Reproducibility of radiomic features in SENSE and compressed SENSE: impact of acceleration factors.

OBJECTIVES: To investigate the impact of acceleration factors on reproducibility of radiomic features in sensitivity encoding (SENSE) and compressed SENSE (CS), compare between SENSE and CS, and identify reproducible radiomic features. METHODS: Three-dimensional turbo spin echo T1-weighted imaging was performed in 14 healthy volunteers (mean age, 57 years; range, 33-67 years; 7 men) under SENSE and CS with accelerator factors of 5.5, 6.8, and 9.7. Eight anatomical locations (brain parenchyma, salivary glands, masseter muscle, tongue, pharyngeal mucosal space, eyeballs) were evaluated. Reproducibility of radiomic features was evaluated by calculating concordance correlation coefficient (CCC) in reference to the original image (SENSE with acceleration factor of 3.5). Reproducibility of radiomic features among acceleration factors and between SENSE and CS was compared. RESULTS: Proportion of radiomic features with CCC > 0.85 in reference to the original image was lower with higher acceleration factors in both SENSE and CS across all anatomical locations (p < .001). Proportion of radiomic features with CCC > 0.85 in reference to the original image was higher in SENSE compared with CS (SENSE, 6.7-7.3% vs CS, 4.4-5.0%; p < .001). Run percentage of gray-level run-length matrix (GLRLM) with wavelet D showed CCC > 0.85 in reference to the original image in both SENSE and CS at acceleration factor of 9.7 in the highest number of anatomical locations. CONCLUSIONS: Higher acceleration factors resulted in lower reproducibility of radiomic features in both SENSE and CS, and SENSE showed higher reproducibility of radiomic features than CS in reference to the original image. Run percentage of GLRLM with wavelet D was identified as the most reproducible feature. KEY POINTS: • Reproducibility of radiomic features in reference to the original image was lower with higher acceleration factors in both sensitivity encoding (SENSE) and compressed SENSE (CS) across all anatomical locations (p < .001). • SENSE showed higher proportions of radiomic features with CCC > 0.85 in reference to the original image (SENSE, 6.7-7.3% vs CS, 4.4-5.0%; p < .001) compared with CS. • Run percentage of gray-level run-length matrix (GLRLM) with wavelet D showed CCC > 0.85 in reference to the original image in both SENSE and CS with the highest acceleration factor.

Diffusion and perfusion MRI radiomics obtained from deep learning segmentation provides reproducible and comparable diagnostic model to human in post-treatment glioblastoma.

OBJECTIVES: Deep learning-based automatic segmentation (DLAS) helps the reproducibility of radiomics features, but its effect on radiomics modeling is unknown. We therefore evaluated whether DLAS can robustly extract anatomical and physiological MRI features, thereby assisting in the accurate assessment of treatment response in glioblastoma patients. METHODS: A DLAS model was trained on 238 glioblastomas and validated on an independent set of 98 pre- and 86 post-treatment glioblastomas from two tertiary hospitals. A total of 1618 radiomics features from contrast-enhanced T1-weighted images (CE-T1w) and histogram features from apparent diffusion coefficient (ADC) and cerebral blood volume (CBV) mapping were extracted. The diagnostic performance of radiomics features and ADC and CBV parameters for identifying treatment response was tested using area under the curve (AUC) from receiver operating characteristics analysis. Feature reproducibility was tested using a 0.80 cutoff for concordance correlation coefficients. RESULTS: Reproducibility was excellent for ADC and CBV features (ICC, 0.82-0.99) and first-order features (pre- and post-treatment, 100% and 94.1% remained), but lower for texture (79.0% and 69.1% remained) and wavelet-transformed (81.8% and 74.9% remained) features of CE-T1w. DLAS-based radiomics showed similar performance to human-performed segmentations in internal validation (AUC, 0.81 [95% CI, 0.64-0.99] vs. AUC, 0.81 [0.60-1.00], p = 0.80), but slightly lower performance in external validation (AUC, 0.78 [0.61-0.95] vs. AUC, 0.65 [0.46-0.84], p = 0.23). CONCLUSION: DLAS-based feature extraction showed high reproducibility for first-order features from anatomical and physiological MRI, and comparable diagnostic performance to human manual segmentations in the identification of pseudoprogression, supporting the utility of DLAS in quantitative MRI analysis. KEY POINTS: • Deep learning-based automatic segmentation (DLAS) enables fast and robust feature extraction from diffusion- and perfusion-weighted MRI. • DLAS showed high reproducibility in first-order feature extraction from anatomical, diffusion, and perfusion MRI across two centers. • DLAS-based radiomics features showed comparable diagnostic accuracy to manual segmentations in post-treatment glioblastoma.

Association of the Irregular 3-Dimensional Shape of White Matter Hyperintensities with Cognitive Function.

INTRODUCTION: The irregular shapes of white matter hyperintensities (WMHs) are associated with poor cognitive function, diabetes, or lacunes. However, the association between the WMH shape and dementia remains understudied. We investigated the association between the calculated shape index of WMH and the diagnosis of dementia and cognitive function. METHODS: The inverse sphericity index (ISIWMH) and volume of WMHs (VOLWMH) were compared among 82 participants with normal cognition, 82 with Alzheimer's dementia (AD), and 82 with subcortical vascular dementia (SVD). We examined the associations of ISIWMH and VOLWMH with the modified Hachinski Ischemic Score (mHIS), diagnosis of AD and SVD, and cognitive test scores, using linear, multinomial, or hierarchical linear regression models. RESULTS: The mHIS was associated with both ISIWMH (ß = 0.326, p < 0.001) and VOLWMH (ß = 0.299, p < 0.001). Both ISIWMH and VOLWMH were associated with the SVD diagnosis (odds ratio [OR] = 2.685, p = 0.002, ISIWMH; OR = 2.597, p = 0.005, VOLWMH), but not with AD. The SVD diagnosis was better explained when the multinomial regression model included both ISIWMH and VOLWMH instead of VOLWMH alone (χ2 = 20.768, df = 2, p < 0.001). The Trail Making Test-D (TMT-D) scores of the SVD patients were associated with both ISIWMH (ß = 0.308) and VOLWMH (ß = 0.293). CONCLUSION: An irregular WMH shape may be associated with the high cerebrovascular component of cognitive impairment and the diagnosis and low cognitive flexibility of SVD, which may improve the prediction of SVD diagnosis when used in combination with WMH volume.

Machine Learning Approach to Identify Stroke Within 4.5 Hours.

Background and Purpose- We aimed to investigate the ability of machine learning (ML) techniques analyzing diffusion-weighted imaging (DWI) and fluid-attenuated inversion recovery (FLAIR) magnetic resonance imaging to identify patients within the recommended time window for thrombolysis. Methods- We analyzed DWI and FLAIR images of consecutive patients with acute ischemic stroke within 24 hours of clear symptom onset by applying automatic image processing approaches. These processes included infarct segmentation, DWI, and FLAIR imaging registration and image feature extraction. A total of 89 vector features from each image sequence were captured and used in the ML. Three ML models were developed to estimate stroke onset time for binary classification (≤4.5 hours): logistic regression, support vector machine, and random forest. To evaluate the performance of ML models, the sensitivity and specificity for identifying patients within 4.5 hours were compared with the sensitivity and specificity of human readings of DWI-FLAIR mismatch. Results- Data from a total of 355 patients were analyzed. DWI-FLAIR mismatch from human readings identified patients within 4.5 hours of symptom onset with 48.5% sensitivity and 91.3% specificity. ML algorithms had significantly greater sensitivities than human readers (75.8% for logistic regression, P=0.020; 72.7% for support vector machine, P=0.033; 75.8% for random forest, P=0.013) in detecting patients within 4.5 hours, but their specificities were comparable (82.6% for logistic regression, P=0.157; 82.6% for support vector machine, P=0.157; 82.6% for random forest, P=0.157). Conclusions- ML algorithms using multiple magnetic resonance imaging features were feasible even more sensitive than human readings in identifying patients with stroke within the time window for acute thrombolysis.

Automated detection algorithm for C4d immunostaining showed comparable diagnostic performance to pathologists in renal allograft biopsy.

A deep learning-based image analysis could improve diagnostic accuracy and efficiency in pathology work. Recently, we proposed a deep learning-based detection algorithm for C4d immunostaining in renal allografts. The objective of this study is to assess the diagnostic performance of the algorithm by comparing pathologists' diagnoses and analyzing the associations of the algorithm with clinical data. C4d immunostaining slides of renal allografts were obtained from two different institutions (100 slides from the Asan Medical Center and 86 slides from the Seoul National University Hospital) and scanned using two different slide scanners. Three pathologists and the algorithm independently evaluated each slide according to the Banff 2017 criteria. Subsequently, they jointly reviewed the results for consensus scoring. The result of the algorithm was compared with that of each pathologist and the consensus diagnosis. Clinicopathological associations of the results of the algorithm with allograft survival, histologic evidence of microvascular inflammation, and serologic results for donor-specific antibodies were also analyzed. As a result, the reproducibility between the pathologists was fair to moderate (kappa 0.36-0.54), which is comparable to that between the algorithm and each pathologist (kappa 0.34-0.51). The C4d scores predicted by the algorithm achieved substantial concordance with the consensus diagnosis (kappa = 0.61), and they were significantly associated with remarkable microvascular inflammation (P = 0.001), higher detection rate of donor-specific antibody (P = 0.003), and shorter graft survival (P < 0.001). In conclusion, the deep learning-based C4d detection algorithm showed a diagnostic performance similar to that of the pathologists.

Magnetic resonance imaging texture predicts progression to dementia due to Alzheimer disease earlier than hippocampal volume

Background: Early identification of people at risk of imminent progression to dementia due to Alzheimer disease is crucial for timely intervention and treatment. We investigated whether the texture of MRI brain scans could predict the progression of mild cognitive impairment (MCI) to Alzheimer disease earlier than volume. Methods: We constructed a development data set (121 people who were cognitively normal and 145 who had mild Alzheimer disease) and a validation data set (113 patients with stable MCI who did not progress to Alzheimer disease for 3 years; 40 with early MCI who progressed to Alzheimer disease after 12­36 months; and 41 with late MCI who progressed to Alzheimer disease within 12 months) from the Alzheimer's Disease Neuroimaging Initiative. We analyzed the texture of the hippocampus, precuneus and posterior cingulate cortex using a grey-level co-occurrence matrix. We constructed texture and volume indices from the development data set using logistic regression. Using area under the curve (AUC) of receiver operator characteristics, we compared the accuracy of hippocampal volume, hippocampal texture and the composite texture of the hippocampus, precuneus and posterior cingulate cortex in predicting conversion from MCI to Alzheimer disease in the validation data set. Results: Compared with hippocampal volume, hippocampal texture (0.790 v. 0.739, p = 0.047) and composite texture (0.811 v. 0.739, p = 0.007) showed larger AUCs for conversion to Alzheimer disease from both early and late MCI. Hippocampal texture showed a marginally larger AUC than hippocampal volume in early MCI (0.795 v. 0.726, p = 0.060). Composite texture showed a larger AUC for conversion to Alzheimer disease than hippocampal volume in both early (0.817 v. 0.726, p = 0.027) and late MCI (0.805 v. 0.753, p = 0.019). Limitations: This study was limited by the absence of histological data, and the pathology reflected by the texture measures remains to be validated. Conclusion: Textures of the hippocampus, precuneus and posterior cingulate cortex predicted conversion from MCI to Alzheimer disease at an earlier time point and with higher accuracy than hippocampal volume.

Artificial Intelligence in Health Care: Current Applications and Issues.

In recent years, artificial intelligence (AI) technologies have greatly advanced and become a reality in many areas of our daily lives. In the health care field, numerous efforts are being made to implement the AI technology for practical medical treatments. With the rapid developments in machine learning algorithms and improvements in hardware performances, the AI technology is expected to play an important role in effectively analyzing and utilizing extensive amounts of health and medical data. However, the AI technology has various unique characteristics that are different from the existing health care technologies. Subsequently, there are a number of areas that need to be supplemented within the current health care system for the AI to be utilized more effectively and frequently in health care. In addition, the number of medical practitioners and public that accept AI in the health care is still low; moreover, there are various concerns regarding the safety and reliability of AI technology implementations. Therefore, this paper aims to introduce the current research and application status of AI technology in health care and discuss the issues that need to be resolved.

Fully Automated and Real-Time Volumetric Measurement of Infarct Core and Penumbra in Diffusion- and Perfusion-Weighted MRI of Patients with Hyper-Acute Stroke.

Multimodal magnetic resonance imaging (MRI) has emerged as a promising tool for diagnosing ischemic stroke and for determining treatment strategies in the acute phase. The detection and quantification of the penumbra and the infarct core regions aid the assessment of the potential risks and benefits of thrombolysis by providing information on salvageable tissue or ischemic lesion age. In this study, we proposed a fully automated and real-time algorithm to compute parameter maps of perfusion-weighted images (PWIs) and to identify an infarct core from diffusion-weighted images (DWIs). DWI and PWI were obtained using a 1.5 Tesla MRI scanner for 15 patients with acute ischemic stroke. Parameter maps of PWI were computed using restricted gamma-variate curve fitting and Fourier-based deconvolution. The ischemic penumbra was identified using time-to-maximum (Tmax) > 6 s as the mutual optimal threshold, while the infarct core was segmented using an adaptive thresholding on DWI. When the penumbra on PWI was compared with that generated using commercial software Pearson's linear correlation coefficient between penumbra volumes was 0.601 (p = 0.030), and the Dice coefficient was 0.51 ± 0.15. The infarct core on DWI was compared with the manually segmented gold standard. Dice coefficient between the manually drawn and automated segmented infarct cores was 0.62 ± 0.18. The processing times for PWI and DWI were 222.9 ± 16.4 and 53.4 ± 4.8 s, respectively. In conclusion, we demonstrate a fully automated and real-time algorithm to segment the penumbra and the infarct core regions based on PWI and DWI.

Intensity-vesselness Gaussian mixture model (IVGMM) for 2D + t segmentation of coronary arteries for X-ray angiography image sequences.

OBJECTIVE: This study aimed to propose an intensity-vesselness Gaussian mixture model (IVGMM) tracking for 2D + t segmentation of coronary arteries for X-ray angiography (XA) image sequences. METHODS: We compose a two dimensional (2D) feature vector of intensity and vesselness to characterize the Gaussian mixture models. In our IVGMM tracking, vessel segmentation is performed for each image frame based on these vessel and background IVGMMs and then the segmentation results of the current image frame is used to update these IVGMMs. The 2D + t segmentation of coronary arteries over the 2D XA image sequence is solved by means of iterating two processes, i.e., segmentation of coronary arteries and update of the IVGMMs. RESULTS: The performance of the proposed IVGMM tracking was evaluated using clinical 2D XA datasets. We evaluated the segmentation accuracy of the IVGMM tracking by comparing with two previous 2D vessel segmentation methods and seven background subtraction (BGS) methods. Of the ten segmentation methods, IVGMM tracking shows the highest similarity to the manual segmentation in terms of precision, recall, Jaccard index (JI), F1 score, and peak signal-to-noise ratio (PSNR). CONCLUSIONS: It is concluded that the IVGMM tracking could obtain reasonable segmentation accuracy outperforming conventional vessel enhancement methods and object tracking methods.

EBNA1 binding and epigenetic regulation of gastrokine tumor suppressor genes in gastric carcinoma cells.

BACKGROUND: Epstein-Barr Virus (EBV) latently infects ~10% of gastric carcinomas (GC). Epstein-Barr Nuclear Antigen 1 (EBNA1) is expressed in EBV-associated GC, and can bind host DNA, where it may impact cellular gene regulation. Here, we show that EBNA1 binds directly to DNA upstream of the divergently transcribed GC-specific tumor suppressor genes gastrokine 1 (GKN1) and gastrokine 2 (GKN2). METHODS: We use ChIP-Seq, ChIP-qPCR, and EMSA to demonstrate that EBNA1 binds directly to the GKN1 and GKN2 promoter locus. We generate AGS-EBV, and AGS-EBNA1 cell lines to study the effects of EBNA1 on GKN1 and GKN2 mRNA expression with or without 5' azacytidine treatment. RESULTS: We show that gastrokine genes are transcriptionally silenced by DNA methylation. We also show that latent EBV infection further reduces GKN1 and GKN2 expression in AGS gastric carcinoma cells, and that siRNA depletion of EBNA1 partially alleviates this repression. However, ectopic expression of EBNA1 slightly increased GKN1 and GKN2 basal mRNA levels, but reduced their responsiveness to demethylating agent. CONCLUSIONS: These findings demonstrate that EBNA1 binds to the divergent promoter of the GKN1 and GKN2 genes in GC cells, and suggest that EBNA1 contributes to the complex transcriptional and epigenetic deregulation of the GKN1 and GKN2 tumor suppressor genes in EBV positive GC.

Deep Learning-Based CT Reconstruction Kernel Conversion in the Quantification of Interstitial Lung Disease: Effect on Reproducibility.

RATIONALE AND OBJECTIVES: The effect of different computed tomography (CT) reconstruction kernels on the quantification of interstitial lung disease (ILD) has not been clearly demonstrated. The study aimed to investigate the effect of reconstruction kernels on the quantification of ILD on CT and determine whether deep learning-based kernel conversion can reduce the variability of automated quantification results between different CT kernels. MATERIALS AND METHODS: Patients with ILD or interstitial lung abnormality who underwent noncontrast high-resolution CT between June 2022 and September 2022 were retrospectively included. Images were reconstructed with three different kernels: B30f, B50f, and B60f. B60f was regarded as the reference standard for quantification, and B30f and B50f images were converted to B60f images using a deep learning-based algorithm. Each disease pattern of ILD and the fibrotic score were quantified using commercial software. The effect of kernel conversion on measurement variability was estimated using intraclass correlation coefficient (ICC) and Bland-Altman method. RESULTS: A total of 194 patients were included in the study. Application of different kernels induced differences in the quantified extent of each pattern. Reticular opacity and honeycombing were underestimated on B30f images and overestimated on B50f images. After kernel conversion, measurement variability was reduced (mean difference, from -2.0 to 3.9 to -0.3 to 0.4%, and 95% limits of agreement [LOA], from [-5.0, 12.7] to [-2.7, 2.1]). The fibrotic score for converted B60f from B50f images was almost equivalent to the original B60f (ICC, 1.000; mean difference, 0.0; and 95% LOA [-0.4, 0.4]). CONCLUSION: Quantitative CT analysis of ILD was affected by the application of different kernels, but deep learning-based kernel conversion effectively reduced measurement variability, improving the reproducibility of quantification.

Do radiomics or diffusion-tensor images provide additional information to predict brain amyloid-beta positivity?

The aim of the present study was to predict amyloid-beta positivity using a conventional T1-weighted image, radiomics, and a diffusion-tensor image obtained by magnetic resonance imaging (MRI). We included 186 patients with mild cognitive impairment (MCI) who underwent Florbetaben positron emission tomography (PET), MRI (three-dimensional T1-weighted and diffusion-tensor images), and neuropsychological tests at the Asan Medical Center. We developed a stepwise machine learning algorithm using demographics, T1 MRI features (volume, cortical thickness and radiomics), and diffusion-tensor image to distinguish amyloid-beta positivity on Florbetaben PET. We compared the performance of each algorithm based on the MRI features used. The study population included 72 patients with MCI in the amyloid-beta-negative group and 114 patients with MCI in the amyloid-beta-positive group. The machine learning algorithm using T1 volume performed better than that using only clinical information (mean area under the curve [AUC]: 0.73 vs. 0.69, p < 0.001). The machine learning algorithm using T1 volume showed better performance than that using cortical thickness (mean AUC: 0.73 vs. 0.68, p < 0.001) or texture (mean AUC: 0.73 vs. 0.71, p = 0.002). The performance of the machine learning algorithm using fractional anisotropy in addition to T1 volume was not better than that using T1 volume alone (mean AUC: 0.73 vs. 0.73, p = 0.60). Among MRI features, T1 volume was the best predictor of amyloid PET positivity. Radiomics or diffusion-tensor images did not provide additional benefits.

Diagnostic Assessment of Deep Learning Algorithms for Frozen Tissue Section Analysis in Women with Breast Cancer.

PURPOSE: Assessing the metastasis status of the sentinel lymph nodes (SLNs) for hematoxylin and eosin-stained frozen tissue sections by pathologists is an essential but tedious and time-consuming task that contributes to accurate breast cancer staging. This study aimed to review a challenge competition (HeLP 2019) for the development of automated solutions for classifying the metastasis status of breast cancer patients. Materials and Methods: A total of 524 digital slides were obtained from frozen SLN sections: 297 (56.7%) from Asan Medical Center (AMC) and 227 (43.4%) from Seoul National University Bundang Hospital (SNUBH), South Korea. The slides were divided into training, development, and validation sets, where the development set comprised slides from both institutions and training and validation set included slides from only AMC and SNUBH, respectively. The algorithms were assessed for area under the receiver operating characteristic curve (AUC) and measurement of the longest metastatic tumor diameter. The final total scores were calculated as the mean of the two metrics, and the three teams with AUC values greater than 0.500 were selected for review and analysis in this study. RESULTS: The top three teams showed AUC values of 0.891, 0.809, and 0.736 and major axis prediction scores of 0.525, 0.459, and 0.387 for the validation set. The major factor that lowered the diagnostic accuracy was micro-metastasis. CONCLUSION: In this challenge competition, accurate deep learning algorithms were developed that can be helpful for making a diagnosis on intraoperative SLN biopsy. The clinical utility of this approach was evaluated by including an external validation set from SNUBH.

Oral and gut dysbiosis leads to functional alterations in Parkinson's disease.

Although several studies have identified a distinct gut microbial composition in Parkinson's disease (PD), few studies have investigated the oral microbiome or functional alteration of the microbiome in PD. We aimed to investigate the connection between the oral and gut microbiome and the functional changes in the PD-specific gut microbiome using shotgun metagenomic sequencing. The taxonomic composition of the oral and gut microbiome was significantly different between PD patients and healthy controls (P = 0.003 and 0.001, respectively). Oral Lactobacillus was more abundant in PD patients and was associated with opportunistic pathogens in the gut (FDR-adjusted P < 0.038). Functional analysis revealed that microbial gene markers for glutamate and arginine biosynthesis were downregulated, while antimicrobial resistance gene markers were upregulated in PD patients than healthy controls (all P < 0.001). We identified a connection between the oral and gut microbiota in PD, which might lead to functional alteration of the microbiome in PD.

Does the amount of tagged stool and fluid significantly affect the radiation exposure in low-dose CT colonography performed with an automatic exposure control?

OBJECTIVE: To determine whether the amount of tagged stool and fluid significantly affects the radiation exposure in low-dose screening CT colonography performed with an automatic tube-current modulation technique. METHODS: The study included 311 patients. The tagging agent was barium (n = 271) or iodine (n = 40). Correlation was measured between mean volume CT dose index (CTDI (vol)) and the estimated x-ray attenuation of the tagged stool and fluid (ATT). Multiple linear regression analyses were performed to determine the effect of ATT on CTDI (vol ) and the effect of ATT on image noise while adjusting for other variables including abdominal circumference. RESULTS: CTDI (vol) varied from 0.88 to 2.54 mGy. There was no significant correlation between CTDI (vol) and ATT (p = 0.61). ATT did not significantly affect CTDI (vol) (p = 0.93), while abdominal circumference was the only factor significantly affecting CTDI (vol) (p < 0.001). Image noise ranged from 59.5 to 64.1 HU. The p value for the regression model explaining the noise was 0.38. CONCLUSION: The amount of stool and fluid tagging does not significantly affect radiation exposure.

Comparison of three image comparison methods for the visual assessment of the image fidelity of compressed computed tomography images.

PURPOSE: This study aimed to comparatively evaluate three different image comparison methods: alternate display without an intervening blank image (AWOB), alternate display with an intervening blank image (AWB), and side-by-side display (SSD), in terms of the perceptual sensitivity to image differences between Joint Photographic Experts Group 2000 (JPEG2000) compressed body CT images and their originals. METHODS: A total of 50 body CT images obtained with five different scan protocols (5-mm-thick abdomen, 0.67-mm-thick abdomen, 5-mm-thick lung, 0.67-mm-thick lung, and 5-mm-thick low-dose lung) were compressed to one of five compression ratios (reversible, 6:1, 8:1, 10:1, and 15:1) using JPEG2000 algorithm. The fidelity of the compressed images was visually assessed on a four-grade scale independently by five radiologists using each of the three image comparison methods of AWOB, AWB, and SSD. The fidelity grading results for the 40 irreversibly compressed images were compared between the three image comparison methods using the Friedman tests with post hoc Tukey tests. The number of image pairs with no perceptible difference was compared using the exact tests for paired proportions. The time required for the fidelity assessment for all of the 50 compressed images was also compared using the Friedman tests with post hoc Tukey tests. RESULTS: For the 40 irreversibly compressed images, the fidelity grade was significantly lower for AWOB than for AWB or SSD (p < 0.01 for all readers); however, there was no significant difference between AWB and SSD (p-value range, 0.06-0.92). The percentage of image pairs with no perceptible difference tended to be smaller for AWOB than for AWB (p < 0.01 for all readers) or SSD (p < 0.01 for readers 1-3, p = 0.04 for reader 4, and p = 0.23 for reader 5). However, there was no significant difference between AWB and SSD (p-value range, 0.12- >0.99). For all of the 50 compressed images, the fidelity grading time significantly increased in the order of AWOB, SSD, and AWB. CONCLUSIONS: In assessing the image fidelity of JPEG2000 compressed body CT images, AWOB yields lower fidelity grade and requires less fidelity grading time than AWB or SSD, indicating that AWOB is most sensitive to image differences among of them.

Introduction of heat map to fidelity assessment of compressed CT images.

PURPOSE: This study aimed to introduce heat map, a graphical data presentation method widely used in gene expression experiments, to the presentation and interpretation of image fidelity assessment data of compressed computed tomography (CT) images. METHODS: The authors used actual assessment data that consisted of five radiologists' responses to 720 computed tomography images compressed using both Joint Photographic Experts Group 2000 (JPEG2000) 2D and JPEG2000 3D compressions. They additionally created data of two artificial radiologists, which were generated by partly modifying the data from two human radiologists. RESULTS: For each compression, the entire data set, including the variations among radiologists and among images, could be compacted into a small color-coded grid matrix of the heat map. A difference heat map depicted the advantage of 3D compression over 2D compression. Dendrograms showing hierarchical agglomerative clustering results were added to the heat maps to illustrate the similarities in the data patterns among radiologists and among images. The dendrograms were used to identify two artificial radiologists as outliers, whose data were created by partly modifying the responses of two human radiologists. CONCLUSIONS: The heat map can illustrate a quick visual extract of the overall data as well as the entirety of large complex data in a compact space while visualizing the variations among observers and among images. The heat map with the dendrograms can be used to identify outliers or to classify observers and images based on the degree of similarity in the response patterns.

Predicting the fidelity of JPEG2000 compressed CT images using DICOM header information.

PURPOSE: To propose multiple logistic regression (MLR) and artificial neural network (ANN) models constructed using digital imaging and communications in medicine (DICOM) header information in predicting the fidelity of Joint Photographic Experts Group (JPEG) 2000 compressed abdomen computed tomography (CT) images. METHODS: Our institutional review board approved this study and waived informed patient consent. Using a JPEG2000 algorithm, 360 abdomen CT images were compressed reversibly (n = 48, as negative control) or irreversibly (n = 312) to one of different compression ratios (CRs) ranging from 4:1 to 10:1. Five radiologists independently determined whether the original and compressed images were distinguishable or indistinguishable. The 312 irreversibly compressed images were divided randomly into training (n = 156) and testing (n = 156) sets. The MLR and ANN models were constructed regarding the DICOM header information as independent variables and the pooled radiologists' responses as dependent variable. As independent variables, we selected the CR (DICOM tag number: 0028, 2112), effective tube current-time product (0018, 9332), section thickness (0018, 0050), and field of view (0018, 0090) among the DICOM tags. Using the training set, an optimal subset of independent variables was determined by backward stepwise selection in a four-fold cross-validation scheme. The MLR and ANN models were constructed with the determined independent variables using the training set. The models were then evaluated on the testing set by using receiver-operating-characteristic (ROC) analysis regarding the radiologists' pooled responses as the reference standard and by measuring Spearman rank correlation between the model prediction and the number of radiologists who rated the two images as distinguishable. RESULTS: The CR and section thickness were determined as the optimal independent variables. The areas under the ROC curve for the MLR and ANN predictions were 0.91 (95% CI; 0.86, 0.95) and 0.92 (0.87, 0.96), respectively. The correlation coefficients of the MLR and ANN predictions with the number of radiologists who responded as distinguishable were 0.76 (0.69, 0.82, p < 0.001) and 0.78 (0.71, 0.83, p < 0.001), respectively. CONCLUSIONS: The MLR and ANN models constructed using the DICOM header information offer promise in predicting the fidelity of JPEG2000 compressed abdomen CT images.

Stability of MRI radiomic features according to various imaging parameters in fast scanned T2-FLAIR for acute ischemic stroke patients.

From May 2015 to June 2016, data on 296 patients undergoing 1.5-Tesla MRI for symptoms of acute ischemic stroke were retrospectively collected. Conventional, echo-planar imaging (EPI) and echo train length (ETL)-T2-FLAIR were simultaneously obtained in 118 patients (first group), and conventional, ETL-, and repetition time (TR)-T2-FLAIR were simultaneously obtained in 178 patients (second group). A total of 595 radiomics features were extracted from one region-of-interest (ROI) reflecting the acute and chronic ischemic hyperintensity, and concordance correlation coefficients (CCC) of the radiomics features were calculated between the fast scanned and conventional T2-FLAIR for paired patients (1st group and 2nd group). Stabilities of the radiomics features were compared with the proportions of features with a CCC higher than 0.85, which were considered to be stable in the fast scanned T2-FLAIR. EPI-T2-FLAIR showed higher proportions of stable features than ETL-T2-FLAIR, and TR-T2-FLAIR also showed higher proportions of stable features than ETL-T2-FLAIR, both in acute and chronic ischemic hyperintensities of whole- and intersection masks (p < .002). Radiomics features in fast scanned T2-FLAIR showed variable stabilities according to the sequences compared with conventional T2-FLAIR. Therefore, radiomics features may be used cautiously in applications for feature analysis as their stability and robustness can be variable.