Multimodal modeling with low-dose CT and clinical information for diagnostic artificial intelligence on mediastinal tumors: a preliminary study.

BACKGROUND: Diagnosing mediastinal tumours, including incidental lesions, using low-dose CT (LDCT) performed for lung cancer screening, is challenging. It often requires additional invasive and costly tests for proper characterisation and surgical planning. This indicates the need for a more efficient and patient-centred approach, suggesting a gap in the existing diagnostic methods and the potential for artificial intelligence technologies to address this gap. This study aimed to create a multimodal hybrid transformer model using the Vision Transformer that leverages LDCT features and clinical data to improve surgical decision-making for patients with incidentally detected mediastinal tumours. METHODS: This retrospective study analysed patients with mediastinal tumours between 2010 and 2021. Patients eligible for surgery (n=30) were considered 'positive,' whereas those without tumour enlargement (n=32) were considered 'negative.' We developed a hybrid model combining a convolutional neural network with a transformer to integrate imaging and clinical data. The dataset was split in a 5:3:2 ratio for training, validation and testing. The model's efficacy was evaluated using a receiver operating characteristic (ROC) analysis across 25 iterations of random assignments and compared against conventional radiomics models and models excluding clinical data. RESULTS: The multimodal hybrid model demonstrated a mean area under the curve (AUC) of 0.90, significantly outperforming the non-clinical data model (AUC=0.86, p=0.04) and radiomics models (random forest AUC=0.81, p=0.008; logistic regression AUC=0.77, p=0.004). CONCLUSION: Integrating clinical and LDCT data using a hybrid transformer model can improve surgical decision-making for mediastinal tumours, showing superiority over models lacking clinical data integration.

Noise reduction by multiple path neural network using Attention mechanisms with an emphasis on robustness against Errors: A pilot study on brain Diffusion-Weighted images.

PURPOSE: In deep learning-based noise reduction, larger networks offer advanced and complex functionality by utilizing its greater degree of freedom, but come with increased unpredictability, raising the potential risk of unforeseen errors. Here, we introduce a novel denoising model for diffusion-weighted images that intentionally limits the network output freedom by incorporating multiple pathways with varying degrees of freedom, with the aim of minimizing the chance of unintended alterations to the input. The purpose of this pilot study is to assess the model's ability to perform effective denoising under the constraints. METHODS: Images from 10 healthy volunteers were used. Key innovations in our model development include: (1) neural network architecture that separated the function for calculating the specific output values from the function for adjusting the calculation for each pixel and (2) training that optimised the network based on both image and secondary obtained diffusion tensor. The generated images were compared with the original ones by measuring the deviation from ground truth images (averaged across eight acquisitions). RESULTS: The generated images demonstrated closer alignment with the ground truth images, both visually and statistically (Q < 0.05), compared to the original images. Furthermore, the advantage of the generated images over the original images was also found in the secondary obtained quantitative parameter maps with significance (Q < 0.05). CONCLUSION: The usefulness of the proposed method was suggested because it was successful in improving both the quality of the generated images and accuracy of the major diffusion parameter maps under the given restrictions.

Accuracy of skull stripping in a single-contrast convolutional neural network model using eight-contrast magnetic resonance images.

In automated analyses of brain morphometry, skull stripping or brain extraction is a critical first step because it provides accurate spatial registration and signal-intensity normalization. Therefore, it is imperative to develop an ideal skull-stripping method in the field of brain image analysis. Previous reports have shown that convolutional neural network (CNN) method is better at skull stripping than non-CNN methods. We aimed to evaluate the accuracy of skull stripping in a single-contrast CNN model using eight-contrast magnetic resonance (MR) images. A total of 12 healthy participants and 12 patients with a clinical diagnosis of unilateral Sturge-Weber syndrome were included in our study. A 3-T MR imaging system and QRAPMASTER were used for data acquisition. We obtained eight-contrast images produced by post-processing T1, T2, and proton density (PD) maps. To evaluate the accuracy of skull stripping in our CNN method, gold-standard intracranial volume (ICVG) masks were used to train the CNN model. The ICVG masks were defined by experts using manual tracing. The accuracy of the intracranial volume obtained from the single-contrast CNN model (ICVE) was evaluated using the Dice similarity coefficient [= 2(ICVE ⋂ ICVG)/(ICVE + ICVG)]. Our study showed significantly higher accuracy in the PD-weighted image (WI), phase-sensitive inversion recovery (PSIR), and PD-short tau inversion recovery (STIR) compared to the other three contrast images (T1-WI, T2-fluid-attenuated inversion recovery [FLAIR], and T1-FLAIR). In conclusion, PD-WI, PSIR, and PD-STIR should be used instead of T1-WI for skull stripping in the CNN models.

Use of a deep-learning-based lumen extraction method to detect significant stenosis on coronary computed tomography angiography in patients with severe coronary calcification.

BACKGROUND: Coronary computed tomography angiography examinations are increasingly becoming established as a minimally invasive method for diagnosing coronary diseases. However, although various imaging and processing methods have been developed, coronary artery calcification remains a major limitation in the evaluation of the vascular lumen. Subtraction coronary computed tomography angiography (Sub-CCTA) is a method known to be able to reduce the influence of coronary artery calcification and is therefore feasible for improving the diagnosis of significant stenosis in patients with severe calcification. However, Sub-CCTA still involves some problems, such as the increased radiation dose due to plain (mask) imaging, extended breath-holding time, and misregistration due to differences in the imaging phase. Therefore, we considered using artificial intelligence instead of Sub-CCTA to visualize the coronary lumen with high calcification. Given this background, the present study aimed to evaluate the diagnostic performance of a deep learning-based lumen extraction method (DL-LEM) to detect significant stenosis on CCTA in 99 consecutive patients (891 segments) with severe coronary calcification from November 2015 to March 2018. We also estimated the impact of DL-LEM on the medical economics in Japan. RESULTS: The DL-LEM slightly improved the per-segment diagnostic accuracy from 74.5 to 76.4%, and the area under the curve (AUC) slightly improved from 0.752 to 0.767 (p = 0.030). When analyzing the 228 segments that could not be evaluated because of severe calcification on the original CCTA images, the DL-LEM improved the accuracy from 35.5 to 42.5%, and the AUC improved from 0.500 to 0.587 (p = 0.00018). As a result, DL-LEM analysis could have avoided invasive coronary angiography in 4/99 cases (per patient). From the calculated results, it was estimated that the number of exams that can be avoided in Japan in one year is approximately 747 for invasive coronary angiography, 219 for fractional flow reserve, and 248 for nuclear exam. The total amount of medical fee that could be reduced was 225,629,368 JPY. CONCLUSIONS: These findings suggest that the DL-LEM may improve the diagnostic performance in detecting significant stenosis in patients with severe coronary calcification. In addition, the results suggest that not a small medical economic effect can be expected.

Development and Validation of a Modified Three-Dimensional U-Net Deep-Learning Model for Automated Detection of Lung Nodules on Chest CT Images From the Lung Image Database Consortium and Japanese Datasets.

RATIONALE AND OBJECTIVES: A more accurate lung nodule detection algorithm is needed. We developed a modified three-dimensional (3D) U-net deep-learning model for the automated detection of lung nodules on chest CT images. The purpose of this study was to evaluate the accuracy of the developed modified 3D U-net deep-learning model. MATERIALS AND METHODS: In this Health Insurance Portability and Accountability Act-compliant, Institutional Review Board-approved retrospective study, the 3D U-net based deep-learning model was trained using the Lung Image Database Consortium and Image Database Resource Initiative dataset. For internal model validation, we used 89 chest CT scans that were not used for model training. For external model validation, we used 450 chest CT scans taken at an urban university hospital in Japan. Each case included at least one nodule of >5 mm identified by an experienced radiologist. We evaluated model accuracy using the competition performance metric (CPM) (average sensitivity at 1/8, 1/4, 1/2, 1, 2, 4, and 8 false-positives per scan). The 95% confidence interval (CI) was computed by bootstrapping 1000 times. RESULTS: In the internal validation, the CPM was 94.7% (95% CI: 89.1%-98.6%). In the external validation, the CPM was 83.3% (95% CI: 79.4%-86.1%). CONCLUSION: The modified 3D U-net deep-learning model showed high performance in both internal and external validation.

Comparison of Brain Volume Measurements Made with 0.3- and 3-T MR Imaging.

The volumes of intracranial tissues of 40 healthy volunteers acquired from 0.3- and 3-T scanners were compared using intraclass correlation coefficients, correlation analyses, and Bland-Altman analyses. We found high intraclass correlation coefficients, high Pearson's correlation coefficients, and low percentage biases in all tissues and most of the brain regions, although small differences were observed in some areas. These findings may support the validity of brain volumetry with low-field magnetic resonance imaging.

Differentiation between multiple sclerosis and neuromyelitis optica spectrum disorders by multiparametric quantitative MRI using convolutional neural network.

Multiple sclerosis and neuromyelitis optica spectrum disorders are both neuroinflammatory diseases and have overlapping clinical manifestations. We developed a convolutional neural network model that differentiates between the two based on magnetic resonance imaging data. Thirty-five patients with relapsing-remitting multiple sclerosis and eighteen age-, sex-, disease duration-, and Expanded Disease Status Scale-matched patients with anti-aquaporin-4 antibody-positive neuromyelitis optica spectrum disorders were included in this study. All patients were scanned on a 3-T scanner using a multi-dynamic multi-echo sequence that simultaneously measures R1 and R2 relaxation rates and proton density. R1, R2, and proton density maps were analyzed using our convolutional neural network model. To avoid overfitting on a small dataset, we aimed to separate features of images into those specific to an image and those common to the group, based on SqueezeNet. We used only common features for classification. Leave-one-out cross validation was performed to evaluate the performance of the model. The area under the receiver operating characteristic curve of the developed convolutional neural network model for differentiating between the two disorders was 0.859. The sensitivity to multiple sclerosis and neuromyelitis optica spectrum disorders, and accuracy were 80.0%, 83.3%, and 81.1%, respectively. In conclusion, we developed a convolutional neural network model that differentiates between multiple sclerosis and neuromyelitis optica spectrum disorders, and which is designed to avoid overfitting on small training datasets. Our proposed algorithm may facilitate a differential diagnosis of these diseases in clinical practice.

A Novel Deep Learning Approach with a 3D Convolutional Ladder Network for Differential Diagnosis of Idiopathic Normal Pressure Hydrocephalus and Alzheimer's Disease.

PURPOSE: Idiopathic normal pressure hydrocephalus (iNPH) and Alzheimer's disease (AD) are geriatric diseases and common causes of dementia. Recently, many studies on the segmentation, disease detection, or classification of MRI using deep learning have been conducted. The aim of this study was to differentiate iNPH and AD using a residual extraction approach in the deep learning method. METHODS: Twenty-three patients with iNPH, 23 patients with AD and 23 healthy controls were included in this study. All patients and volunteers underwent brain MRI with a 3T unit, and we used only whole-brain three-dimensional (3D) T1-weighted images. We designed a fully automated, end-to-end 3D deep learning classifier to differentiate iNPH, AD and control. We evaluated the performance of our model using a leave-one-out cross-validation test. We also evaluated the validity of the result by visualizing important areas in the process of differentiating AD and iNPH on the original input image using the Gradient-weighted Class Activation Mapping (Grad-CAM) technique. RESULTS: Twenty-one out of 23 iNPH cases, 19 out of 23 AD cases and 22 out of 23 controls were correctly diagnosed. The accuracy was 0.90. In the Grad-CAM heat map, brain parenchyma surrounding the lateral ventricle was highlighted in about half of the iNPH cases that were successfully diagnosed. The medial temporal lobe or inferior horn of the lateral ventricle was highlighted in many successfully diagnosed cases of AD. About half of the successful cases showed nonspecific heat maps. CONCLUSION: Residual extraction approach in a deep learning method achieved a high accuracy for the differential diagnosis of iNPH, AD, and healthy controls trained with a small number of cases.

Deep Learning Approach for Generating MRA Images From 3D Quantitative Synthetic MRI Without Additional Scans.

OBJECTIVES: Quantitative synthetic magnetic resonance imaging (MRI) enables synthesis of various contrast-weighted images as well as simultaneous quantification of T1 and T2 relaxation times and proton density. However, to date, it has been challenging to generate magnetic resonance angiography (MRA) images with synthetic MRI. The purpose of this study was to develop a deep learning algorithm to generate MRA images based on 3D synthetic MRI raw data. MATERIALS AND METHODS: Eleven healthy volunteers and 4 patients with intracranial aneurysms were included in this study. All participants underwent a time-of-flight (TOF) MRA sequence and a 3D-QALAS synthetic MRI sequence. The 3D-QALAS sequence acquires 5 raw images, which were used as the input for a deep learning network. The input was converted to its corresponding MRA images by a combination of a single-convolution and a U-net model with a 5-fold cross-validation, which were then compared with a simple linear combination model. Image quality was evaluated by calculating the peak signal-to-noise ratio (PSNR), structural similarity index measurements (SSIMs), and high frequency error norm (HFEN). These calculations were performed for deep learning MRA (DL-MRA) and linear combination MRA (linear-MR), relative to TOF-MRA, and compared with each other using a nonparametric Wilcoxon signed-rank test. Overall image quality and branch visualization, each scored on a 5-point Likert scale, were blindly and independently rated by 2 board-certified radiologists. RESULTS: Deep learning MRA was successfully obtained in all subjects. The mean PSNR, SSIM, and HFEN of the DL-MRA were significantly higher, higher, and lower, respectively, than those of the linear-MRA (PSNR, 35.3 ± 0.5 vs 34.0 ± 0.5, P < 0.001; SSIM, 0.93 ± 0.02 vs 0.82 ± 0.02, P < 0.001; HFEN, 0.61 ± 0.08 vs 0.86 ± 0.05, P < 0.001). The overall image quality of the DL-MRA was comparable to that of TOF-MRA (4.2 ± 0.7 vs 4.4 ± 0.7, P = 0.99), and both types of images were superior to that of linear-MRA (1.5 ± 0.6, for both P < 0.001). No significant differences were identified between DL-MRA and TOF-MRA in the branch visibility of intracranial arteries, except for ophthalmic artery (1.2 ± 0.5 vs 2.3 ± 1.2, P < 0.001). CONCLUSIONS: Magnetic resonance angiography generated by deep learning from 3D synthetic MRI data visualized major intracranial arteries as effectively as TOF-MRA, with inherently aligned quantitative maps and multiple contrast-weighted images. Our proposed algorithm may be useful as a screening tool for intracranial aneurysms without requiring additional scanning time.

Diagnostic accuracy of 3D deep-learning-based fully automated estimation of patient-level minimum fractional flow reserve from coronary computed tomography angiography.

AIMS: Although deep-learning algorithms have been used to compute fractional flow reserve (FFR) from coronary computed tomography angiography (CCTA), no study has achieved 'fully automated' (i.e. free from human input) FFR calculation using deep-learning algorithms. The purpose of the study was to evaluate the accuracy of a fully automated 3D deep-learning model for estimating minimum FFR from CCTA data, with invasive FFR as the reference standard. METHODS AND RESULTS: This retrospective study of 1052 patients included 131 patients whose CCTA studies showed 30-90% stenosis and underwent invasive FFR (abnormal FFR observed in 72/131, 55%), and 921 patients who underwent clinically indicated CCTA without invasive FFR. We designed a fully automated 3D deep-learning model that inputs CCTA data and outputs minimum FFR without requiring human input. The model comprised a series of deep-learning algorithms: a conditional generative adversarial network, a 3D convolutional ladder network, and two independent neural networks with integrated virtual adversarial training. We used Monte Carlo cross-validation to evaluate the accuracy of the model for estimating FFR, with invasive FFR as the reference standard. The deep-learning FFR achieved area under the receiver-operating characteristic curve of 0.78 for detection of abnormal FFR; and was significantly higher than for visually determined CCTA >50% stenosis (area under the curve = 0.56). The deep-learning FFR model achieved 76% accuracy for detecting abnormal FFR, with sensitivity of 85% (79-89%) and specificity of 63% (54-70%). CONCLUSION: The 3D deep-learning model, which performs fully automatic estimation of minimum FFR from cardiac CT data, achieved 76% accuracy in detecting abnormal FFR.

Convolutional neural network-based segmentation can help in assessing the substantia nigra in neuromelanin MRI.

PURPOSE: This study aimed to evaluate the accuracy and diagnostic test performance of the U-net-based segmentation method in neuromelanin magnetic resonance imaging (NM-MRI) compared to the established manual segmentation method for Parkinson's disease (PD) diagnosis. METHODS: NM-MRI datasets from two different 3T-scanners were used: a "principal dataset" with 122 participants and an "external validation dataset" with 24 participants, including 62 and 12 PD patients, respectively. Two radiologists performed SNpc manual segmentation. Inter-reader precision was determined using Dice coefficients. The U-net was trained with manual segmentation as ground truth and Dice coefficients used to measure accuracy. Training and validation steps were performed on the principal dataset using a 4-fold cross-validation method. We tested the U-net on the external validation dataset. SNpc hyperintense areas were estimated from U-net and manual segmentation masks, replicating a previously validated thresholding method, and their diagnostic test performances for PD determined. RESULTS: For SNpc segmentation, U-net accuracy was comparable to inter-reader precision in the principal dataset (Dice coefficient: U-net, 0.83 ± 0.04; inter-reader, 0.83 ± 0.04), but lower in external validation dataset (Dice coefficient: U-net, 079 ± 0.04; inter-reader, 0.85 ± 0.03). Diagnostic test performances for PD were comparable between U-net and manual segmentation methods in both principal (area under the receiver operating characteristic curve: U-net, 0.950; manual, 0.948) and external (U-net, 0.944; manual, 0.931) datasets. CONCLUSION: U-net segmentation provided relatively high accuracy in the evaluation of the SNpc in NM-MRI and yielded diagnostic performance comparable to that of the established manual method.

Analysis of antihypertensive treatment using real-world Japanese data-the retrospective study of antihypertensives for lowering blood pressure (REAL) study.

Hypertension requires strict treatment because it causes diseases that can lead to death. Although various classes of antihypertensive drugs are available, the actual status of antihypertensive drug selection and the transition in prescription patterns over time have not been fully examined. Therefore, we conducted a claims-based study using two claims databases (2008-16) to determine this status in Japan. We examined the prescription rate for each class of antihypertensive drugs in hypertensive patients and compared the patients' ages and the sizes of the medical institutions treating these patients. Among the 1 560 865 and 302 433 hypertensive patients in each database, calcium channel blockers (CCBs) (>60%) and angiotensin II receptor blockers (ARBs) (>55%) were the most frequently prescribed classes. The prescription rate of CCBs increased and ARBs decreased with the patients' ages. Although the Japanese guidelines for management of hypertension in 2014 changed the recommendation and indicated that ß-blockers should not be used as first-line drugs, their prescription status did not change during this study period up to 2016. Use of CCBs and ARBs as first-line drugs differed by the types of patient comorbidities. Although ARBs or angiotensin-converting enzyme inhibitors were recommended for patients with some comorbidities, CCBs were used relatively frequently. In conclusion, the patients' ages and comorbidities and the sizes of the medical institutions affect the selection of antihypertensive drugs. Selection and use of drugs may not always follow the guidelines.

Current treatment status and medical cost for multiple sclerosis based on analysis of a Japanese claims database.

OBJECTIVE: To assist policymakers as they reflect on treatment protocols and approaches for the efficient delivery of medical care for multiple sclerosis (MS) patients in Japan. METHODS: We analyzed data from a large Japanese health insurance claims database. Using an algorithm based on diagnosis codes, all patients with a diagnosis of MS were identified; patients having a non-MS demyelinating disease were excluded from the population. MS patient data were used for cross-sectional analysis carried out on the data collected at a certain period. We identified a total of 1808 MS patients, and we analyzed data for 1133 patients with an observation period of ≥6 months from October 2013 to September 2014. Newly diagnosed MS patients were identified within the MS patients, and their data were used for longitudinal analysis, tracking each patient over a period of time. RESULTS: The total per patient per month cost for MS was ¥93 542 (US$781, €695 as of October 2015). Disease-modifying therapy drugs costs constituted half of the overall medical costs. For newly diagnosed MS patients, hospitalization costs were the largest component in the initial month, while drug costs were the largest component more than several months after the initial visit. There was a positive correlation between relapse frequency and medical cost. CONCLUSIONS: These results provide up-to-date information on the demographics, medical treatment and cost status of MS in almost real-time by using a claims database. They suggest that claims data analysis can effectively support medical policymaking.

Analysis of second- and third-line antihypertensive treatments after initial therapy with an angiotensin II receptor blocker using real-world Japanese data.

Combination therapy using two or three classes of drugs is often required to treat hypertension to prevent cardiovascular disease. In this study, we examined combination therapies administered following initial therapy with an angiotensin II receptor blocker (ARB) in hypertensive Japanese patients. To determine which classes of antihypertensives are being prescribed as second- or third-line treatments for patients who were initially treated with a single ARB, we analyzed prescription claims data from two Japanese health-care databases for 2008 to 2015. Among the 26 998 patients who were initially treated with a single ARB (from one database), calcium channel blockers (CCBs) were the most frequently prescribed second-line antihypertensive, as these medicines were added for >20% of patients within 1 year of ARB prescription initiation. The addition rates of CCBs as a second-line therapy differed depending on the initial ARB type. In contrast, <10% of patients received a diuretic as a second-line antihypertensive. Among the 48 813 patients who were prescribed an ARB in combination with a CCB (as shown in the other database), diuretics were prescribed as third-line antihypertensives more frequently than increased doses of CCBs or ARBs. Diuretics were added for 8% of patients within 2 years of CCB addition, and the addition rates differed based on the CCB dose used for combination therapy. We also found that the addition rates of diuretics differed depending on patient clinical histories among ARB and CCB recipients.