. 2010–014–1161), and informed consent was obtained from all participants.

      Funding statement

      This work was supported by the Korean Medical Device Development Fund grant funded by the Korean Government (the Ministry of Science and ICT, the Ministry of Trade Industry and Energy, the Ministry of Health & Welfare, Republic of Korea, the Ministry of Food and Drug Safety ) (Project Number: 202011A03).

      CRediT authorship contribution statement

      Soon Ho Yoon: Conceptualization, Data curation, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision. Da Som Kim: Data curation, Formal analysis, Methodology, Software. Nakwon Kwak: Data curation, Writing – review & editing. Jae-Joon Yim: Data curation. Seung-Jin Yoo: Validation, Visualization, Writing – review & editing. Jung Hee Hong: Validation, Visualization, Writing-review & editing.

      Conflicts of interest

      Outside of this study, Soon Ho Yoon works in MEDICAL IP as a chief medical officer and has a stock option of the firm. Other authors had no conflicts of interest to declare.

      Data Availability

      The datasets generated or analyzed during the study are available from the corresponding author on reasonable request.

      Acknowledgments

      The authors would like to acknowledge Andrew Dombrowski, PhD (Compecs, Inc.) for his assistance in improving the use of English in this manuscript.

      Appendix A. Supplementary material

      References

        • Rubin G.D.
        • Ryerson C.J.
        • Haramati L.B.
        • Sverzellati N.
        • Kanne J.P.
        • Raoof S.
        • et al.
        The role of chest imaging in patient management during the covid-19 pandemic: a multinational consensus statement from the fleischner society.
        Radiology. 2020; 296: 172-180
        • Kim H.
        • Hong H.
        • Yoon S.H.
        Diagnostic performance of ct and reverse transcriptase polymerase chain reaction for coronavirus disease 2019: a meta-analysis.
        Radiology. 2020; 296: E145-E155
        • Chen Q.
        • Zu Z.Y.
        • Jiang M.D.
        • Lu L.
        • Lu G.M.
        • Zhang L.J.
        Infection control and management strategy for covid-19 in the radiology department: focusing on experiences from china.
        Korean J. Radio. 2020; 21: 851-858
        • Kalender W.A.
        • Kyriakou Y.
        Flat-detector computed tomography (fd-ct).
        Eur. Radio. 2007; 17: 2767-2779
        • Rho J.Y.
        • Yoon K.H.
        • Jeong S.
        • Lee J.H.
        • Park C.
        • Kim H.W.
        Usefulness of mobile computed tomography in patients with coronavirus disease 2019 pneumonia: a case series.
        Korean J. Radio. 2020; 21: 1018-1023
        • Kim H.S.
        • Lee K.S.
        • Koh W.J.
        • Jeon K.
        • Lee E.J.
        • Kang H.
        • et al.
        Serial ct findings of mycobacterium massiliense pulmonary disease compared with mycobacterium abscessus disease after treatment with antibiotic therapy.
        Radiology. 2012; 263: 260-270
        • Ruuskanen O.
        • Lahti E.
        • Jennings L.C.
        • Murdoch D.R.
        Viral pneumonia.
        Lancet. 2011; 377: 1264-1275
        • Zhou P.
        • Yang X.L.
        • Wang X.G.
        • Hu B.
        • Zhang L.
        • Zhang W.
        • et al.
        A pneumonia outbreak associated with a new coronavirus of probable bat origin.
        Nature. 2020; 579: 270-273
        • Koo H.J.
        • Lim S.
        • Choe J.
        • Choi S.H.
        • Sung H.
        • Do K.H.
        Radiographic and ct features of viral pneumonia.
        Radiographics. 2018; 38: 719-739
        • Svahn T.M.
        • Sjoberg T.
        • Ast J.C.
        Dose estimation of ultra-low-dose chest ct to different sized adult patients.
        Eur. Radio. 2019; 29: 4315-4323