The Feasibility of Deep Learning-Based Reconstruction for Low-Tube-Voltage CT Angiography for Transcatheter Aortic Valve Implantation.

OBJECTIVE: The purpose of this study is to evaluate the efficacy of deep learning reconstruction (DLR) on low-tube-voltage computed tomographic angiography (CTA) for transcatheter aortic valve implantation (TAVI). METHODS: We enrolled 30 patients who underwent TAVI-CT on a 320-row CT scanner. Electrocardiogram-gated coronary CTA (CCTA) was performed at 100 kV, followed by nongated aortoiliac CTA at 80 kV using a single bolus of contrast material. We used hybrid-iterative reconstruction (HIR), model-based IR (MBIR), and DLR to reconstruct these images. The contrast-to-noise ratios (CNRs) were calculated. Five-point scales were used for the overall image quality analysis. The diameter of the aortic annulus was measured in each reconstructed image, and we compared the interobserver and intraobserver agreements. RESULTS: In the CCTA, the CNR and image quality score for DLR were significantly higher than those for HIR and MBIR ( P < 0.01). In the aortoiliac CTA, the CNR for DLR was significantly higher than that for HIR ( P < 0.01) and significantly lower than that for MBIR ( P ≤ 0.02). The image quality score for DLR was significantly higher than that for HIR ( P < 0.01). No significant differences were observed between the image quality scores for DLR and MBIR. The measured aortic annulus diameter had high interobserver and intraobserver agreement regardless of the reconstruction method (all intraclass correlation coefficients, >0.89). CONCLUSIONS: In low tube voltage TAVI-CT, DLR provides higher image quality than HIR, and DLR provides higher image quality than MBIR in CCTA and is visually comparable to MBIR in aortoiliac CTA.

Reproducibility of quantitative ADC, T1, and T2 measurement on the cerebral cortex: Utility of whole brain echo-planar DWI with compressed SENSE (EPICS-DWI): A pilot study.

Purpose: To assess the reproducibility of ADC, T1, T2, and proton density (PD) measurements on the cortex across the entire brain using high-resolution pseudo-3D diffusion-weighted imaging using echo-planar imaging with compressed SENSE (EPICS-DWI) and 3D quantification with an interleaved Look-Locker acquisition sequence with T2 preparation pulse (3D-QALAS) in normal healthy adults. Methods: Twelve healthy participants (median age, 33 years; range, 28-51 years) were recruited to evaluate the reproducibility of whole-brain EPICS-DWI and synthetic MRI. EPICS-DWI utilizes a compressed SENSE reconstruction framework while maintaining the EPI sampling pattern. The 3D-QALAS sequence is based on multi-acquisition 3D gradient echo, with five acquisitions equally spaced in time, interleaved with a T2 preparation pulse and an inversion pulse. EPICS-DWI (b values, 0 and 1000 s/mm2) and 3D-QALAS sequence with identical voxel size on a 3.0-T MR system were performed twice (for test-retest scan). Intraclass correlation coefficients (ICCs) for ADC, T1, T2, and PD for all parcellated volume of interest (VOI) per subject on scan-rescan tests were calculated to assess reproducibility. Bland-Altman plots were used to investigate discrepancies in ADCs, T1s, T2s, and PDs obtained from the two MR scans. Results: The ICC of ADCs was 0.785, indicating "good" reproducibility. The ICCs of T1s, T2s, and PDs were 0.986, 0.978, and 0.968, indicating "excellent" reproducibility. Conclusion: The combination of EPICS-DWI and 3D-QALAS sequences with identical voxel size could reproducible ADC, T1, T2, and PD measurements for the cortex across the entire brain in healthy adults.

A deep learning model based on fusion images of chest radiography and X-ray sponge images supports human visual characteristics of retained surgical items detection.

PURPOSE: Although a novel deep learning software was proposed using post-processed images obtained by the fusion between X-ray images of normal post-operative radiography and surgical sponge, the association of the retained surgical item detectability with human visual evaluation has not been sufficiently examined. In this study, we investigated the association of retained surgical item detectability between deep learning and human subjective evaluation. METHODS: A deep learning model was constructed from 2987 training images and 1298 validation images, which were obtained from post-processing of the image fusion between X-ray images of normal post-operative radiography and surgical sponge. Then, another 800 images were used, i.e., 400 with and 400 without surgical sponge. The detection characteristics of retained sponges between the model and a general observer with 10-year clinical experience were analyzed using the receiver operator characteristics. RESULTS: The following values from the deep learning model and observer were, respectively, derived: Cutoff values of probability were 0.37 and 0.45; areas under the curves were 0.87 and 0.76; sensitivity values were 85% and 61%; and specificity values were 73% and 92%. CONCLUSION: For the detection of surgical sponges, we concluded that the deep learning model has higher sensitivity, while the human observer has higher specificity. These characteristics indicate that the deep learning system that is complementary to humans could support the clinical workflow in operation rooms for prevention of retained surgical items.

Assessment of cerebral perfusion in moyamoya disease with dynamic pseudo-continuous arterial spin labeling using a variable repetition time scheme with optimized background suppression.

PURPOSE: Accurate assessment of cerebral perfusion in moyamoya disease is necessary to determine the indication for treatment. We aimed to investigate the usefulness of dynamic PCASL using a variable TR scheme with optimized background suppression in the evaluation of cerebral perfusion in moyamoya disease. METHODS: We retrospectively analyzed the images of 24 patients (6 men and 18 women, mean age 31.4 ± 18.2 years) with moyamoya disease; each of whom was imaged with both dynamic PCASL using the variable-TR scheme and 123IMP SPECT with acetazolamide challenge. ASL dynamic data at 10 phases are acquired by changing the LD and PLD. The background suppression timing was optimized for each phase. CBF and ATT were measured with ASL, and CBF and CVR to an acetazolamide challenge were measured with SPECT. RESULTS: A significant moderate correlation was found between the CBF measured by dynamic PCASL and that by SPECT (r = 0.53, P < 0.001). The CBF measured by dynamic PCASL (52.5 ± 13.3 ml/100 mg/min) was significantly higher than that measured by SPECT (43.0 ± 12.6 ml/100 mg/min, P < 0.001). The ATT measured by dynamic PCASL showed a significant correlation with the CVR measured by SPECT (r = 0.44, P < 0.001). ATT was significantly longer in areas where the CVR was impaired (CVR < 18.4%, ATT = 1812 ± 353 ms) than in areas where it was preserved (CVR > 18.4%, ATT = 1301 ± 437 ms, P < 0.001). The ROC analysis showed a moderate accuracy (AUC = 0.807, sensitivity = 87.7%, specificity = 70.4%) when the cutoff value of ATT was set at 1518 ms. CONCLUSION: Dynamic PCASL using this scheme was found to be useful for assessing cerebral perfusion in moyamoya disease.

Intravoxel incoherent motion diffusion-weighted imaging (IVIM-DWI) of the major salivary glands: an assessment of the optimal number and combination of b-values.

Purpose: To examine the optimal number and combination of b-values in intravoxel incoherent motion (IVIM) diffusionweighted imaging (DWI) of the major salivary glands. Material and methods: IVIM-DWI was performed on 10 healthy volunteers using 13 b-values (low b-values: 0-100 s/mm2; high b-values: 200-1000 s/mm2). The IVIM parameters and apparent diffusion coefficient of the bilateral major salivary glands were calculated using 13 b-values and were considered the standard values. We sequentially reduced the number of b-values to 10, 8, 6, and 5. The parameters in each combination were calculated. The standard values were compared with the parameters from each reduced b-value in IVIM-DWI. The Wilcoxon signed-rank test was used to determine whether there were any differences between the parameters in each combination. Bonferroni correction was conducted for multiple comparisons. Results: There were no significant differences between the standard values and parameters from the 2 combinations of 6 b-values. However, significant differences were observed between the standard values and parameters from some combinations of only 2 low or only 2 high b-values. Conclusions: IVIM-DWI of the major salivary glands could be performed using a minimum of 6 b-values. However, they should contain 3 low and 3 high b-values.

Importance of the heart rate in ultra-high-resolution coronary CT angiography with 0.35 s gantry rotation time.

PURPOSE: We investigated the effects of the heart rate (HR) on the motion artifact in coronary computed tomography angiography (CCTA) with ultra-high-resolution-CT (U-HRCT), and we clarified the upper limit of optimal HR in CCTA with U-HRCT in a comparison with conventional-resolution-CT (CRCT) on a cardiac phantom and in patients with CCTA. MATERIALS AND METHODS: A pulsating cardiac phantom equipped with coronary models was scanned at static and HR simulations of 40-90 beats/min (bpm) at 10-bpm intervals using U-HRCT and CRCT, respectively. The sharpness and lumen diameter of the coronary model were quantitatively compared between U-HRCT and CRCT stratified by HR in the phantom study. We also assessed the visual inspections of clinical images in CCTA with U-HRCT. RESULTS: At the HRs ≤ 60 bpm, the error of the lumen diameter of the U-HRCT tended to be smaller than that of the CRCT. However, at the HRs > 60 bpm, the inverse was shown. For the image sharpness, the U-HRCT was significantly superior to the CRCT (p < 0.05). In the visual assessment, the scores were negatively correlated with HRs in patients (Spearman r = - 0.71, p < 0.01). A receiver-operating characteristic analysis revealed the HR of 61 bpm as the optimal cutoff of the non-diagnostic image quality, with an area under the curve of 0.87, 95% sensitivity, and 71% specificity. CONCLUSION: At HRs ≤ 60 bpm, U-HRCT was more accurate in the imaging of coronary arteries than CRCT. The upper limit of the optimal HR in CCTA with U-HRCT was approx. 60 bpm.

Low-radiation dose scan protocol for preoperative imaging for dental implant surgery using deep learning-based reconstruction in multidetector CT.

OBJECTIVES: This study aimed to investigate the impact of a deep learning-based reconstruction (DLR) technique on image quality and reduction of radiation exposure, and to propose a low-dose multidetector-row computed tomography (MDCT) scan protocol for preoperative imaging for dental implant surgery. METHODS: The PB-1 phantom and a Catphan phantom 600 were scanned using volumetric scanning with a 320-row MDCT scanner. All scans were performed with a tube voltage of 120 kV, and the tube current varied from 120 to 60 to 40 to 30 mA. Images of the mandible were reconstructed using DLR. Additionally, images acquired with the 120-mA protocol were reconstructed using filtered back projection as a reference. Two observers independently graded the image quality of the mandible images using a 4-point scale (4, superior to reference; 1, unacceptable). The system performance function (SPF) was calculated to comprehensively evaluate image quality. The Wilcoxon signed-rank test was employed for statistical analysis, with statistical significance set at p value < 0.05. RESULTS: There was no significant difference between the image quality acquired with the 40-mA tube current and reconstructed with the DLR technique (40DLR), and that acquired with the reference protocol (3.00, 3.00, p = 1.00). The SPF at 1.0 cycles/mm acquired with 40DLR was improved by 156.7% compared to that acquired with the reference protocol. CONCLUSIONS: Our proposed protocol, which achieves a two-thirds reduction in radiation dose, can provide a minimally invasive MDCT scan of acceptable image quality for dental implant surgery.

Deep-learning reconstruction for ultra-low-dose lung CT: Volumetric measurement accuracy and reproducibility of artificial ground-glass nodules in a phantom study.

OBJECTIVES: The lung nodule volume determined by CT is used for nodule diagnoses and monitoring tumor responses to therapy. Increased image noise on low-dose CT degrades the measurement accuracy of the lung nodule volume. We compared the volumetric accuracy among deep-learning reconstruction (DLR), model-based iterative reconstruction (MBIR), and hybrid iterative reconstruction (HIR) at an ultra-low-dose setting. METHODS: Artificial ground-glass nodules (6 mm and 10 mm diameters, -660 HU) placed at the lung-apex and the middle-lung field in chest phantom were scanned by 320-row CT with the ultra-low-dose setting of 6.3 mAs. Each scan data set was reconstructed by DLR, MBIR, and HIR. The volumes of nodules were measured semi-automatically, and the absolute percent volumetric error (APEvol) was calculated. The APEvol provided by each reconstruction were compared by the Tukey-Kramer method. Inter- and intraobserver variabilities were evaluated by a Bland-Altman analysis with limits of agreements. RESULTS: DLR provided a lower APEvol compared to MBIR and HIR. The APEvol of DLR (1.36%) was significantly lower than those of the HIR (8.01%, p = 0.0022) and MBIR (7.30%, p = 0.0053) on a 10-mm-diameter middle-lung nodule. DLR showed narrower limits of agreement compared to MBIR and HIR in the inter- and intraobserver agreement of the volumetric measurement. CONCLUSIONS: DLR showed higher accuracy compared to MBIR and HIR for the volumetric measurement of artificial ground-glass nodules by ultra-low-dose CT. ADVANCES IN KNOWLEDGE: DLR with ultra-low-dose setting allows a reduction of dose exposure, maintaining accuracy for the volumetry of lung nodule, especially in patients which deserve a long-term follow-up.

Image quality improvement with deep learning-based reconstruction on abdominal ultrahigh-resolution CT: A phantom study.

PURPOSE: In an ultrahigh-resolution CT (U-HRCT), deep learning-based reconstruction (DLR) is expected to drastically reduce image noise without degrading spatial resolution. We assessed a new algorithm's effect on image quality at different radiation doses assuming an abdominal CT protocol. METHODS: For the normal-sized abdominal models, a Catphan 600 was scanned by U-HRCT with 100%, 50%, and 25% radiation doses. In all acquisitions, DLR was compared to model-based iterative reconstruction (MBIR), filtered back projection (FBP), and hybrid iterative reconstruction (HIR). For the quantitative assessment, we compared image noise, which was defined as the standard deviation of the CT number, and spatial resolution among all reconstruction algorithms. RESULTS: Deep learning-based reconstruction yielded lower image noise than FBP and HIR at each radiation dose. DLR yielded higher image noise than MBIR at the 100% and 50% radiation doses (100%, 50%, DLR: 15.4, 16.9 vs MBIR: 10.2, 15.6 Hounsfield units: HU). However, at the 25% radiation dose, the image noise in DLR was lower than that in MBIR (16.7 vs. 26.6 HU). The spatial frequency at 10% of the modulation transfer function (MTF) in DLR was 1.0 cycles/mm, slightly lower than that in MBIR (1.05 cycles/mm) at the 100% radiation dose. Even when the radiation dose decreased, the spatial frequency at 10% of the MTF of DLR did not change significantly (50% and 25% doses, 0.98 and 0.99 cycles/mm, respectively). CONCLUSION: Deep learning-based reconstruction performs more consistently at decreasing dose in abdominal ultrahigh-resolution CT compared to all other commercially available reconstruction algorithms evaluated.

Metal artefact reduction in the oral cavity using deep learning reconstruction algorithm in ultra-high-resolution computed tomography: a phantom study.

OBJECTIVES: This study aimed to improve the impact of the metal artefact reduction (MAR) algorithm for the oral cavity by assessing the effect of acquisition and reconstruction parameters on an ultra-high-resolution CT (UHRCT) scanner. METHODS: The mandible tooth phantom with and without the lesion was scanned using super-high-resolution, high-resolution (HR), and normal-resolution (NR) modes. Images were reconstructed with deep learning-based reconstruction (DLR) and hybrid iterative reconstruction (HIR) using the MAR algorithm. Two dental radiologists independently graded the degree of metal artefact (1, very severe; 5, minimum) and lesion shape reproducibility (1, slight; 5, almost perfect). The signal-to-artefact ratio (SAR), accuracy of the CT number of the lesion, and image noise were calculated quantitatively. The Tukey-Kramer method with a p-value of less than 0.05 was used to determine statistical significance. RESULTS: The HRDLR visual score was better than the NRHIR score in terms of degree of metal artefact (4.6 ± 0.5 and 2.6 ± 0.5, p < 0.0001) and lesion shape reproducibility (4.5 ± 0.5 and 2.9 ± 1.1, p = 0.0005). The SAR of HRDLR was significantly better than that of NRHIR (4.9 ± 0.4 and 2.1 ± 0.2, p < 0.0001), and the absolute percentage error of the CT number in HRDLR was lower than that in NRHIR (0.8% in HRDLR and 23.8% in NRIR). The image noise of HRDLR was lower than that of NRHIR (15.7 ± 1.4 and 51.6 ± 15.3, p < 0.0001). CONCLUSIONS: Our study demonstrated that the combination of HR mode and DLR in UHRCT scanner improved the impact of the MAR algorithm in the oral cavity.

Improved scan method for dental imaging using multidetector computed tomography: a phantom study.

OBJECTIVES: This study aimed to propose an improved scan method to shorten irradiation time and reduce radiation exposure. METHODS: The maxilla of a human head CT phantom and a Catphan phantom were used for qualitative and quantitative assessment, respectively. The phantoms were scanned by a 160-row multidetector CT scanner using volumetric and helical scanning. In volumetric scanning, the tube current varied from 120 to 60 to 30 to 20 mA with a tube voltage of 120 kV. Images were reconstructed with a bone kernel using iterative reconstruction (IR) and filtered back projection. As a reference protocol, helical scanning was performed using our clinical setting with 120 kV. Two dental radiologists independently graded the quality of dental images using a 4-point scale (4, superior to reference; 1, unacceptable). For the quantitative assessment, we assessed the system performance from each scan. RESULTS: There was no significant difference between the image quality of volumetric scanning using the 60 mA protocol reconstructed with IR and that of the reference (3.08 and 3.00, p = 0.3388). The system performance values at 1.0 cycles/mm of volumetric scanning and 60 mA protocol reconstructed with IR and reference were 0.0038 and 0.0041, respectively. The effective dose of volumetric scanning using the 60 mA protocol was 51.8 µSv, which is a 64.2% reduction to that of the reference. CONCLUSIONS: We proposed an improved scan method resulting in a 64.2% reduction of radiation dose with one-fourth of irradiation time by combining volumetric scanning and IR technique in multidetector CT.

Correction to: Development of a new phantom simulating extracellular space of tumor cell growth and cell edema for diffusion-weighted magnetic resonance imaging.

The original version of this article unfortunately contained a mistake. Second column of "Cell edema" should read as.

Effect of scan mode and focal spot size in airway dimension measurements for ultra-high-resolution computed tomography of chronic obstructive pulmonary disease: A COPDGene phantom study.

PURPOSE: Quantitative evaluations of airway dimensions through computed tomography (CT) have revealed a good correlation with airflow limitation in chronic obstructive pulmonary disease. However, large inaccuracies have been known to occur in CT airway measurements. Ultra-high-resolution CT (UHRCT) might improve measurement accuracy using precise scan modes with minimal focal spot. We assessed the effects of scan mode and focal spot size on airway measurements in UHRCT. METHODS: COPDGene Ⅱ phantom, comprising a plastic tube mimicking human airway of inner diameter 3 mm, wall thickness 0.6 mm, and inclination 30 degrees was scanned at super high resolution (SHR, beam collimation of 0.25 mm × 160 rows) and high resolution (HR, beam collimation of 0.5 mm × 80 rows) modes using UHRCT. Each acquisition was performed both with small (0.4 × 0.5 mm) and large (0.6 × 1.3 mm) focal spots. The wall area percentage (WA%) was calculated as the percentage of total airway area occupied by the airway wall. Statistical analysis was performed to compare the WA% measurement errors for each scan mode and focal spot size. RESULTS: The WA% measurement errors in the SHR mode were 9.8% with a small focal spot and 18.8% with a large one. The measurement errors in the HR mode were 13.3% with a small focal spot and 21.4% with a large one. There were significant differences between each scan mode and focal spot size (p < 0.05). CONCLUSIONS: The SHR mode with a small focal spot could improve airway measurement accuracy of UHRCT.

Development of a new phantom simulating extracellular space of tumor cell growth and cell edema for diffusion-weighted magnetic resonance imaging.

OBJECTIVE: A phantom for diffusion-weighted imaging is required to standardize quantitative evaluation. The objectives were to develop a phantom simulating various cell densities and to evaluate repeatability. MATERIALS AND METHODS: The acrylic fine particles with three different diameters were used to simulate human cells. Four-degree cell density components were developed by adjusting the volume of 10-µm particles (5, 20, 35, and 50% volume, respectively). Two-degree components to simulate cell edema were also developed by adjusting the diameter without changing number (17% and 40% volume, respectively). Spearman's rank correlation coefficient was used to find a significant correlation between apparent diffusion coefficient (ADC) and particle density. Coefficient of variation (CV) for ADC was calculated for each component for 6 months. A p value < 0.05 represented a statistically significance. RESULTS: Each component (particle ratio of 5, 17, 20, 35, 40, and 50% volume, respectively) presented ADC values of 1.42, 1.30, 1.30, 1.12, 1.09, and 0.89 (× 10-3 mm2/s), respectively. A negative correlation (r = - 0.986, p < 0.05) was observed between ADC values and particle ratio. CV for ADC was less than 5%. DISCUSSION: A phantom simulating the diffusion restriction correlating with cell density and size could be developed.

The reproducibility of measurements using a standardization phantom for the evaluation of fractional anisotropy (FA) derived from diffusion tensor imaging (DTI).

OBJECTIVES: It is necessary to standardize the examination procedure and diagnostic criteria of diffusion tensor imaging (DTI). Thus, the purpose of this study was to examine the reproducibility of measurements using a standardization phantom composed of different fibre materials with different fibre densities (FDs) for the evaluation of fractional anisotropy (FA) derived from DTI. MATERIALS AND METHODS: Two types of fibre materials wrapped in heat-shrinkable tubes were used as fibre phantoms. We designed fibre phantoms with three different FDs of each fibre material. The standardization phantom was examined using DTI protocol six times a day, and each examination session was repeated once a month for 7 consecutive months. Fibre tracking was performed by setting regions of interest in the FA map, and FA was measured in each fibre phantom. Coefficients of variation (CVs) were used to evaluate the inter-examination reproducibility of FA values. Furthermore, Bland-Altman plots were used to evaluate the intra-operator reproducibility of FA measurements. RESULTS: All CVs for each fibre phantom were within 2% throughout the 7-month study of repeated DTI sessions. The high intra-operator reproducibility of the FA measurement was confirmed. DISCUSSION: High reproducibility of measurements using a standardization phantom for the evaluation of FA was achieved.

[Improvement of Image Quality in the Axial Section Using High-resolution Scan Mode and Hybrid Iterative Reconstruction in Ultra-high-resolution Computed Tomography].

The purpose of this study is to compare the physical characteristics and visibility of high-resolution and conventional images acquired with the same X-ray dose, and to investigate the superiority of super high-resolution imaging. A Catphan phantom was scanned in the normal resolution (NR), high-resolution (HR), and super high-resolution (SHR) modes of ultra-high-resolution computed tomography at 120 kV and 75 mAs. All images were reconstructed into a 5-mm thick image slices with filtered back-projection (FBP) and hybrid image reconstruction (HIR), which included normal and enhanced adaptive iterative dose reduction 3D (AIDR and eAIDR, respectively). The modulation transfer function (MTF) and noise power spectrum (NPS) were measured using the circular edge method and radial frequency method, respectively. The signal-to-noise ratio (SNR) was then calculated. High-contrast resolution and low-contrast detectability were evaluated visually by five radiological technologists. The MTFs of HReAIDR and HRFBP images were higher than those of NRFBP images. However, the NPSs of HReAIDR and HRFBP images were larger than those of NRFBP images. The SNR of HReAIDR images was higher than that of NRFBP and HRFBP images. The scores of high-contrast resolution of HReAIDR, NRFBP, and HRFBP images were 13, 8, and 13 cycles/cm, respectively, and the scores of low-contrast detectability were 5, 5, and 6 mm, respectively. Hence, an improvement in high-contrast resolution of signal more than 400 HU in the axial section can be achieved without increasing the radiation dose and decreasing low-contrast detectability with 10 HU using the HR mode and eAIDR.

Comparison of intravoxel incoherent motion diffusion-weighted imaging between turbo spin-echo and echo-planar imaging of the head and neck.

OBJECTIVES: To compare image quality, apparent diffusion coefficient (ADC), and intravoxel incoherent motion (IVIM)-derived parameters between turbo spin-echo (TSE)-diffusion-weighted imaging (DWI) and echo-planar imaging (EPI)-DWI of the head and neck. METHODS: Fourteen volunteers underwent head and neck imaging using TSE-DWI and EPI-DWI. Distortion ratio (DR), signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), ADC and IVIM-derived parameters were compared between the two techniques. Bland-Altman analysis was performed to analyse reproducibility between the quantitative parameters of TSE-DWI and EPI-DWI. RESULTS: DR of TSE-DWI was significantly smaller than that of EPI-DWI. SNR and CNR of TSE-DWI were significantly higher than those of EPI-DWI. ADC and IVIM-derived parameters of TSE-DWI showed higher values than those of EPI-DWI, although the difference was not significant. Bland-Altman analysis showed wide limits of agreement between the two sequences. CONCLUSION: TSE-DWI can produce better image quality than EPI-DWI, while TSE-DWI possibly exhibits different values of quantitative parameters. Therefore, TSE-DWI could be a good alternative to EPI-DWI for patients sensitive to distortion. However, it is not recommended to use both TSE-DWI and EPI-DWI on follow-up. KEY POINTS: • Head and neck DWI is especially sensitive to magnetic inhomogeneity. • The distortion of images was less with TSE-DWI than with EPI-DWI. • TSE-DWI can possibly exhibit higher ADC and IVIM-derived parameters than EPI-DWI. • Bland-Altman analysis showed unacceptable LoA in quantitative analysis between TSE-DWI and EPI-DWI. • It is not recommended to use both TSE-DWI and EPI-DWI for follow-up.