Deep learning reconstruction improves image quality of abdominal ultra-high-resolution CT

Deep learning enables time-efficient soft tissue enhancement in CBCT: Proof-of-concept study for dentomaxillofacial applications

PURPOSE

The use of iterative and deep learning reconstruction methods, which would allow effective noise reduction, is limited in cone-beam computed tomography (CBCT). As a consequence, the visibility of soft tissues is limited with CBCT. The study aimed to improve this issue through time-efficient deep learning enhancement (DLE) methods.

METHODS

Two DLE networks, UNIT and U-Net, were trained with simulated CBCT data. The performance of the networks was tested with three different test data sets. The quantitative evaluation measured the structural similarity index measure (SSIM) and the peak signal-to-noise ratio (PSNR) of the DLE reconstructions with respect to the ground truth iterative reconstruction method. In the second assessment, a dentomaxillofacial radiologist assessed the resolution of hard tissue structures, visibility of soft tissues, and overall image quality of real patient data using the Likert scale. Finally, the technical image quality was determined using modulation transfer function, noise power spectrum, and noise magnitude analyses.

RESULTS

The study demonstrated that deep learning CBCT denoising is feasible and time efficient. The DLE methods, trained with simulated CBCT data, generalized well, and DLE provided quantitatively (SSIM/PSNR) and visually similar noise-reduction as conventional IR, but with faster processing time. The DLE methods improved soft tissue visibility compared to the conventional Feldkamp-Davis-Kress (FDK) algorithm through noise reduction. However, in hard tissue quantification tasks, the radiologist preferred the FDK over the DLE methods.

CONCLUSION

Post-reconstruction DLE allowed feasible reconstruction times while yielding improvements in soft tissue visibility in each dataset.

Patient-derived PixelPrint phantoms for evaluating clinical imaging performance of a deep learning CT reconstruction algorithm

Objective

Deep learning reconstruction (DLR) algorithms exhibit object-dependent resolution and noise performance. Thus, traditional geometric CT phantoms cannot fully capture the clinical imaging performance of DLR. This study uses a patient-derived 3D-printed PixelPrint lung phantom to evaluate a commercial DLR algorithm across a wide range of radiation dose levels.

Approach

The lung phantom used in this study is based on a patient chest CT scan containing ground glass opacities and was fabricated using PixelPrint 3D-printing technology. The phantom was placed inside two different sized extension rings to mimic a small and medium sized patient and was scanned on a conventional CT scanner at exposures between 0.5 and 20 mGy. Each scan was reconstructed using filtered back projection (FBP), iterative reconstruction, and DLR at five levels of denoising. Image noise, contrast to noise ratio (CNR), root mean squared error (RMSE), structural similarity index (SSIM), and multi-scale SSIM (MS SSIM) were calculated for each image.

Main Results

DLR demonstrated superior performance compared to FBP and iterative reconstruction for all measured metrics in both phantom sizes, with better performance for more aggressive denoising levels. DLR was estimated to reduce dose by 25-83% in the small phantom and by 50-83% in the medium phantom without decreasing image quality for any of the metrics measured in this study. These dose reduction estimates are more conservative compared to the estimates obtained when only considering noise and CNR with a non-anatomical physics phantom.

Significance

DLR has the capability of producing diagnostic image quality at up to 83% lower radiation dose which can improve the clinical utility and viability of lower dose CT scans. Furthermore, the PixelPrint phantom used in this study offers an improved testing environment with more realistic tissue structures compared to traditional CT phantoms, allowing for structure-based image quality evaluation beyond noise and contrast-based assessments.

Effectiveness of deep learning reconstruction on standard to ultra-low-dose high-definition chest CT images

PURPOSE

Deep learning reconstruction (DLR) has been introduced by major vendors, tested for CT examinations of a variety of organs, and compared with other reconstruction methods. The purpose of this study was to compare the capabilities of DLR for image quality improvement and lung texture evaluation with those of hybrid-type iterative reconstruction (IR) for standard-, reduced- and ultra-low-dose CTs (SDCT, RDCT and ULDCT) obtained with high-definition CT (HDCT) and reconstructed at 0.25-mm, 0.5-mm and 1-mm section thicknesses with 512 × 512 or 1024 × 1024 matrixes for patients with various pulmonary diseases.

MATERIALS AND METHODS

Forty age-, gender- and body mass index-matched patients with various pulmonary diseases underwent SDCT (CT dose index volume : mean ± standard deviation, 9.0 ± 1.8 mGy), RDCT (CTDIvol: 1.7 ± 0.2 mGy) and ULDCT (CTDIvol: 0.8 ± 0.1 mGy) at a HDCT. All CT data set were then reconstructed with 512 × 512 or 1024 × 1024 matrixes by means of hybrid-type IR and DLR. SNR of lung parenchyma and probabilities of all lung textures were assessed for each CT data set. SNR and detection performance of each lung texture reconstructed with DLR and hybrid-type IR were then compared by means of paired t tests and ROC analyses for all CT data at each section thickness.

RESULTS

Data for each radiation dose showed DLR attained significantly higher SNR than hybrid-type IR for each of the CT data (p < 0.0001). On assessments of all findings except consolidation and nodules or masses, areas under the curve (AUCs) for ULDCT with hybrid-type IR for each section thickness (0.91 ≤ AUC ≤ 0.97) were significantly smaller than those with DLR (0.97 ≤ AUC ≤ 1, p < 0.05) and the standard protocol (0.98 ≤ AUC ≤ 1, p < 0.05).

CONCLUSION

DLR is potentially more effective for image quality improvement and lung texture evaluation than hybrid-type IR on all radiation dose CTs obtained at HDCT and reconstructed with each section thickness with both matrixes for patients with a variety of pulmonary diseases.

Improving image quality with super-resolution deep-learning-based reconstruction in coronary CT angiography

OBJECTIVES

To evaluate the effect of super-resolution deep-learning-based reconstruction (SR-DLR) on the image quality of coronary CT angiography (CCTA).

METHODS

Forty-one patients who underwent CCTA using a 320-row scanner were retrospectively included. Images were reconstructed with hybrid (HIR), model-based iterative reconstruction (MBIR), normal-resolution deep-learning-based reconstruction (NR-DLR), and SR-DLR algorithms. For each image series, image noise, and contrast-to-noise ratio (CNR) at the left main trunk, right coronary artery, left anterior descending artery, and left circumflex artery were quantified. Blooming artifacts from calcified plaques were measured. Image sharpness, noise magnitude, noise texture, edge smoothness, overall quality, and delineation of the coronary wall, calcified and noncalcified plaques, cardiac muscle, and valves were subjectively ranked on a 4-point scale (1, worst; 4, best). The quantitative parameters and subjective scores were compared among the four reconstructions. Task-based image quality was assessed with a physical evaluation phantom. The detectability index for the objects simulating the coronary lumen, calcified plaques, and noncalcified plaques was calculated from the noise power spectrum (NPS) and task-based transfer function (TTF).

RESULTS

SR-DLR yielded significantly lower image noise and blooming artifacts with higher CNR than HIR, MBIR, and NR-DLR (all p < 0.001). The best subjective scores for all the evaluation criteria were attained with SR-DLR, with significant differences from all other reconstructions (p < 0.001). In the phantom study, SR-DLR provided the highest NPS average frequency, TTF50%, and detectability for all task objects.

CONCLUSION

SR-DLR considerably improved the subjective and objective image qualities and object detectability of CCTA relative to HIR, MBIR, and NR-DLR algorithms.

CLINICAL RELEVANCE STATEMENT

The novel SR-DLR algorithm has the potential to facilitate accurate assessment of coronary artery disease on CCTA by providing excellent image quality in terms of spatial resolution, noise characteristics, and object detectability.

KEY POINTS

• SR-DLR designed for CCTA improved image sharpness, noise property, and delineation of cardiac structures with reduced blooming artifacts from calcified plaques relative to HIR, MBIR, and NR-DLR. • In the task-based image-quality assessments, SR-DLR yielded better spatial resolution, noise property, and detectability for objects simulating the coronary lumen, coronary calcifications, and noncalcified plaques than other reconstruction techniques. • The image reconstruction times of SR-DLR were shorter than those of MBIR, potentially serving as a novel standard-of-care reconstruction technique for CCTA performed on a 320-row CT scanner.

Stress-only dynamic computed tomography perfusion protocol (CTP) alone without computed tomography coronary angiography (CCTA) has limited specificity to diagnose ischemia: A retrospective two-center study

PURPOSE

To investigate diagnostic performance of stress-only dynamic myocardial computed tomography perfusion (CTP) without computed tomography coronary angiography (CCTA) to diagnose ischemia with invasive fractional flow reserve (FFR) as a reference standard.

METHOD

135 datasets (68 positive for ischemia with invasive FFR < 0.8) acquired with a 256-slice CT system (Revolution, GE Healthcare, Chicago, IL, USA) were retrieved, postprocessed with a deep learning-based algorithm (Advanced intelligent Clear-IQ Engine (AiCE), Canon Medical Systems, Otawara, Japan) (FC03/cardiac kernel, 8 mm slice thickness), analyzed using a dedicated workstation (Vitrea research 7.11.0. Vital Images, Minnetonka, MN, USA), and loaded into a clinical workstation (CardIQ, GE Healthcare, Chicago, IL, USA) for review. Ten observers with various experience from two research sites evaluated the post-processed images, perfusion slices and maps to indicate presence vs absence of perfusion defect and its probability (five-point Likert scale). Binary decisions and probability scores were used to calculate sensitivity and specificity for each reader, and to create receiver operating characteristics (ROC) curves, respectively. Furthermore, the correlation coefficient (ICC) was computed. ROC AUC of a purely quantitative analysis was obtained thanks to a color-coded map with a fixed scale superimposed on myocardial walls displaying myocardial blood flow (MBF) values.

RESULTS

The overall case-based sensitivity and specificity for the detection of perfusion deficit were 0.79 and 0.30, respectively. No significant differences were detected in the AUC across readers (p value = 0.66). The AUC values were 0.50, 0.58, 0.63, 0.59, 0.45, 0.60, 0.56, 0.61, 0.52, 0.61. Absolute reader agreement ICC was 0.60 (good agreement) for an average case.

CONCLUSION

Dynamic CTP alone has good sensitivity, but low specificity when analyzed without CCTA. These findings reinforce the need to guide the interpretation functional test with the knowledge of coronary artery anatomy.

Evaluation of three-dimensional dual-energy CT cholangiopancreatography image quality in patients with pancreatobiliary dilatation: Comparison with conventional single-energy CT

Objective

This study aimed to evaluate three-dimensional (3D) negative-contrast CT cholangiopancreatography (nCTCP) image quality using dual-energy CT (DECT) with iterative reconstruction (IR) technique in patients with pancreatobiliary dilatation compared with single-energy CT (SECT).

Methods

Of the patients, 67 and 56 underwent conventional SECT (SECT set) and DECT with IR technique (DECT set), respectively. All patients were retrospectively analyzed during the portal phase to compare objective image quality and other data including patient demographics, hepatic and pancreatic parenchymal enhancement, noise, and attenuation difference (AD) between dilated ducts and enhanced hepatic parenchyma, signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and CT volume dose index (CTDIvol). Two radiologists used the five-point Likert scale to evaluate the subjective image quality of 3D nCTCP regarding image noise, sharpness of dilated ducts, and overall image quality. Statistical analyses used the Mann-Whitney U test.

Results

No significant difference in patient demographics in either CT set was showed during objective evaluation (p > 0.05). However, higher hepatic and pancreatic parenchymal enhancement, AD, SNR, and CNR and lower hepatic and pancreatic noise (p < 0.005) as well as CTDIvol (p = 0.005) on DECT than on SECT were observed. Higher mean grades on DECT than on SECT were showed for image noise (4.65 vs 3.92), sharpness of dilated ducts (4.52 vs 3.94), and overall image quality (4.45 vs 3.91; p < 0.001), respectively during subjective evaluation.

Conclusion

A higher overall image quality and lower radiation dose on 3D nCTCP can be obtained by DECT with IR technique than with conventional SECT in patients with pancreatobiliary dilatation.

Deep learning imaging reconstruction of reduced-dose 40 keV virtual monoenergetic imaging for early detection of colorectal cancer liver metastases

OBJECTIVE

To explore whether reduced-dose (RD) gemstone spectral imaging (GSI) and deep learning image reconstruction (DLIR) of 40 keV virtual monoenergetic image (VMI) enhanced the early detection and diagnosis of colorectal cancer liver metastases (CRLM).

METHODS

Thirty-five participants with pathologically confirmed colorectal cancer were prospectively enrolled from March to August 2022 after routine care abdominal computed tomography (CT). GSI mode was used for contrast-enhanced CT, and two portal venous phase CT images were obtained [standard-dose (SD) CT dose index (CTDIvol) = 15.51 mGy, RD CTDIvol = 7.95 mGy]. The 40 keV-VMI were reconstructed via filtered back projection (FBP) and iterative reconstruction (ASIR-V 60 %, AV60) of both SD and RD images. RD medium-strength deep learning image reconstruction (DLIR-M) and RD high-strength deep learning image reconstruction (DLIR-H) were used to reconstruct the 40 keV-VMI. The contrast-to-noise ratio (CNR) and signal-to-noise ratio (SNR) of the liver and the lesions were objectively evaluated. The overall image quality, lesion conspicuity, and diagnostic confidence were subjectively evaluated, to compare the differences in evaluation results among the different images.

RESULTS

All 35 participants (mean age: 59.51 ± 11.01 years; 14 females) underwent SD and RD GSI portal venous-phase CT scans. The dose-length product of the RD GSI scan was reduced by 49-53 % lower than that of the SD GSI scan (420.22 ± 31.95) vs (817.58 ± 60.56). A total of 219 lesions were identified, including 55 benign lesions and 164 metastases, with an average size of 7.37 ± 4.14 mm. SD-FBP detected 207 lesions, SD-AV60 detected 201 lesions, and DLIR-M and DLIR-H detected 199 and 190 lesions, respectively. For lesions ≤ 5 mm, there was no statistical difference between SD-FBP vs DLIR-M (χ2McNemar = 1.00, P = 0.32) and SD-AV60 vs DLIR-M (χ2McNemar = 0.33, P = 0.56) in the detection rate. The CNR, SNR, and noise of DLIR-M and DLIR-H 40 keV-VMI images were better than those of SD-FBP images (P < 0.01) but did not differ significantly from those of SD-AV60 images (P > 0.05). When the lesions ≤ 5 mm, there were statistical differences in the overall diagnostic sensitivity of lesions compared with SD-FBP, SD-AV60, DLIR-M and DLIR-H (P＜0.01). There were no statistical differences in the sensitivity of lesions diagnosis between SD-FBP, SD-AV60 and DLIR-M (both P＞0.05). However, the DLIR-M subjective image quality and lesion diagnostic confidence were higher for SD-FBP (both P < 0.01).

CONCLUSION

Reduced dose DLIR-M of 40 keV-VMI can be used for routine follow-up care of colorectal cancer patients, to optimize evaluations and ensure CT image quality. Meanwhile, the detection rate and diagnostic sensitivity and specificity of small lesions, early liver metastases is not obviously reduced.

Clinical feasibility of deep learning reconstruction in liver diffusion-weighted imaging: Improvement of image quality and impact on apparent diffusion coefficient value

PURPOSE

Diffusion-weighted imaging (DWI) of the liver suffers from low resolution, noise, and artifacts. This study aimed to investigate the effect of deep learning reconstruction (DLR) on image quality and apparent diffusion coefficient (ADC) quantification of liver DWI at 3 Tesla.

METHOD

In this prospective study, images of the liver obtained at DWI with b-values of 0 (DWI0), 50 (DWI50) and 800 s/mm2 (DWI800) from consecutive patients with liver lesions from February 2022 to February 2023 were reconstructed with and without DLR (non-DLR). Image quality was assessed qualitatively using Likert scoring system and quantitatively using signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR) and liver/parenchyma boundary sharpness from region-of-interest (ROI) analysis. ADC value of lesion were measured. Phantom experiment was also performed to investigate the factors that determine the effect of DLR on ADC value. Qualitative score, SNR, CNR, boundary sharpness, and apparent diffusion coefficients (ADCs) for DWI were compared using paired t-test and Wilcoxon signed rank test. P < 0.05 was considered statistically significant.

RESULTS

A total of 85 patients with 170 lesions were included. DLR group showed a higher qualitative score than the non-DLR group. for example, with DWI800 the score was 4.77 ± 0.52 versus 4.30 ± 0.63 (P < 0.001). DLR group also showed higher SNRs, CNRs and boundary sharpness than the non-DLR group. DLR reduced the ADC of malignant tumors (1.105[0.904, 1.340] versus 1.114[0.904, 1.320]) (P < 0.001), but there was no significant difference in the diagnostic value of malignancy for DLR and non-DLR groups (P = 57.3). The phantom study confirmed a reduction of ADC in images with low resolution, and a stronger reduction of ADC in heterogeneous structures than in homogeneous ones (P < 0.001).

CONCLUSIONS

DLR improved image quality of liver DWI. DLR reduced the ADC value of lesions, but did not affect the diagnostic performance of ADC in distinguishing malignant tumors on a 3.0-T MRI system.

Impact of a Deep Learning-based Super-resolution Image Reconstruction Technique on High-contrast Computed Tomography: A Phantom Study

RATIONALE AND OBJECTIVES

Deep-learning-based super-resolution image reconstruction (DLSRR) is a novel image reconstruction technique that is expected to contribute to improvement in spatial resolution as well as noise reduction through learning from high-resolution computed tomography (CT). This study aims to evaluate image quality obtained with DLSRR and assess its clinical potential.

MATERIALS AND METHODS

CT images of a Mercury CT 4.0 phantom were obtained using a 320-row multi-detector scanner at tube currents of 100, 200, and 300 mA. Image data were reconstructed by filtered back projection (FBP), hybrid iterative reconstruction (HIR), model-based iterative reconstruction (MBIR), deep-learning-based image reconstruction (DLR), and DLSRR at image reconstruction strength levels of mild, standard, and strong. Noise power spectrum (NPS), task transfer function (TTF), and detectability index were calculated.

RESULTS

The magnitude of the noise-reducing effect in comparison with FBP was in the order MBIR

CONCLUSION

The present results suggest that DLSRR can achieve greater noise reduction and improved spatial resolution in the high-contrast region compared with conventional DLR and iterative reconstruction techniques.

Collapse

Key Words

• deep-learning-based super-resolution image reconstruction
• detectability index
• iterative reconstruction
• noise power spectrum
• task transfer function

Collapse

MESH Headings Collapse

Grants Collapse

Affiliation(s)

Hideyuki Sato Department of Radiological Technology, Juntendo University Hospital, Tokyo, Japan
Shinichiro Fujimoto Department of Cardiovascular Biology and Medicine, Juntendo University Graduate School of Medicine, Tokyo, Japan.
Nobuo Tomizawa Department of Radiology, Juntendo University Graduate School of Medicine, Tokyo, Japan
Hidekazu Inage Department of Radiological Technology, Juntendo University Hospital, Tokyo, Japan
Takuya Yokota Department of Radiological Technology, Juntendo University Hospital, Tokyo, Japan
Hikaru Kudo Department of Radiological Technology, Juntendo University Hospital, Tokyo, Japan
Ruiheng Fan Department of Radiological Technology, Juntendo University Hospital, Tokyo, Japan
Keiichi Kawamoto Department of Radiological Technology, Juntendo University Hospital, Tokyo, Japan
Yuri Honda Department of Radiological Technology, Juntendo University Hospital, Tokyo, Japan
Takayuki Kobayashi Department of Radiological Technology, Kitasato University Kitasato Institute Hospital, Tokyo, Japan
Tohru Minamino Department of Cardiovascular Biology and Medicine, Juntendo University Graduate School of Medicine, Tokyo, Japan
Yosuke Kogure Department of Radiological Technology, Juntendo University Hospital, Tokyo, Japan

Collapse

Collaborators Collapse

Value of deep-learning image reconstruction at submillisievert CT for evaluation of the female pelvis

AIM

To assess the value of deep-learning reconstruction (DLR) at submillisievert computed tomography (CT) for the evaluation of the female pelvis, with standard dose (SD) hybrid iterative reconstruction (IR) images as reference.

MATERIALS AND METHODS

The present study enrolled 50 female patients consecutively who underwent contrast-enhanced abdominopelvic CT for clinically indicated reasons. Submillisievert pelvic images were acquired using a noise index of 15 for low-dose (LD) scans, which were reconstructed with DLR (body and body sharp), hybrid-IR, and model-based IR (MBIR). Additionally, SD scans were reconstructed with a noise index of 7.5 using hybrid-IR. Radiation dose, quantitative image quality, overall image quality, image appearance using a five-point Likert scale (1-5: worst to best), and lesion evaluation in both SD and LD images were analysed and compared.

RESULTS

The submillisievert pelvic CT examinations showed a 61.09 ± 4.13% reduction in the CT dose index volume compared to SD examinations. Among the LD images, DLR (body sharp) had the highest quantitative quality, followed by DLR (body), MBIR, and hybrid-IR. LD DLR (body) had overall image quality comparable to the reference (p=0.084) and favourable image appearance (p=0.209). In total, 40 pelvic lesions were detected in both SD and LD images. LD DLR (body and body sharp) exhibited similar diagnostic confidence (p=0.317 and 0.096) compared with SD hybrid-IR.

CONCLUSION

DLR algorithms, providing comparable image quality and diagnostic confidence, are feasible in submillisievert abdominopelvic CT. The DLR (body) algorithm with favourable image appearance is recommended in clinical settings.

Detection of Vertebral Mass and Diagnosis of Spinal Cord Compression in Computed Tomography With Deep Learning Reconstruction: Comparison With Hybrid Iterative Reconstruction

PURPOSE

To compare the impact of deep learning reconstruction (DLR) and hybrid-iterative reconstruction (hybrid-IR) on vertebral mass depiction, detection, and diagnosis of spinal cord compression on computed tomography (CT).

METHODS

This retrospective study included 29 and 20 patients with and without vertebral masses. CT images were reconstructed using DLR and hybrid-IR. Three readers performed vertebral mass detection tests and evaluated the presence of spinal cord compression, the depiction of vertebral masses, and image noise. Quantitative image noise was measured by placing regions of interest on the aorta and spinal cord.

RESULTS

Deep learning reconstruction tended to improve the performance of readers with less diagnostic imaging experience in detecting vertebral masses (area under the receiver operating characteristic curve [AUC] = .892-.966) compared with hybrid-IR (AUC = .839-.917). Diagnostic performance in evaluating spinal cord compression in DLR (AUC = .887-.995) also improved compared with that in hybrid-IR (AUC = .866-.942) for some readers. The depiction of vertebral masses and subjective image noise in DLR were significantly improved compared with those in hybrid-IR (P < .041). Quantitative image noise in DLR was also significantly reduced compared with that in hybrid-IR (P < .001).

CONCLUSION

Deep learning reconstruction improved the depiction of vertebral masses, which resulted in a tendency to improve the performance of CT compared to hybrid-IR in detecting vertebral masses and diagnosing spinal cord compression for some readers.

Implementation of AI image reconstruction in CT-how is it validated and what dose reductions can be achieved

CT reconstruction has undergone a substantial change over the last decade with the introduction of iterative reconstruction (IR) and now with deep learning reconstruction (DLR). In this review, DLR will be compared to IR and filtered back-projection (FBP) reconstructions. Comparisons will be made using image quality metrics such as noise power spectrum, contrast-dependent task-based transfer function, and non-prewhitening filter detectability index (dNPW'). Discussion on how DLR has impacted CT image quality, low-contrast detectability, and diagnostic confidence will be provided. DLR has shown the ability to improve in areas that IR is lacking, namely: noise magnitude reduction does not alter noise texture to the degree that IR did, and the noise texture found in DLR is more aligned with noise texture of an FBP reconstruction. Additionally, the dose reduction potential for DLR is shown to be greater than IR. For IR, the consensus was dose reduction should be limited to no more than 15-30% to preserve low-contrast detectability. For DLR, initial phantom and patient observer studies have shown acceptable dose reduction between 44 and 83% for both low- and high-contrast object detectability tasks. Ultimately, DLR is able to be used for CT reconstruction in place of IR, making it an easy "turnkey" upgrade for CT reconstruction. DLR for CT is actively being improved as more vendor options are being developed and current DLR options are being enhanced with second generation algorithms being released. DLR is still in its developmental early stages, but is shown to be a promising future for CT reconstruction.

Data drift in medical machine learning: implications and potential remedies

Data drift refers to differences between the data used in training a machine learning (ML) model and that applied to the model in real-world operation. Medical ML systems can be exposed to various forms of data drift, including differences between the data sampled for training and used in clinical operation, differences between medical practices or context of use between training and clinical use, and time-related changes in patient populations, disease patterns, and data acquisition, to name a few. In this article, we first review the terminology used in ML literature related to data drift, define distinct types of drift, and discuss in detail potential causes within the context of medical applications with an emphasis on medical imaging. We then review the recent literature regarding the effects of data drift on medical ML systems, which overwhelmingly show that data drift can be a major cause for performance deterioration. We then discuss methods for monitoring data drift and mitigating its effects with an emphasis on pre- and post-deployment techniques. Some of the potential methods for drift detection and issues around model retraining when drift is detected are included. Based on our review, we find that data drift is a major concern in medical ML deployment and that more research is needed so that ML models can identify drift early, incorporate effective mitigation strategies and resist performance decay.

Can 1.25 mm thin-section images generated with Deep Learning Image Reconstruction technique replace standard-of-care 5 mm images in abdominal CT?

BACKGROUND

CT image reconstruction has evolved from filtered back projection to hybrid- and model-based iterative reconstruction. Deep learning-based image reconstruction is a relatively new technique that uses deep convolutional neural networks to improve image quality.

OBJECTIVE

To evaluate and compare 1.25 mm thin-section abdominal CT images reconstructed with deep learning image reconstruction (DLIR) with 5 mm thick images reconstructed with adaptive statistical iterative reconstruction (ASIR-V).

METHODS

This retrospective study included 52 patients (31 F; 56.9±16.9 years) who underwent abdominal CT scans between August-October 2019. Image reconstruction was performed to generate 5 mm images at 40% ASIR-V and 1.25 mm DLIR images at three strengths (low [DLIR-L], medium [DLIR-M], and high [DLIR-H]). Qualitative assessment was performed to determine image noise, contrast, visibility of small structures, sharpness, and artifact based on a 5-point-scale. Image preference determination was based on a 3-point-scale. Quantitative assessment included measurement of attenuation, image noise, and contrast-to-noise ratios (CNR).

RESULTS

Thin-section images reconstructed with DLIR-M and DLIR-H yielded better image quality scores than 5 mm ASIR-V reconstructed images. Mean qualitative scores of DLIR-H for noise (1.77 ± 0.71), contrast (1.6 ± 0.68), small structure visibility (1.42 ± 0.66), sharpness (1.34 ± 0.55), and image preference (1.11 ± 0.34) were the best (p<0.05). DLIR-M yielded intermediate scores. All DLIR reconstructions showed superior ratings for artifacts compared to ASIR-V (p<0.05), whereas each DLIR group performed comparably (p>0.05, 0.405-0.763). In the quantitative assessment, there were no significant differences in attenuation values between all reconstructions (p>0.05). However, DLIR-H demonstrated the lowest noise (9.17 ± 3.11) and the highest CNR (CNRliver = 26.88 ± 6.54 and CNRportal vein = 7.92 ± 3.85) (all p<0.001).

CONCLUSION

DLIR allows generation of thin-section (1.25 mm) abdominal CT images, which provide improved image quality with higher inter-reader agreement compared to 5 mm thick images reconstructed with ASIR-V.

CLINICAL IMPACT

Improved image quality of thin-section CT images reconstructed with DLIR has several benefits in clinical practice, such as improved diagnostic performance without radiation dose penalties.

Milestones in CT: Past, Present, and Future

In 1971, the first patient CT examination by Ambrose and Hounsfield paved the way for not only volumetric imaging of the brain but of the entire body. From the initial 5-minute scan for a 180° rotation to today's 0.24-second scan for a 360° rotation, CT technology continues to reinvent itself. This article describes key historical milestones in CT technology from the earliest days of CT to the present, with a look toward the future of this essential imaging modality. After a review of the beginnings of CT and its early adoption, the technical steps taken to decrease scan times-both per image and per examination-are reviewed. Novel geometries such as electron-beam CT and dual-source CT have also been developed in the quest for ever-faster scans and better in-plane temporal resolution. The focus of the past 2 decades on radiation dose optimization and management led to changes in how exposure parameters such as tube current and tube potential are prescribed such that today, examinations are more customized to the specific patient and diagnostic task than ever before. In the mid-2000s, CT expanded its reach from gray-scale to color with the clinical introduction of dual-energy CT. Today's most recent technical innovation-photon-counting CT-offers greater capabilities in multienergy CT as well as spatial resolution as good as 125 μm. Finally, artificial intelligence is poised to impact both the creation and processing of CT images, as well as automating many tasks to provide greater accuracy and reproducibility in quantitative applications.

75% radiation dose reduction using deep learning reconstruction on low-dose chest CT

OBJECTIVE

Few studies have explored the clinical feasibility of using deep-learning reconstruction to reduce the radiation dose of CT. We aimed to compare the image quality and lung nodule detectability between chest CT using a quarter of the low dose (QLD) reconstructed with vendor-agnostic deep-learning image reconstruction (DLIR) and conventional low-dose (LD) CT reconstructed with iterative reconstruction (IR).

MATERIALS AND METHODS

We retrospectively collected 100 patients (median age, 61 years [IQR, 53-70 years]) who received LDCT using a dual-source scanner, where total radiation was split into a 1:3 ratio. QLD CT was generated using a quarter dose and reconstructed with DLIR (QLD-DLIR), while LDCT images were generated using a full dose and reconstructed with IR (LD-IR). Three thoracic radiologists reviewed subjective noise, spatial resolution, and overall image quality, and image noise was measured in five areas. The radiologists were also asked to detect all Lung-RADS category 3 or 4 nodules, and their performance was evaluated using area under the jackknife free-response receiver operating characteristic curve (AUFROC).

RESULTS

The median effective dose was 0.16 (IQR, 0.14-0.18) mSv for QLD CT and 0.65 (IQR, 0.57-0.71) mSv for LDCT. The radiologists' evaluations showed no significant differences in subjective noise (QLD-DLIR vs. LD-IR, lung-window setting; 3.23 ± 0.19 vs. 3.27 ± 0.22; P = .11), spatial resolution (3.14 ± 0.28 vs. 3.16 ± 0.27; P = .12), and overall image quality (3.14 ± 0.21 vs. 3.17 ± 0.17; P = .15). QLD-DLIR demonstrated lower measured noise than LD-IR in most areas (P < .001 for all). No significant difference was found between QLD-DLIR and LD-IR for the sensitivity (76.4% vs. 72.2%; P = .35) or the AUFROCs (0.77 vs. 0.78; P = .68) in detecting Lung-RADS category 3 or 4 nodules. Under a noninferiority limit of -0.1, QLD-DLIR showed noninferior detection performance (95% CI for AUFROC difference, -0.04 to 0.06).

CONCLUSION

QLD-DLIR images showed comparable image quality and noninferior nodule detectability relative to LD-IR images.

Prediction of Anemia From Cerebral Venous Sinus Attenuation on Deep-Learning Reconstructed Brain Computed Tomography Images

OBJECTIVE

The aim of the study is to evaluate whether the prediction of anemia is possible using quantitative analyses of unenhanced cranial computed tomography (CT) with deep learning reconstruction (DLR) compared with conventional methods.

METHODS

This cross-sectional retrospective study included 116 participants (76 males; mean age, 66.7) who had hemoglobin (Hb) levels obtained within 24 hours of unenhanced cranial CT, which included 2 reconstruction methods: DLR and hybrid iterative reconstruction. Regions of interest were the confluence of sinuses (CoS) and the right and left transverse sinuses. In addition, edge rise distance of cerebrospinal fluid and venous was measured.

RESULTS

Spearman rank correlation coefficient demonstrated a positive association between Hb levels and sinus attenuation values. Among these, the CoS in DLR had the best correlation ( r = 0.703, P < 0.001). For the prediction of anemia (Hb < 11 g/dL), the area under the curve of CoS in DLR (area under the curve = 0.874; 95% confidence interval, 0.798-0.949; P < 0.001) was the highest; however, there were no significant differences among reconstruction method and sinus. The attenuation values of DLR were significantly higher than those of hybrid iterative reconstruction ( P < 0.001, paired t test), and the differences between the 2 methods were 4.1 (standard deviation [SD], 1.6) for CoS, 5.2 (SD, 2.2) for right transverse sinuses, and 5.8 (SD, 2.4) for left transverse sinuses. The signal-to-noise ratio ( P < 0.001, paired t test) and edge rise distance ( P < 0.001, Wilcoxon signed rank test) of DLR was significantly higher.

CONCLUSIONS

Higher CT attenuation values should be considered for predicting anemia based on brain DLR images.

Dose reduction and image enhancement in micro-CT using deep learning

BACKGROUND

In preclinical settings, micro-computed tomography (CT) provides a powerful tool to acquire high resolution anatomical images of rodents and offers the advantage to in vivo non-invasively assess disease progression and therapy efficacy. Much higher resolutions are needed to achieve scale-equivalent discriminatory capabilities in rodents as those in humans. High resolution imaging however comes at the expense of increased scan times and higher doses. Specifically, with preclinical longitudinal imaging, there are concerns that dose accumulation may affect experimental outcomes of animal models.

PURPOSE

Dose reduction efforts under the ALARA (as low as reasonably achievable) principles are thus a key point of attention. However, low dose CT acquisitions inherently induce higher noise levels which deteriorate image quality and negatively impact diagnostic performance. Many denoising techniques already exist, and deep learning (DL) has become increasingly popular for image denoising, but research has mostly focused on clinical CT with limited studies conducted on preclinical CT imaging. We investigate the potential of convolutional neural networks (CNN) for restoring high quality micro-CT images from low dose (noisy) images. The novelty of the CNN denoising frameworks presented in this work consists of utilizing image pairs with realistic CT noise present in the input as well as the target image used for the model training; a noisier image acquired with a low dose protocol is matched to a less noisy image acquired with a higher dose scan of the same mouse.

METHODS

Low and high dose ex vivo micro-CT scans of 38 mice were acquired. Two CNN models, based on a 2D and 3D four-layer U-Net, were trained with mean absolute error (30 training, 4 validation and 4 test sets). To assess denoising performance, ex vivo mice and phantom data were used. Both CNN approaches were compared to existing methods, like spatial filtering (Gaussian, Median, Wiener) and iterative total variation image reconstruction algorithm. Image quality metrics were derived from the phantom images. A first observer study (n = 23) was set-up to rank overall quality of differently denoised images. A second observer study (n = 18) estimated the dose reduction factor of the investigated 2D CNN method.

RESULTS

Visual and quantitative results show that both CNN algorithms exhibit superior performance in terms of noise suppression, structural preservation and contrast enhancement over comparator methods. The quality scoring by 23 medical imaging experts also indicates that the investigated 2D CNN approach is consistently evaluated as the best performing denoising method. Results from the second observer study and quantitative measurements suggest that CNN-based denoising could offer a 2-4× dose reduction, with an estimated dose reduction factor of about 3.2 for the considered 2D network.

CONCLUSIONS

Our results demonstrate the potential of DL in micro-CT for higher quality imaging at low dose acquisition settings. In the context of preclinical research, this offers promising future prospects for managing the cumulative severity effects of radiation in longitudinal studies.

Deep Learning-Based Versus Iterative Image Reconstruction for Unenhanced Brain CT: A Quantitative Comparison of Image Quality

This exploratory retrospective study aims to quantitatively compare the image quality of unenhanced brain computed tomography (CT) reconstructed with an iterative (AIDR-3D) and a deep learning-based (AiCE) reconstruction algorithm. After a preliminary phantom study, AIDR-3D and AiCE reconstructions (0.5 mm thickness) of 100 consecutive brain CTs acquired in the emergency setting on the same 320-detector row CT scanner were retrospectively analyzed, calculating image noise reduction attributable to the AiCE algorithm, artifact indexes in the posterior cranial fossa, and contrast-to-noise ratios (CNRs) at the cortical and thalamic levels. In the phantom study, the spatial resolution of the two datasets proved to be comparable; conversely, AIDR-3D reconstructions showed a broader noise pattern. In the human study, median image noise was lower with AiCE compared to AIDR-3D (4.7 vs. 5.3, p < 0.001, median 19.6% noise reduction), whereas AIDR-3D yielded a lower artifact index than AiCE (7.5 vs. 8.4, p < 0.001). AiCE also showed higher median CNRs at the cortical (2.5 vs. 1.8, p < 0.001) and thalamic levels (2.8 vs. 1.7, p < 0.001). These results highlight how image quality improvements granted by deep learning-based (AiCE) and iterative (AIDR-3D) image reconstruction algorithms vary according to different brain areas.

Bayesian Reconstruction Algorithms for Low-Dose Computed Tomography Are Not Yet Suitable in Clinical Context

Computed tomography (CT) is a widely used examination technique that usually requires a compromise between image quality and radiation exposure. Reconstruction algorithms aim to reduce radiation exposure while maintaining comparable image quality. Recently, unsupervised deep learning methods have been proposed for this purpose. In this study, a promising sparse-view reconstruction method (posterior temperature optimized Bayesian inverse model; POTOBIM) is tested for its clinical applicability. For this study, 17 whole-body CTs of deceased were performed. In addition to POTOBIM, reconstruction was performed using filtered back projection (FBP). An evaluation was conducted by simulating sinograms and comparing the reconstruction with the original CT slice for each case. A quantitative analysis was performed using peak signal-to-noise ratio (PSNR) and structural similarity index measure (SSIM). The quality was assessed visually using a modified Ludewig's scale. In the qualitative evaluation, POTOBIM was rated worse than the reference images in most cases. A partially equivalent image quality could only be achieved with 80 projections per rotation. Quantitatively, POTOBIM does not seem to benefit from more than 60 projections. Although deep learning methods seem suitable to produce better image quality, the investigated algorithm (POTOBIM) is not yet suitable for clinical routine.

Coronary artery calcium quantification: comparison between filtered-back projection, hybrid iterative reconstruction, and deep learning reconstruction techniques

BACKGROUND

The reference protocol for the quantification of coronary artery calcium (CAC) should be updated to meet the standards of modern imaging techniques.

PURPOSE

To assess the influence of filtered-back projection (FBP), hybrid iterative reconstruction (IR), and three levels of deep learning reconstruction (DLR) on CAC quantification on both in vitro and in vivo studies.

MATERIAL AND METHODS

In vitro study was performed with a multipurpose anthropomorphic chest phantom and small pieces of bones. The real volume of each piece was measured using the water displacement method. In the in vivo study, 100 patients (84 men; mean age = 71.2 ± 8.7 years) underwent CAC scoring with a tube voltage of 120 kVp and image thickness of 3 mm. The image reconstruction was done with FBP, hybrid IR, and three levels of DLR including mild (DLRmild), standard (DLRstd), and strong (DLRstr).

RESULTS

In the in vitro study, the calcium volume was equivalent (P = 0.949) among FBP, hybrid IR, DLRmild, DLRstd, and DLRstr. In the in vivo study, the image noise was significantly lower in images that used DLRstr-based reconstruction, when compared images other reconstructions (P < 0.001). There were no significant differences in the calcium volume (P = 0.987) and Agatston score (P = 0.991) among FBP, hybrid IR, DLRmild, DLRstd, and DLRstr. The highest overall agreement of Agatston scores was found in the DLR groups (98%) and hybrid IR (95%) when compared to standard FBP reconstruction.

CONCLUSION

The DLRstr presented the lowest bias of agreement in the Agatston scores and is recommended for the accurate quantification of CAC.

Deep-learning CT reconstruction in clinical scans of the abdomen: a systematic review and meta-analysis

OBJECTIVE

To perform a systematic literature review and meta-analysis of the two most common commercially available deep-learning algorithms for CT.

METHODS

We used PubMed, Scopus, Embase, and Web of Science to conduct systematic searches for studies assessing the most common commercially available deep-learning CT reconstruction algorithms: True Fidelity (TF) and Advanced intelligent Clear-IQ Engine (AiCE) in the abdomen of human participants since only these two algorithms currently have adequate published data for robust systematic analysis.

RESULTS

Forty-four articles fulfilled inclusion criteria. 32 studies evaluated TF and 12 studies assessed AiCE. DLR algorithms produced images with significantly less noise (22-57.3% less than IR) but preserved a desirable noise texture with increased contrast-to-noise ratios and improved lesion detectability on conventional CT. These improvements with DLR were similarly noted in dual-energy CT which was only assessed for a single vendor. Reported radiation reduction potential was 35.1-78.5%. Nine studies assessed observer performance with the two dedicated liver lesion studies being performed on the same vendor reconstruction (TF). These two studies indicate preserved low contrast liver lesion detection (> 5 mm) at CTDIvol 6.8 mGy (BMI 23.5 kg/m2) to 12.2 mGy (BMI 29 kg/m2). If smaller lesion detection and improved lesion characterization is needed, a CTDIvol of 13.6-34.9 mGy is needed in a normal weight to obese population. Mild signal loss and blurring have been reported at high DLR reconstruction strengths.

CONCLUSION

Deep learning reconstructions significantly improve image quality in CT of the abdomen. Assessment of other dose levels and clinical indications is needed. Careful choice of radiation dose levels is necessary, particularly for small liver lesion assessment.

Application of Deep Learning-Based Denoising Technique for Radiation Dose Reduction in Dynamic Abdominal CT: Comparison with Standard-Dose CT Using Hybrid Iterative Reconstruction Method

The purpose is to evaluate whether deep learning-based denoising (DLD) algorithm provides sufficient image quality for abdominal computed tomography (CT) with a 30% reduction in radiation dose, compared to standard-dose CT reconstructed with conventional hybrid iterative reconstruction (IR). The subjects consisted of 50 patients who underwent abdominal CT with standard dose and reconstructed with hybrid IR (ASiR-V50%) and another 50 patients who underwent abdominal CT with approximately 30% less dose and reconstructed with ASiR-V50% and DLD at low-, medium- and high-strength (DLD-L, DLD-M and DLD-H, respectively). The standard deviation of attenuation in liver parenchyma was measured as image noise. Contrast-to-noise ratio (CNR) for portal vein on portal venous phase was calculated. Lesion conspicuity in 23 abdominal solid mass on the reduced-dose CT was rated on a 5-point scale: 0 (best) to -4 (markedly inferior). Compared with hybrid IR of standard-dose CT, DLD-H of reduced-dose CT provided significantly lower image noise (portal phase: 9.0 (interquartile range, 8.7-9.4) HU vs 12.0 (11.4-12.7) HU, P < 0.0001) and significantly higher CNR (median, 5.8 (4.4-7.4) vs 4.3 (3.3-5.3), P = 0.0019). As for DLD-M of reduced-dose CT, no significant difference was found in image noise and CNR compared to hybrid IR of standard-dose CT (P > 0.99). Lesion conspicuity scores for DLD-H and DLD-M were significantly better than hybrid IR (P < 0.05). Dynamic contrast-enhanced abdominal CT acquired with approximately 30% lower radiation dose and generated with the DLD algorithm exhibit lower image noise and higher CNR compared to standard-dose CT with hybrid IR.

Evaluation of Image Quality and Detectability of Deep Learning Image Reconstruction (DLIR) Algorithm in Single- and Dual-energy CT

This study is aimed to evaluate effects of deep learning image reconstruction (DLIR) on image quality in single-energy CT (SECT) and dual-energy CT (DECT), in reference to adaptive statistical iterative reconstruction-V (ASIR-V). The Gammex 464 phantom was scanned in SECT and DECT modes at three dose levels (5, 10, and 20 mGy). Raw data were reconstructed using six algorithms: filtered back-projection (FBP), ASIR-V at 40% (AV-40) and 100% (AV-100) strength, and DLIR at low (DLIR-L), medium (DLIR-M), and high strength (DLIR-H), to generate SECT 120kVp images and DECT 120kVp-like images. Objective image quality metrics were computed, including noise power spectrum (NPS), task transfer function (TTF), and detectability index (d'). Subjective image quality evaluation, including image noise, texture, sharpness, overall quality, and low- and high-contrast detectability, was performed by six readers. DLIR-H reduced overall noise magnitudes from FBP by 55.2% in a more balanced way of low and high frequency ranges comparing to AV-40, and improved the TTF values at 50% for acrylic inserts by average percentages of 18.32%. Comparing to SECT 20 mGy AV-40 images, the DECT 10 mGy DLIR-H images showed 20.90% and 7.75% improvement in d' for the small-object high-contrast and large-object low-contrast tasks, respectively. Subjective evaluation showed higher image quality and better detectability. At 50% of the radiation dose level, DECT with DLIR-H yields a gain in objective detectability index compared to full-dose AV-40 SECT images used in daily practice.

Fast and low-dose medical imaging generation empowered by hybrid deep-learning and iterative reconstruction

Fast and low-dose reconstructions of medical images are highly desired in clinical routines. We propose a hybrid deep-learning and iterative reconstruction (hybrid DL-IR) framework and apply it for fast magnetic resonance imaging (MRI), fast positron emission tomography (PET), and low-dose computed tomography (CT) image generation tasks. First, in a retrospective MRI study (6,066 cases), we demonstrate its capability of handling 3- to 10-fold under-sampled MR data, enabling organ-level coverage with only 10- to 100-s scan time; second, a low-dose CT study (142 cases) shows that our framework can successfully alleviate the noise and streak artifacts in scans performed with only 10% radiation dose (0.61 mGy); and last, a fast whole-body PET study (131 cases) allows us to faithfully reconstruct tumor-induced lesions, including small ones (<4 mm), from 2- to 4-fold-accelerated PET acquisition (30-60 s/bp). This study offers a promising avenue for accurate and high-quality image reconstruction with broad clinical value.

Deep learning techniques in liver tumour diagnosis using CT and MR imaging - A systematic review

Deep learning has become a thriving force in the computer aided diagnosis of liver cancer, as it solves extremely complicated challenges with high accuracy over time and facilitates medical experts in their diagnostic and treatment procedures. This paper presents a comprehensive systematic review on deep learning techniques applied for various applications pertaining to liver images, challenges faced by the clinicians in liver tumour diagnosis and how deep learning bridges the gap between clinical practice and technological solutions with an in-depth summary of 113 articles. Since, deep learning is an emerging revolutionary technology, recent state-of-the-art research implemented on liver images are reviewed with more focus on classification, segmentation and clinical applications in the management of liver diseases. Additionally, similar review articles in literature are reviewed and compared. The review is concluded by presenting the contemporary trends and unaddressed research issues in the field of liver tumour diagnosis, offering directions for future research in this field.

Advanced Neuroimaging With Photon-Counting Detector CT

ABSTRACT

Photon-counting detector computed tomography (PCD-CT) is an emerging technology and promises the next step in CT evolution. Photon-counting detectors count the number of individual incoming photons and assess the energy level of each of them. These mechanisms differ substantially from conventional energy-integrating detectors. The new technique has several advantages, including lower radiation exposure, higher spatial resolution, reconstruction of images with less beam-hardening artifacts, and advanced opportunities for spectral imaging. Research PCD-CT systems have already demonstrated promising results, and recently, the first whole-body full field-of-view PCD-CT scanners became clinically available. Based on published studies of preclinical systems and the first experience with clinically approved scanners, the performance can be translated to valuable neuroimaging applications, including brain imaging, intracranial and extracranial CT angiographies, or head and neck imaging with detailed assessment of the temporal bone. In this review, we will provide an overview of the current status in neuroimaging with upcoming and potential clinical applications.

Deep learning-based reconstruction can improve the image quality of low radiation dose head CT

OBJECTIVES

To evaluate the image quality of deep learning-based reconstruction (DLR), model-based (MBIR), and hybrid iterative reconstruction (HIR) algorithms for lower-dose (LD) unenhanced head CT and compare it with those of standard-dose (STD) HIR images.

METHODS

This retrospective study included 114 patients who underwent unenhanced head CT using the STD (n = 57) or LD (n = 57) protocol on a 320-row CT. STD images were reconstructed with HIR; LD images were reconstructed with HIR (LD-HIR), MBIR (LD-MBIR), and DLR (LD-DLR). The image noise, gray and white matter (GM-WM) contrast, and contrast-to-noise ratio (CNR) at the basal ganglia and posterior fossa levels were quantified. The noise magnitude, noise texture, GM-WM contrast, image sharpness, streak artifact, and subjective acceptability were independently scored by three radiologists (1 = worst, 5 = best). The lesion conspicuity of LD-HIR, LD-MBIR, and LD-DLR was ranked through side-by-side assessments (1 = worst, 3 = best). Reconstruction times of three algorithms were measured.

RESULTS

The effective dose of LD was 25% lower than that of STD. Lower image noise, higher GM-WM contrast, and higher CNR were observed in LD-DLR and LD-MBIR than those in STD (all, p ≤ 0.035). Compared with STD, the noise texture, image sharpness, and subjective acceptability were inferior for LD-MBIR and superior for LD-DLR (all, p < 0.001). The lesion conspicuity of LD-DLR (2.9 ± 0.2) was higher than that of HIR (1.2 ± 0.3) and MBIR (1.8 ± 0.4) (all, p < 0.001). Reconstruction times of HIR, MBIR, and DLR were 11 ± 1, 319 ± 17, and 24 ± 1 s, respectively.

CONCLUSION

DLR can enhance the image quality of head CT while preserving low radiation dose level and short reconstruction time.

KEY POINTS

• For unenhanced head CT, DLR reduced the image noise and improved the GM-WM contrast and lesion delineation without sacrificing the natural noise texture and image sharpness relative to HIR. • The subjective and objective image quality of DLR was better than that of HIR even at 25% reduced dose without considerably increasing the image reconstruction times (24 s vs. 11 s). • Despite the strong noise reduction and improved GM-WM contrast performance, MBIR degraded the noise texture, sharpness, and subjective acceptance with prolonged reconstruction times relative to HIR, potentially hampering its feasibility.

Artificial Intelligence in CT and MR Imaging for Oncological Applications

Cancer care increasingly relies on imaging for patient management. The two most common cross-sectional imaging modalities in oncology are computed tomography (CT) and magnetic resonance imaging (MRI), which provide high-resolution anatomic and physiological imaging. Herewith is a summary of recent applications of rapidly advancing artificial intelligence (AI) in CT and MRI oncological imaging that addresses the benefits and challenges of the resultant opportunities with examples. Major challenges remain, such as how best to integrate AI developments into clinical radiology practice, the vigorous assessment of quantitative CT and MR imaging data accuracy, and reliability for clinical utility and research integrity in oncology. Such challenges necessitate an evaluation of the robustness of imaging biomarkers to be included in AI developments, a culture of data sharing, and the cooperation of knowledgeable academics with vendor scientists and companies operating in radiology and oncology fields. Herein, we will illustrate a few challenges and solutions of these efforts using novel methods for synthesizing different contrast modality images, auto-segmentation, and image reconstruction with examples from lung CT as well as abdome, pelvis, and head and neck MRI. The imaging community must embrace the need for quantitative CT and MRI metrics beyond lesion size measurement. AI methods for the extraction and longitudinal tracking of imaging metrics from registered lesions and understanding the tumor environment will be invaluable for interpreting disease status and treatment efficacy. This is an exciting time to work together to move the imaging field forward with narrow AI-specific tasks. New AI developments using CT and MRI datasets will be used to improve the personalized management of cancer patients.

Ultra-High-Resolution CT of the Head and Neck with Deep Learning Reconstruction-Assessment of Image Quality and Radiation Exposure and Intraindividual Comparison with Normal-Resolution CT

OBJECTIVES

To assess the benefits of ultra-high-resolution CT (UHR-CT) with deep learning-based image reconstruction engine (AiCE) regarding image quality and radiation dose and intraindividually compare it to normal-resolution CT (NR-CT).

METHODS

Forty consecutive patients with head and neck UHR-CT with AiCE for diagnosed head and neck malignancies and available prior NR-CT of a different scanner were retrospectively evaluated. Two readers evaluated subjective image quality using a 5-point Likert scale regarding image noise, image sharpness, artifacts, diagnostic acceptability, and assessability of various anatomic regions. For reproducibility, inter-reader agreement was analyzed. Furthermore, signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and slope of the gray-value transition between different tissues were calculated. Radiation dose was evaluated by comparing CTDIvol, DLP, and mean effective dose values.

RESULTS

UHR-CT with AiCE reconstruction led to significant improvement in subjective (image noise and diagnostic acceptability: p < 0.000; ICC ≥ 0.91) and objective image quality (SNR: p < 0.000; CNR: p < 0.025) at significantly lower radiation doses (NR-CT 2.03 ± 0.14 mSv; UHR-CT 1.45 ± 0.11 mSv; p < 0.0001) compared to NR-CT.

CONCLUSIONS

Compared to NR-CT, UHR-CT combined with AiCE provides superior image quality at a markedly lower radiation dose. With improved soft tissue assessment and potentially improved tumor detection, UHR-CT may add further value to the role of CT in the assessment of head and neck pathologies.

Reducing both radiation and contrast doses for overweight patients in coronary CT angiography with 80-kVp and deep learning image reconstruction

PURPOSE

To investigate the use of an 80-kVp tube voltage combined with a deep learning image reconstruction (DLIR) algorithm in coronary CT angiography (CCTA) for overweight patients to reduce radiation and contrast doses in comparison with the 120-kVp protocol and adaptive statistical iterative reconstruction (ASIR-V).

METHODS

One hundred consecutive CCTA patients were prospectively enrolled and randomly divided into a low-dose group (n = 50) with 80-kVp, smart mA for noise index (NI) of 36 HU, contrast dose rate of 18 mgI/kg/s and DLIR and 60 % ASIR-V and a standard-dose group (n = 50) with 120-kVp, smart mA for NI of 25 HU, contrast dose rate of 32 mgI/kg/s and 60 % ASIR-V. The radiation and contrast dose, subjective image quality score, attenuation values, noise, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) were compared.

RESULTS

The low-dose group achieved a significant reduction in the effective radiation dose (1.01 ± 0.45 mSv vs 1.85 ± 0.40 mSv, P < 0.001) and contrast dose (33.69 ± 3.87 mL vs 59.11 ± 5.60 mL, P < 0.001) compared to the standard-dose group. The low-dose group with DLIR presented similar enhancement but lower noise, higher SNR and CNR and higher subjective quality scores than the standard-dose group. Moreover, the same patient comparison in the low-dose group between different reconstructions showed that DLIR images had slightly and consistently higher CT values in small vessels, indicating better defined vessels, much lower image noise, higher SNR and CNR and higher subjective quality scores than ASIR-V images (all P < 0.001).

CONCLUSIONS

The application of 80-kVp and DLIR allows for significant radiation and dose reduction while further improving image quality in CCTA for overweight patients.

First Results of a New Deep Learning Reconstruction Algorithm on Image Quality and Liver Metastasis Conspicuity for Abdominal Low-Dose CT

The study's aim was to assess the impact of a deep learning image reconstruction algorithm (Precise Image; DLR) on image quality and liver metastasis conspicuity compared with an iterative reconstruction algorithm (IR). This retrospective study included all consecutive patients with at least one liver metastasis having been diagnosed between December 2021 and February 2022. Images were reconstructed using level 4 of the IR algorithm (i4) and the Standard/Smooth/Smoother levels of the DLR algorithm. Mean attenuation and standard deviation were measured by placing the ROIs in the fat, muscle, healthy liver, and liver tumor. Two radiologists assessed the image noise and image smoothing, overall image quality, and lesion conspicuity using Likert scales. The study included 30 patients (mean age 70.4 ± 9.8 years, 17 men). The mean CTDI_vol was 6.3 ± 2.1 mGy, and the mean dose-length product 314.7 ± 105.7 mGy.cm. Compared with i4, the HU values were similar in the DLR algorithm at all levels for all tissues studied. For each tissue, the image noise significantly decreased with DLR compared with i4 (p < 0.01) and significantly decreased from Standard to Smooth (-26 ± 10%; p < 0.01) and from Smooth to Smoother (-37 ± 8%; p < 0.01). The subjective image assessment confirmed that the image noise significantly decreased between i4 and DLR (p < 0.01) and from the Standard to Smoother levels (p < 0.01), but the opposite occurred for the image smoothing. The highest scores for overall image quality and conspicuity were found for the Smooth and Smoother levels.

Low dose of contrast agent and low radiation liver computed tomography with deep-learning-based contrast boosting model in participants at high-risk for hepatocellular carcinoma: prospective, randomized, double-blind study

OBJECTIVE

To investigate the image quality and lesion conspicuity of a deep-learning-based contrast-boosting (DL-CB) algorithm on double-low-dose (DLD) CT of simultaneous reduction of radiation and contrast doses in participants at high-risk for hepatocellular carcinoma (HCC).

METHODS

Participants were recruited and underwent four-phase dynamic CT (NCT04722120). They were randomly assigned to either standard-dose (SD) or DLD protocol. All CT images were initially reconstructed using iterative reconstruction, and the images of the DLD protocol were further processed using the DL-CB algorithm (DLD-DL). The primary endpoint was the contrast-to-noise ratio (CNR), the secondary endpoint was qualitative image quality (noise, hepatic lesion, and vessel conspicuity), and the tertiary endpoint was lesion detection rate. The t-test or repeated measures analysis of variance was used for analysis.

RESULTS

Sixty-eight participants with 57 focal liver lesions were enrolled (20 with HCC and 37 with benign findings). The DLD protocol had a 19.8% lower radiation dose (DLP, 855.1 ± 254.8 mGy·cm vs. 713.3 ± 94.6 mGy·cm, p = .003) and 27% lower contrast dose (106.9 ± 15.0 mL vs. 77.9 ± 9.4 mL, p < .001) than the SD protocol. The comparative analysis demonstrated that CNR (p < .001) and portal vein conspicuity (p = .002) were significantly higher in the DLD-DL than in the SD protocol. There was no significant difference in lesion detection rate for all lesions (82.7% vs. 73.3%, p = .140) and HCCs (75.7% vs. 70.4%, p = .644) between the SD protocol and DLD-DL.

CONCLUSIONS

DL-CB on double-low-dose CT provided improved CNR of the aorta and portal vein without significant impairment of the detection rate of HCC compared to the standard-dose acquisition, even in participants at high risk for HCC.

KEY POINTS

• Deep-learning-based contrast-boosting algorithm on double-low-dose CT provided an improved contrast-to-noise ratio compared to standard-dose CT. • The detection rate of focal liver lesions was not significantly differed between standard-dose CT and a deep-learning-based contrast-boosting algorithm on double-low-dose CT. • Double-low-dose CT without a deep-learning algorithm presented lower CNR and worse image quality.

Clinical impact of artificial intelligence-based solutions on imaging of the pancreas and liver

Artificial intelligence (AI) has experienced substantial progress over the last ten years in many fields of application, including healthcare. In hepatology and pancreatology, major attention to date has been paid to its application to the assisted or even automated interpretation of radiological images, where AI can generate accurate and reproducible imaging diagnosis, reducing the physicians’ workload. AI can provide automatic or semi-automatic segmentation and registration of the liver and pancreatic glands and lesions. Furthermore, using radiomics, AI can introduce new quantitative information which is not visible to the human eye to radiological reports. AI has been applied in the detection and characterization of focal lesions and diffuse diseases of the liver and pancreas, such as neoplasms, chronic hepatic disease, or acute or chronic pancreatitis, among others. These solutions have been applied to different imaging techniques commonly used to diagnose liver and pancreatic diseases, such as ultrasound, endoscopic ultrasonography, computerized tomography (CT), magnetic resonance imaging, and positron emission tomography/CT. However, AI is also applied in this context to many other relevant steps involved in a comprehensive clinical scenario to manage a gastroenterological patient. AI can also be applied to choose the most convenient test prescription, to improve image quality or accelerate its acquisition, and to predict patient prognosis and treatment response. In this review, we summarize the current evidence on the application of AI to hepatic and pancreatic radiology, not only in regard to the interpretation of images, but also to all the steps involved in the radiological workflow in a broader sense. Lastly, we discuss the challenges and future directions of the clinical application of AI methods.

[Possible Radiation Dose Reduction in Abdominal Plain CT Using Deep Learning Reconstruction]

PURPOSE

The purposes of this study were to evaluate the low-contrast detectability of CT images assuming hepatocellular carcinoma and to determine whether dose reduction in abdominal plain CT imaging is possible.

METHODS

A Catphan 600 was imaged at 350, 250, 150, and 50 mA using an Aquilion ONE PRISM Edition (Canon) and reconstructed using deep learning reconstruction (DLR) and model-based iterative reconstruction (MBIR). A low-contrast object-specific contrast-to-noise ratio (CNR_LO) was measured and compared in a 5-mm module with a CT value difference of 10 HU, assuming hepatocellular carcinoma; a visual examination was also performed. Moreover, an NPS within a uniform module was measured.

RESULTS

CNR_LO was higher for DLR at all doses (1.12 at 150 mA for DLR and 1.07 at 250 mA for MBIR). On visual evaluation, DLR could detect up to 150 mA and MBIR up to 250 mA. The NPS was lower for DLR at 0.1 cycles/mm at 150 mA.

CONCLUSION

The low-contrast detection performance was better with DLR than with MBIR, indicating the possibility of dose reduction.

Deep Learning Image Reconstruction for CT: Technical Principles and Clinical Prospects

Filtered back projection (FBP) has been the standard CT image reconstruction method for 4 decades. A simple, fast, and reliable technique, FBP has delivered high-quality images in several clinical applications. However, with faster and more advanced CT scanners, FBP has become increasingly obsolete. Higher image noise and more artifacts are especially noticeable in lower-dose CT imaging using FBP. This performance gap was partly addressed by model-based iterative reconstruction (MBIR). Yet, its "plastic" image appearance and long reconstruction times have limited widespread application. Hybrid iterative reconstruction partially addressed these limitations by blending FBP with MBIR and is currently the state-of-the-art reconstruction technique. In the past 5 years, deep learning reconstruction (DLR) techniques have become increasingly popular. DLR uses artificial intelligence to reconstruct high-quality images from lower-dose CT faster than MBIR. However, the performance of DLR algorithms relies on the quality of data used for model training. Higher-quality training data will become available with photon-counting CT scanners. At the same time, spectral data would greatly benefit from the computational abilities of DLR. This review presents an overview of the principles, technical approaches, and clinical applications of DLR, including metal artifact reduction algorithms. In addition, emerging applications and prospects are discussed.

Deep Learning Denoising of Low-Dose Computed Tomography Chest Images: A Quantitative and Qualitative Image Analysis

PURPOSE

To assess deep learning denoised (DLD) computed tomography (CT) chest images at various low doses by both quantitative and qualitative perceptual image analysis.

METHODS

Simulated noise was inserted into sinogram data from 32 chest CTs acquired at 100 mAs, generating anatomically registered images at 40, 20, 10, and 5 mAs. A DLD model was developed, with 23 scans selected for training, 5 for validation, and 4 for test.Quantitative analysis of perceptual image quality was assessed with Structural SIMilarity Index (SSIM) and Fréchet Inception Distance (FID). Four thoracic radiologists graded overall diagnostic image quality, image artifact, visibility of small structures, and lesion conspicuity. Noise-simulated and denoised image series were evaluated in comparison with one another, and in comparison with standard 100 mAs acquisition at the 4 mAs levels. Statistical tests were conducted at the 2-sided 5% significance level, with multiple comparison correction.

RESULTS

At the same mAs levels, SSIM and FID between noise-simulated and reconstructed DLD images indicated that images were closer to a perfect match with increasing mAs (closer to 1 for SSIM, and 0 for FID).In comparing noise-simulated and DLD images to standard-dose 100-mAs images, DLD improved SSIM and FID. Deep learning denoising improved SSIM of 40-, 20-, 10-, and 5-mAs simulations in comparison with standard-dose 100-mAs images, with change in SSIM from 0.91 to 0.94, 0.87 to 0.93, 0.67 to 0.87, and 0.54 to 0.84, respectively. Deep learning denoising improved FID of 40-, 20-, 10-, and 5-mAs simulations in comparison with standard-dose 100-mAs images, with change in FID from 20 to 13, 46 to 21, 104 to 41, and 148 to 69, respectively.Qualitative image analysis showed no significant difference in lesion conspicuity between DLD images at any mAs in comparison with 100-mAs images. Deep learning denoising images at 10 and 5 mAs were rated lower for overall diagnostic image quality ( P < 0.001), and at 5 mAs lower for overall image artifact and visibility of small structures ( P = 0.002), in comparison with 100 mAs.

CONCLUSIONS

Deep learning denoising resulted in quantitative improvements in image quality. Qualitative assessment demonstrated DLD images at or less than 10 mAs to be rated inferior to standard-dose images.

Abdominopelvic CT Image Quality: Evaluation of Thin (0.5-mm) Slices Using Deep Learning Reconstruction

BACKGROUND. Because thick-section images (typically 3-5 mm) have low image noise, radiologists typically use them to perform clinical interpretation, although they may additionally refer to thin-section images (typically 0.5-0.625 mm) for problem solving. Deep learning reconstruction (DLR) can yield thin-section images with low noise. OBJECTIVE. The purpose of this study is to compare abdominopelvic CT image quality between thin-section DLR images and thin- and thick-section hybrid iterative reconstruction (HIR) images. METHODS. This retrospective study included 50 patients (31 men and 19 women; median age, 64 years) who underwent abdominopelvic CT between June 15, 2020, and July 29, 2020. Images were reconstructed at 0.5-mm section using DLR and at 0.5-mm and 3.0-mm sections using HIR. Five radiologists independently performed pairwise comparisons (0.5-mm DLR and either 0.5-mm or 3.0-mm HIR) and recorded the preferred image for subjective image quality measures (scale, -2 to 2). The pooled scores of readers were compared with a score of 0 (denoting no preference). Image noise was quantified using the SD of ROIs on regions of homogeneous liver. RESULTS. For comparison of 0.5-mm DLR images and 0.5-mm HIR images, the median pooled score was 2 (indicating a definite preference for DLR) for noise and overall image quality and 1 (denoting a slight preference for DLR) for sharpness and natural appearance. For comparison of 0.5-mm DLR and 3.0-mm HIR, the median pooled score was 1 for the four previously mentioned measures. These assessments were all significantly different (p < .001) from 0. For artifacts, the median pooled score for both comparisons was 0, which was not significant for comparison with 3.0-mm HIR (p = .03) but was significant for comparison with 0.5-mm HIR (p < .001) due to imbalance in scores of 1 (n = 28) and -1 (slight preference for HIR, n = 1). Noise for 0.5-mm DLR was lower by mean differences of 12.8 HU compared with 0.5-mm HIR and 4.4 HU compared with 3.0-mm HIR (both p < .001). CONCLUSION. Thin-section DLR improves subjective image quality and reduces image noise compared with currently used thin- and thick-section HIR, without causing additional artifacts. CLINICAL IMPACT. Although further diagnostic performance studies are warranted, the findings suggest the possibility of replacing current use of both thin- and thick-section HIR with the use of thin-section DLR only during clinical interpretations.

Improving image quality with deep learning image reconstruction in double-low-dose head CT angiography compared with standard dose and adaptive statistical iterative reconstruction

OBJECTIVE

To demonstrate similar image quality with deep learning image reconstruction (DLIR) in reduced contrast medium (CM) and radiation dose (double-low-dose) head CT angiography (CTA), in comparison with standard-dose and adaptive statistical iterative reconstruction-Veo (ASIR-V).

METHODS

A prospective study was performed in 63 patients who under head CTA using 256-slice CT. Patients were randomized into either the standard-dose group (n = 38) with 40 ml of Iopromide (370 mgI ml-1 at 4.5 ml s-1); or a double-low-dose group (n = 25) with CM of 25 ml at 3.0 ml s-1. For image reconstruction, the double-low-dose group used DLIR-M and DLIR-H strength, and the standard-dose group used ASIR-V with 50% strength. The CT value and standard deviation, signal-to-noise ratio and contrast-to-noise ratio of posterior fossa, neck muscles, carotid, vertebral and middle cerebral arteries were measured. The image noise, vessel edge and structure blurring and overall image quality were assessed by using a 5-grade method. The double-low-dose group reduced CM dose by 37.5% and CT dose index by 41% compared with the standard-dose group. DLIR further reduced the standard deviation value of the middle cerebral artery and posterior fossa and provided better overall subjective image quality (p < 0.05).

CONCLUSION

DLIR significantly reduces image noise and provides higher overall image quality in the double-low-dose CTA.

ADVANCES IN KNOWLEDGE

It is feasible to reduce CM dose by 37.5% and volume CT dose index by 41% with the combination of 80 kVp and DLIR in head CTA. Compared with ASIR-V, DLIR further reduces image noise and achieves better image quality with reduced contrast and radiation dose.

Deep learning reconstruction allows for usage of contrast agent of lower concentration for coronary CTA than filtered back projection and hybrid iterative reconstruction

BACKGROUND

The demand for homogeneous and higher vascular contrast enhancement is critical to provide an appropriate interpretation of abnormal vascular findings in coronary computed tomography angiography (CTA).

PURPOSE

To evaluate the effect of various contrast media concentrations (Iohexol-370, Iohexol-300, Iohexol-240) and image reconstructions (filtered back projection [FBP], hybrid iterative reconstruction [IR], and deep learning reconstruction [DLR]) on coronary CTA.

MATERIAL AND METHODS

A total of 63 patients referred for coronary CTA between July and October 2021 were enrolled in this prospective study, and they randomly received one of three contrast media. CTA images were reconstructed with FBP, hybrid IR, and DLR. The CT attenuation, image noise, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) were calculated for all three images. The images were subjectively evaluated by two radiologists in terms of overall image quality, artifacts, image noise, and vessel wall delineation on a 5-point Likert scale.

RESULTS

The application of DLR resulted in significantly lower image noise; higher CT attenuation, SNR, and CNR; and better subjective analysis among the three different concentrations of contrast media groups (P < 0.001). There was no significant difference in the CT attenuation of the left ventricle (P = 0.089) and coronary arteries (P = 0.072) between hybrid IR at Iohexol-300 and DLR at Iohexol-240. Furthermore, application of DLR to the Iohexol-240 significantly improved SNR and CNR; it achieved higher subjective scores compared with hybrid IR at Iohexol-300 (P < 0.001).

CONCLUSION

We suggest that using DLR with Iohexol-240 contrast media is preferable to hybrid IR with Iohexol-300 contrast media in coronary CTA.

Lung-Optimized Deep-Learning-Based Reconstruction for Ultralow-Dose CT

RATIONALE AND OBJECTIVES

To evaluate the image properties of lung-specialized deep-learning-based reconstruction (DLR) and its applicability in ultralow-dose CT (ULDCT) relative to hybrid- (HIR) and model-based iterative-reconstructions (MBIR).

MATERIALS AND METHODS

An anthropomorphic chest phantom was scanned on a 320-row scanner at 50-mA (low-dose-CT 1 [LDCT-1]), 25-mA (LDCT-2), and 10-mA (ULDCT). LDCT were reconstructed with HIR; ULDCT images were reconstructed with HIR (ULDCT-HIR), MBIR (ULDCT-MBIR), and DLR (ULDCT-DLR). Image noise and contrast-to-noise ratio (CNR) were quantified. With the LDCT images as reference standards, ULDCT image qualities were subjectively scored on a 5-point scale (1 = substantially inferior to LDCT-2, 3 = comparable to LDCT-2, 5 = comparable to LDCT-1). For task-based image quality analyses, a physical evaluation phantom was scanned at seven doses to achieve the noise levels equivalent to chest phantom; noise power spectrum (NPS) and task-based transfer function (TTF) were evaluated. Clinical ULDCT (10-mA) images obtained in 14 nonobese patients were reconstructed with HIR, MBIR, and DLR; the subjective acceptability was ranked.

RESULTS

Image noise was lower and CNR was higher in ULDCT-DLR and ULDCT-MBIR than in LDCT-1, LDCT-2, and ULDCT-HIR (p < 0.01). The overall quality of ULDCT-DLR was higher than of ULDCT-HIR and ULDCT-MBIR (p < 0.01), and almost comparable with that of LDCT-2 (mean score: 3.4 ± 0.5). DLR yielded the highest NPS peak frequency and TTF_50% for high-contrast object. In clinical ULDCT images, the subjective acceptability of DLR was higher than of HIR and MBIR (p < 0.01).

CONCLUSION

DLR optimized for lung CT improves image quality and provides possible greater dose optimization opportunity than HIR and MBIR.

Deep learning image reconstruction algorithm: impact on image quality in coronary computed tomography angiography

PURPOSE

To perform a comprehensive intraindividual objective and subjective image quality evaluation of coronary CT angiography (CCTA) reconstructed with deep learning image reconstruction (DLIR) and to assess correlation with routinely applied hybrid iterative reconstruction algorithm (ASiR-V).

MATERIAL AND METHODS

Fifty-one patients (29 males) undergoing clinically indicated CCTA from April to December 2021 were prospectively enrolled. Fourteen datasets were reconstructed for each patient: three DLIR strength levels (DLIR_L, DLIR_M, and DLIR_H), ASiR-V from 10% to 100% in 10%-increment, and filtered back-projection (FBP). Signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) determined objective image quality. Subjective image quality was assessed with a 4-point Likert scale. Concordance between reconstruction algorithms was assessed by Pearson correlation coefficient.

RESULTS

DLIR algorithm did not impact vascular attenuation (P ≥ 0.374). DLIR_H showed the lowest noise, comparable with ASiR-V 100% (P = 1) and significantly lower than other reconstructions (P ≤ 0.021). DLIR_H achieved the highest objective quality, with SNR and CNR comparable to ASiR-V 100% (P = 0.139 and 0.075, respectively). DLIR_M obtained comparable objective image quality with ASiR-V 80% and 90% (P ≥ 0.281), while achieved the highest subjective image quality (4, IQR: 4-4; P ≤ 0.001). DLIR and ASiR-V datasets returned a very strong correlation in the assessment of CAD (r = 0.874, P = 0.001).

CONCLUSION

DLIR_M significantly improves CCTA image quality and has very strong correlation with routinely applied ASiR-V 50% dataset in the diagnosis of CAD.

Image quality comparison of lower extremity CTA between CT routine reconstruction algorithms and deep learning reconstruction

BACKGROUND

To evaluate the image quality of lower extremity computed tomography angiography (CTA) with deep learning-based reconstruction (DLR) compared to model-based iterative reconstruction (MBIR), hybrid-iterative reconstruction (HIR), and filtered back projection (FBP).

METHODS

Fifty patients (38 males, average age 59.8 ± 19.2 years) who underwent lower extremity CTA between January and May 2021 were included. Images were reconstructed with DLR, MBIR, HIR, and FBP. The standard deviation (SD), contrast-to-noise ratio (CNR), signal-to-noise ratio (SNR), noise power spectrum (NPS) curves, and the blur effect, were calculated. The subjective image quality was independently evaluated by two radiologists. The diagnostic accuracy of DLR, MBIR, HIR, and FBP reconstruction algorithms was calculated.

RESULTS

The CNR and SNR were significantly higher in DLR images than in the other three reconstruction algorithms, and the SD was significantly lower in DLR images of the soft tissues. The noise magnitude was the lowest with DLR. The NPS average spatial frequency (f_av) values were higher using DLR than HIR. For blur effect evaluation, DLR and FBP were similar for soft tissues and the popliteal artery, which was better than HIR and worse than MBIR. In the aorta and femoral arteries, the blur effect of DLR was worse than MBIR and FBP and better than HIR. The subjective image quality score of DLR was the highest. The sensitivity and specificity of the lower extremity CTA with DLR were the highest in the four reconstruction algorithms with 98.4% and 97.2%, respectively.

CONCLUSIONS

Compared to the other three reconstruction algorithms, DLR showed better objective and subjective image quality. The blur effect of the DLR was better than that of the HIR. The diagnostic accuracy of lower extremity CTA with DLR was the best among the four reconstruction algorithms.

Ultra-low-dose hepatic multiphase CT using deep learning-based image reconstruction algorithm focused on arterial phase in chronic liver disease: A non-inferiority study

PURPOSE

This study determined whether image quality and detectability of ultralow-dose hepatic multiphase CT (ULDCT, 33.3% dose) using a vendor-agnostic deep learning model(DLM) are noninferior to those of standard-dose CT (SDCT, 100% dose) using model-based iterative reconstruction(MBIR) in patients with chronic liver disease focusing on arterial phase.

METHODS

Sixty-seven patients underwent hepatic multiphase CT using a dual-source scanner to obtain two different radiation dose CT scans (100%, SDCT and 33.3%, ULDCT). ULDCT using DLM and SDCT using MBIR were compared. A margin of -0.5 for the difference between the two protocols was pre-defined as noninferiority of the overall image quality of the arterial phase image. Quantitative image analysis (signal to noise ratio[SNR] and contrast to noise ratio[CNR]) was also conducted. The detectability of hepatic arterial focal lesions was compared using the Jackknife free-response receiver operating characteristic analysis. Non-inferiority was satisfied if the margin of the lower limit of 95%CI of the difference in figure-of-merit was less than -0.1.

RESULTS

Mean overall arterial phase image quality scores with ULDCT using DLM and SDCT using MBIR were 4.35 ± 0.57 and 4.08 ± 0.58, showing noninferiority (difference: -0.269; 95 %CI, -0.374 to -0.164). ULDCT using DLM showed a significantly superior contrast-to-noise ratio of arterial enhancing lesion (p < 0.05). Figure-of-merit for detectability of arterial hepatic focal lesion was 0.986 for ULDCT using DLM and 0.963 for SDCT using MBIR, showing noninferiority (difference: -0.023, 95 %CI: -0.016 to 0.063).

CONCLUSION

ULDCT using DLM with 66.7% dose reduction showed non-inferior overall image quality and detectability of arterial focal hepatic lesion compared to SDCT using MBIR.

"Image quality evaluation of the Precise image CT deep learning reconstruction algorithm compared to Filtered Back-projection and iDose4: a phantom study at different dose levels"

PURPOSE

To characterize the performance of the Precise Image (PI) deep learning reconstruction (DLR) algorithm for abdominal Computed Tomography (CT) imaging.

METHODS

CT images of the Catphan-600 phantom (equipped with an external annulus) were acquired using an abdominal protocol at four dose levels and reconstructed using FBP, iDose⁴ (levels 2,5) and PI ('Soft Tissue' definition, levels 'Sharper','Sharp','Standard','Smooth','Smoother'). Image noise, image non-uniformity, noise power spectrum (NPS), target transfer function (TTF), detectability index (d'), CT numbers accuracy and image histograms were analyzed.

RESULTS

The behavior of the PI algorithm depended strongly on the selected level of reconstruction. The phantom analysis suggested that the PI image noise decreased linearly by varying the level of reconstruction from Sharper to Smoother, expressing a noise reduction up to 80% with respect to FBP. Additionally, the non-uniformity decreased, the histograms became narrower, and d' values increased as PI reconstruction levels changed from Sharper to Smoother. PI had no significant impact on the average CT number of different contrast objects. The conventional FBP NPS was deeply altered only by Smooth and Smoother levels of reconstruction. Furthermore, spatial resolution was found to be dose- and contrast-dependent, but in each analyzed condition it was greater than or comparable to FBP and iDose⁴ TTFs.

CONCLUSIONS

The PI algorithm can reduce image noise with respect to FBP and iDose⁴; spatial resolution, CT numbers and image uniformity are generally preserved by the algorithm but changes in NPS for the Smooth and Smoother levels need to be considered in protocols implementation.

Contribution of an artificial intelligence deep-learning reconstruction algorithm for dose optimization in lumbar spine CT examination: A phantom study

PURPOSE

The purpose of this study was to assess the impact of the new artificial intelligence deep-learning reconstruction (AI-DLR) algorithm on image quality and radiation dose compared with iterative reconstruction algorithm in lumbar spine computed tomography (CT) examination.

MATERIALS AND METHODS

Acquisitions on phantoms were performed using a tube current modulation system for four DoseRight Indexes (DRI) (i.e., 26/23/20/15). Raw data were reconstructed using the Level 4 of iDose⁴ (i4) and three levels of AI-DLR (Smoother/Smooth/Standard) with a bone reconstruction kernel. The Noise power spectrum (NPS), task-based transfer function (TTF) and detectability index (d') were computed (d' modeled detection of a lytic and a sclerotic bone lesions). Image quality was subjectively assessed on an anthropomorphic phantom by two radiologists.

RESULTS

The Noise magnitude was lower with AI-DLR than i4 and decreased from Standard to Smooth (-31 ± 0.1 [SD]%) and Smooth to Smoother (-48 ± 0.1 [SD]%). The average NPS spatial frequency was similar with i4 (0.43 ± 0.01 [SD] mm^-1) and Standard (0.42 ± 0.01 [SD] mm^-1) but decreased from Standard to Smoother (0.36 ± 0.01 [SD] mm^-1). TTF values at 50% decreased as the dose decreased but were similar with i4 and all AI-DLR levels. For both simulated lesions, d' values increased from Standard to Smoother levels. Higher detectabilities were found with a DRI at 15 and Smooth and Smoother levels than with a DRI at 26 and i4. The images obtained with these dose and AI-DLR levels were rated satisfactory for clinical use by the radiologists.

CONCLUSION

Using Smooth and Smoother levels with CT allows a significant dose reduction (up to 72%) with a high detectability of lytic and sclerotic bone lesions and a clinical overall image quality.

Validation of a convolutional neural network for the automated creation of curved planar reconstruction images along the main pancreatic duct

PURPOSE

To evaluate the accuracy and time-efficiency of newly developed software in automatically creating curved planar reconstruction (CPR) images along the main pancreatic duct (MPD), which was developed based on a 3-dimensional convolutional neural network, and compare them with those of conventional manually generated CPR ones.

MATERIALS AND METHODS

A total of 100 consecutive patients with MPD dilatation (≥ 3 mm) who underwent contrast-enhanced computed tomography between February 2021 and July 2021 were included in the study. Two radiologists independently performed blinded qualitative analysis of automated and manually created CPR images. They rated overall image quality based on a four-point scale and weighted κ analysis was employed to compare between manually created and automated CPR images. A quantitative analysis of the time required to create CPR images and the total length of the MPD measured from CPR images was performed.

RESULTS

The κ value was 0.796, and a good correlation was found between the manually created and automated CPR images. The average time to create automated and manually created CPR images was 61.7 s and 174.6 s, respectively (P < 0.001). The total MPD length of the automated and manually created CPR images was 110.5 and 115.6 mm, respectively (P = 0.059).

CONCLUSION

The automated CPR software significantly reduced reconstruction time without compromising image quality.

Systematic assessment of coronary calcium detectability and quantification on four generations of CT reconstruction techniques: a patient and phantom study

In computed tomography, coronary artery calcium (CAC) scores are influenced by image reconstruction. The effect of a newly introduced deep learning-based reconstruction (DLR) on CAC scoring in relation to other algorithms is unknown. The aim of this study was to evaluate the effect of four generations of image reconstruction techniques (filtered back projection (FBP), hybrid iterative reconstruction (HIR), model-based iterative reconstruction (MBIR), and DLR) on CAC detectability, quantification, and risk classification. First, CAC detectability was assessed with a dedicated static phantom containing 100 small calcifications varying in size and density. Second, CAC quantification was assessed with a dynamic coronary phantom with velocities equivalent to heart rates of 60-75 bpm. Both phantoms were scanned and reconstructed with four techniques. Last, scans of fifty patients were included and the Agatston calcium score was calculated for all four reconstruction techniques. FBP was used as a reference. In the phantom studies, all reconstruction techniques resulted in less detected small calcifications, up to 22%. No clinically relevant quantification changes occurred with different reconstruction techniques (less than 10%). In the patient study, the cardiovascular risk classification resulted, for all reconstruction techniques, in excellent agreement with the reference (κ = 0.96-0.97). However, MBIR resulted in significantly higher Agatston scores (61 (5.5-435.0) vs. 81.5 (9.25-435.0); p < 0.001) and 6% reclassification rate. In conclusion, HIR and DLR reconstructed scans resulted in similar Agatston scores with excellent agreement and low-risk reclassification rate compared with routine reconstructed scans (FBP). However, caution should be taken with low Agatston scores, as based on phantom study, detectability of small calcifications varies with the used reconstruction algorithm, especially with MBIR and DLR.

Impact of a new deep-learning-based reconstruction algorithm on image quality in ultra-high-resolution CT: clinical observational and phantom studies

OBJECTIVES

To demonstrate the effect of an improved deep learning-based reconstruction (DLR) algorithm on Ultra-High-Resolution Computed Tomography (U-HRCT) scanners.

METHODS

Clinical and phantom studies were conducted. Thirty patients who underwent contrast-enhanced CT examination during the follow-up period were enrolled. Images were reconstructed using improved DLR [termed, New DLR, i.e., Advanced Intelligent Clear-IQ Engine (AiCE) Body Sharp] and conventional DLR (Conv DLR, AiCE Body) algorithms. Two radiologists assessed the overall image quality using a 5-point scale (5 = excellent; 1 = unacceptable). The noise power spectra (NPSs) were calculated to assess the frequency characteristics of the image noise, and the square root of area under the curve (√AUC NPS) between 0.05 and 0.50 cycle/mm was calculated as an indicator of the image noise. Dunnett's test was used for statistical analysis of the visual evaluation score, with statistical significance set at p < 0.05.

RESULTS

The overall image quality of New DLR was better than that of the Conv DLR (4.2 ± 0.4 and 3.3 ± 0.4, respectively; p < 0.0001). All New DLR images had an overall image quality score above the average or excellent. The √AUCNPS value of New DLR was lower than that of Conv DLR (13.8 and 14.2, respectively). The median values of reconstruction time required with New DLR and Conv DLR were 5.0 and 7.8 min, respectively.

CONCLUSIONS

The new DLR algorithm improved the image quality within a practical reconstruction time.

ADVANCES IN KNOWLEDGE

The new DLR enables us to choose whether to improve image quality or reduce the dose.

Deep learning image reconstruction for quality assessment of iodine concentration in computed tomography: A phantom study

BACKGROUND

Recently, deep learning reconstruction (DLR) technology aiming to improve image quality with minimal radiation dose has been applied not only to pediatric scans, but also to computed tomography angiography (CTA).

OBJECTIVE

To evaluate image quality characteristics of filtered back projection (FBP), hybrid iterative reconstruction [Adaptive Iterative Dose Reduction 3D (AIDR 3D)], and DLR (AiCE) using different iodine concentrations and scan parameters.

METHODS

Phantoms with eight iodine concentrations (ranging from 1.2 to 25.9 mg/mL) located at the edge of a cylindrical water phantom with a diameter of 19 cm were scanned. Data were reconstructed with FBP, AIDR 3D, and AiCE using various scan parameters of tube current and voltage using a 320 row-detector CT scanner. Data obtained using different reconstruction techniques were quantitatively compared by analyzing Hounsfield units (HU), noise, and contrast-to-noise ratios (CNRs).

RESULTS

HU values of FBP and AIDR 3D were constant even when the iodine concentration was changed, whereas AiCE showed the highest HU value when the iodine concentration was low, but the HU value reversed when the iodine concentration exceeded a certain value. In the AIDR 3D and AiCE, the noise decreased as the tube current increased, and the change in noise when the iodine concentration was inconsistent. AIDR 3D and AiCE yielded better noise reduction rates than with FBP at a low tube current. The noise reduction rate of AIDR 3D and AiCE compared to that of FBP showed characteristics ranging from 7% to 35%, and the noise reduction rate of AiCE compared to that of AIDR 3D ranged from 2.0% to 13.3%.

CONCLUSIONS

The evaluated reconstruction techniques showed different image quality characteristics (HU value, noise, and CNR) according to dose and scan parameters, and users must consider these results and characteristics before performing patient scans.