Journal Of Digital Imaging[JOURNAL]
Sun
200 papers
Workload Balancing in Emergency Night Shifts for a Multicenter Diagnostic Imaging Department: a RIS-Integrated Solution.
Nitrosi A, Pattacini P, Bertolini M … +9 more , Trojani V, Sghedoni R, Spaggiari L, Zanichelli M, Besutti G, Notari P, Canovi L, Colli M, Iori M
J Digit Imaging · 2023 Oct · PMID 37349619 · Full text
The purpose of this short report is to illustrate the implementation of a RIS function for balancing radiological activities and workloads between two different teams of radiologists from the same Diagnostic Department d...
The purpose of this short report is to illustrate the implementation of a RIS function for balancing radiological activities and workloads between two different teams of radiologists from the same Diagnostic Department during emergency nights and holiday shifts. One group is from the main hospital, Arcispedale S.Maria Nuova di Reggio Emilia, and the other group belongs to the five minor hospitals in the district of Reggio Emilia.The implementation of a dedicated balancing function in the RIS system successfully allows the balancing of the radiological activity between two or more teams of different radiologists, while preserving the care continuity of care and the involved workers' experience and confidence in reporting.
Brain Tumor Segmentation for Multi-Modal MRI with Missing Information.
Feng X, Ghimire K, Kim DD … +13 more , Chandra RS, Zhang H, Peng J, Han B, Huang G, Chen Q, Patel S, Bettagowda C, Sair HI, Jones C, Jiao Z, Yang L, Bai H
J Digit Imaging · 2023 Oct · PMID 37340197 · Full text
Deep convolutional neural networks (DCNNs) have shown promise in brain tumor segmentation from multi-modal MRI sequences, accommodating heterogeneity in tumor shape and appearance. The fusion of multiple MRI sequences al...
Deep convolutional neural networks (DCNNs) have shown promise in brain tumor segmentation from multi-modal MRI sequences, accommodating heterogeneity in tumor shape and appearance. The fusion of multiple MRI sequences allows networks to explore complementary tumor information for segmentation. However, developing a network that maintains clinical relevance in situations where certain MRI sequence(s) might be unavailable or unusual poses a significant challenge. While one solution is to train multiple models with different MRI sequence combinations, it is impractical to train every model from all possible sequence combinations. In this paper, we propose a DCNN-based brain tumor segmentation framework incorporating a novel sequence dropout technique in which networks are trained to be robust to missing MRI sequences while employing all other available sequences. Experiments were performed on the RSNA-ASNR-MICCAI BraTS 2021 Challenge dataset. When all MRI sequences were available, there were no significant differences in performance of the model with and without dropout for enhanced tumor (ET), tumor (TC), and whole tumor (WT) (p-values 1.000, 1.000, 0.799, respectively), demonstrating that the addition of dropout improves robustness without hindering overall performance. When key sequences were unavailable, the network with sequence dropout performed significantly better. For example, when tested on only T1, T2, and FLAIR sequences together, DSC for ET, TC, and WT increased from 0.143 to 0.486, 0.431 to 0.680, and 0.854 to 0.901, respectively. Sequence dropout represents a relatively simple yet effective approach for brain tumor segmentation with missing MRI sequences.
Quantitative Validation of the Correlation Between Optimized Pyramidal Tract Delineation After Brain Shift Compensation and Direct Electrical Subcortical Stimulation During Brain Tumor Surgery.
Li Y, Hou Y, Li X … +3 more , Li Q, Lu J, Tang J
J Digit Imaging · 2023 Oct · PMID 37340196 · Full text
It remains unclear whether tractography of pyramidal tracts is correlated with the intraoperative direct electrical subcortical stimulation (DESS), and brain shift further complicates the issue. The objective of this res...
It remains unclear whether tractography of pyramidal tracts is correlated with the intraoperative direct electrical subcortical stimulation (DESS), and brain shift further complicates the issue. The objective of this research is to quantitatively verify the correlation between optimized tractography (OT) of pyramidal tracts after brain shift compensation and DESS during brain tumor surgery. OT was performed for 20 patients with lesions in proximity to the pyramidal tracts based on preoperative diffusion-weighted magnetic resonance imaging. During surgery, tumor resection was guided by DESS. A total of 168 positive stimulation points and their corresponding stimulation intensity thresholds were recorded. Using the brain shift compensation algorithm based on hierarchical B-spline grids combined with a Gaussian resolution pyramid, we warped the preoperative pyramidal tract models and used receiver operating characteristic (ROC) curves to investigate the reliability of our brain shift compensation method based on anatomic landmarks. Additionally, the minimum distance between the DESS points and warped OT (wOT) model was measured and correlated with DESS intensity threshold. Brain shift compensation was achieved in all cases, and the area under the ROC curve was 0.96 in the registration accuracy analysis. The minimum distance between the DESS points and the wOT model was found to have a significantly high correlation with the DESS stimulation intensity threshold (r = 0.87, P < 0.001), with a linear regression coefficient of 0.96. Our OT method can provide comprehensive and accurate visualization of the pyramidal tracts for neurosurgical navigation and was quantitatively verified by intraoperative DESS after brain shift compensation.
What Matters in Radiological Image Segmentation? Effect of Segmentation Errors on the Diagnostic Related Features.
Chen Z, Chen J, Zhao J … +7 more , Liu B, Jiang S, Si D, Ding H, Nian Y, Yang X, Xiao J
J Digit Imaging · 2023 Oct · PMID 37340195 · Full text
Segmentation is a crucial step in extracting the medical image features for clinical diagnosis. Though multiple metrics have been proposed to evaluate the segmentation performance, there is no clear study on how or to wh...
Segmentation is a crucial step in extracting the medical image features for clinical diagnosis. Though multiple metrics have been proposed to evaluate the segmentation performance, there is no clear study on how or to what extent the segmentation errors will affect the diagnostic related features used in clinical practice. Therefore, we proposed a segmentation robustness plot (SRP) to build the link between segmentation errors and clinical acceptance, where relative area under the curve (R-AUC) was designed to help clinicians to identify the robust diagnostic related image features. In experiments, we first selected representative radiological series from time series (cardiac first-pass perfusion) and spatial series (T2 weighted images on brain tumors) of magnetic resonance images, respectively. Then, dice similarity coefficient (DSC) and Hausdorff distance (HD), as the widely used evaluation metrics, were used to systematically control the degree of the segmentation errors. Finally, the differences between diagnostic related image features extracted from the ground truth and the derived segmentation were analyzed, using the statistical method large sample size T-test to calculate the corresponding p values. The results are denoted in the SRP, where the x-axis indicates the segmentation performance using the aforementioned evaluation metric, and the y-axis shows the severity of the corresponding feature changes, which are expressed in either the p values for a single case or the proportion of patients without significant change. The experimental results in SRP show that when DSC is above 0.95 and HD is below 3 mm, the segmentation errors will not change the features significantly in most cases. However, when segmentation gets worse, additional metrics are required for further analysis. In this way, the proposed SRP indicates the impact of the segmentation errors on the severity of the corresponding feature changes. By using SRP, one could easily define the acceptable segmentation errors in a challenge. Additionally, the R-AUC calculated from SRP provides an objective reference to help the selection of reliable features in image analysis.
Consistency of Artificial Intelligence (AI)-based Diagnostic Support Software in Short-term Digital Mammography Reimaging After Core Needle Biopsy.
Youk JH, Han K, Lee SE … +1 more , Kim EK
J Digit Imaging · 2023 Oct · PMID 37326891 · Full text
To evaluate the consistency in the performance of Artificial Intelligence (AI)-based diagnostic support software in short-term digital mammography reimaging after core needle biopsy. Of 276 women who underwent short-term...
To evaluate the consistency in the performance of Artificial Intelligence (AI)-based diagnostic support software in short-term digital mammography reimaging after core needle biopsy. Of 276 women who underwent short-term (<3 mo) serial digital mammograms followed by breast cancer surgery from Jan. to Dec. 2017, 550 breasts were included. All core needle biopsies for breast lesions were performed between serial exams. All mammography images were analyzed using a commercially available AI-based software providing an abnormality score (0-100). Demographic data for age, interval between serial exams, biopsy, and final diagnosis were compiled. Mammograms were reviewed for mammographic density and finding. Statistical analysis was performed to evaluate the distribution of variables according to biopsy and to test the interaction effects of variables with the difference in AI-based score according to biopsy. AI-based score of 550 exams (benign or normal in 263 and malignant in 287) showed significant difference between malignant and benign/normal exams (0.48 vs. 91.97 in first exam and 0.62 vs. 87.13 in second exam, P<0.0001). In comparison of serial exams, no significant difference was found in AI-based score. AI-based score difference between serial exams was significantly different according to biopsy performed or not (-0.25 vs. 0.07, P = 0.035). In linear regression analysis, there was no significant interaction effect of all clinical and mammographic characteristics with mammographic examinations performed after biopsy or not. The results from AI-based diagnostic support software for digital mammography was relatively consistent in short-term reimaging even after core needle biopsy.
Lessons Learned in Change Management in Deploying Novel Informatics Solutions: Experience Implementing a Point-of-Care Patient Photography System with Radiography.
Wick CA, Tridandapani S, Heilbrun ME … +3 more , Hanna T, Safdar N, Bhatti P
J Digit Imaging · 2023 Oct · PMID 37322308 · Full text
We describe implementation of a point-of-care system for simultaneous acquisition of patient photographs along with portable radiographs at a large academic hospital. During the implementation process, we observed severa...
We describe implementation of a point-of-care system for simultaneous acquisition of patient photographs along with portable radiographs at a large academic hospital. During the implementation process, we observed several technical challenges in the areas of (1) hardware-automatic triggering for photograph acquisition, camera hardware enclosure, networking, and system server hardware and (2) software-post-processing of photographs. Additionally, we also faced cultural challenges involving workflow issues, communication with technologists and users, and system maintenance. We describe our solutions to address these challenges. We anticipate that these experiences will provide useful insights into deploying and iterating new technologies in imaging informatics.
Effect of Gaussian Smoothing Filter Size for CT-Based Attenuation Correction on Quantitative Assessment of Bone SPECT/CT: A Phantom Study.
Tomita Y, Ichikawa Y, Hashizume K … +1 more , Sakuma H
J Digit Imaging · 2023 Oct · PMID 37322307 · Full text
This study aims to determine the effect of Gaussian filter size for CT-based attenuation correction (CTAC) on the quantitative assessment of bone SPECT. An experiment was performed using a cylindrical phantom containing...
This study aims to determine the effect of Gaussian filter size for CT-based attenuation correction (CTAC) on the quantitative assessment of bone SPECT. An experiment was performed using a cylindrical phantom containing six rods, of which one was filled with water and five were filled with various concentrations of KHPO solution (120-960 mg/cm) to simulate different bone densities. Tc-solution of 207 kBq/ml was also included within the rods. SPECT data were acquired at 120 views for 30 s/view. CT for attenuation correction were obtained at 120 kVp and 100 mA. Sixteen different CTAC maps processed with different Gaussian filter sizes (ranging from 0 to 30 mm in 2 mm increments) were generated. SPECT images were reconstructed for each of the 16 CTAC maps. Attenuation coefficients and radioactivity concentrations in the rods were compared with those in the water-filled rod without KHPO solution as a reference. Gaussian filter sizes below 14-16 mm resulted in an overestimation of radioactivity concentrations for rods with high concentrations of KHPO (≥ 666 mg/cm). The overestimation of radioactivity concentration measurement was 3.8% and 5.5% for 666 mg/cm and 960 mg/cm KHPO solutions, respectively. The difference in radioactivity concentration between the water rod and the KHPO rods was minimal at 18-22 mm. The use of Gaussian filter sizes smaller than 14-16 mm caused an overestimation of radioactivity concentration in regions of high CT values. Setting the Gaussian filter size to 18-22 mm enables radioactivity concentration to be measured with the least influence on bone density.
A Multi-Stage Faster RCNN-Based iSPLInception for Skin Disease Classification Using Novel Optimization.
Josphineleela R, Raja Rao PBV, Shaikh A … +1 more , Sudhakar K
J Digit Imaging · 2023 Oct · PMID 37322306 · Full text
Nowadays, skin cancer is considered a serious disorder in which early identification and treatment of the disease are essential to ensure the stability of the patients. Several existing skin cancer detection methods are...
Nowadays, skin cancer is considered a serious disorder in which early identification and treatment of the disease are essential to ensure the stability of the patients. Several existing skin cancer detection methods are introduced by employing deep learning (DL) to perform skin disease classification. Convolutional neural networks (CNNs) can classify melanoma skin cancer images. But, it suffers from an overfitting problem. Therefore, to overcome this problem and to classify both benign and malignant tumors efficiently, the multi-stage faster RCNN-based iSPLInception (MFRCNN-iSPLI) method is proposed. Then, the test dataset is used for evaluating the proposed model performance. The faster RCNN is employed directly to perform image classification. This may heavily raise computation time and network complications. So, the iSPLInception model is applied in the multi-stage classification. In this, the iSPLInception model is formulated using the Inception-ResNet design. For candidate box deletion, the prairie dog optimization algorithm is utilized. We have utilized two skin disease datasets, namely, ISIC 2019 Skin lesion image classification and the HAM10000 dataset for conducting experimental results. The methods' accuracy, precision, recall, and F1 score values are calculated, and the results are compared with the existing methods such as CNN, hybrid DL, Inception v3, and VGG19. With 95.82% accuracy, 96.85% precision, 96.52% recall, and 0.95% F1 score values, the output analysis of each measure verified the prediction and classification effectiveness of the method.
Patient Identification Based on Deep Metric Learning for Preventing Human Errors in Follow-up X-Ray Examinations.
Ueda Y, Morishita J
J Digit Imaging · 2023 Oct · PMID 37308675 · Full text
Biological fingerprints extracted from clinical images can be used for patient identity verification to determine misfiled clinical images in picture archiving and communication systems. However, such methods have not be...
Biological fingerprints extracted from clinical images can be used for patient identity verification to determine misfiled clinical images in picture archiving and communication systems. However, such methods have not been incorporated into clinical use, and their performance can degrade with variability in the clinical images. Deep learning can be used to improve the performance of these methods. A novel method is proposed to automatically identify individuals among examined patients using posteroanterior (PA) and anteroposterior (AP) chest X-ray images. The proposed method uses deep metric learning based on a deep convolutional neural network (DCNN) to overcome the extreme classification requirements for patient validation and identification. It was trained on the NIH chest X-ray dataset (ChestX-ray8) in three steps: preprocessing, DCNN feature extraction with an EfficientNetV2-S backbone, and classification with deep metric learning. The proposed method was evaluated using two public datasets and two clinical chest X-ray image datasets containing data from patients undergoing screening and hospital care. A 1280-dimensional feature extractor pretrained for 300 epochs performed the best with an area under the receiver operating characteristic curve of 0.9894, an equal error rate of 0.0269, and a top-1 accuracy of 0.839 on the PadChest dataset containing both PA and AP view positions. The findings of this study provide considerable insights into the development of automated patient identification to reduce the possibility of medical malpractice due to human errors.
Histopathological Image Deep Feature Representation for CBIR in Smart PACS.
Tommasino C, Merolla F, Russo C … +2 more , Staibano S, Rinaldi AM
J Digit Imaging · 2023 Oct · PMID 37296349 · Full text
Pathological Anatomy is moving toward computerizing processes mainly due to the extensive digitization of histology slides that resulted in the availability of many Whole Slide Images (WSIs). Their use is essential, espe...
Pathological Anatomy is moving toward computerizing processes mainly due to the extensive digitization of histology slides that resulted in the availability of many Whole Slide Images (WSIs). Their use is essential, especially in cancer diagnosis and research, and raises the pressing need for increasingly influential information archiving and retrieval systems. Picture Archiving and Communication Systems (PACSs) represent an actual possibility to archive and organize this growing amount of data. The design and implementation of a robust and accurate methodology for querying them in the pathology domain using a novel approach are mandatory. In particular, the Content-Based Image Retrieval (CBIR) methodology can be involved in the PACSs using a query-by-example task. In this context, one of many crucial points of CBIR concerns the representation of images as feature vectors, and the accuracy of retrieval mainly depends on feature extraction. Thus, our study explored different representations of WSI patches by features extracted from pre-trained Convolution Neural Networks (CNNs). In order to perform a helpful comparison, we evaluated features extracted from different layers of state-of-the-art CNNs using different dimensionality reduction techniques. Furthermore, we provided a qualitative analysis of obtained results. The evaluation showed encouraging results for our proposed framework.
Deep Ensembles Are Robust to Occasional Catastrophic Failures of Individual DNNs for Organs Segmentations in CT Images.
Petrov Y, Malik B, Fredrickson J … +2 more , Jemaa S, Carano RAD
J Digit Imaging · 2023 Oct · PMID 37291384 · Full text
Deep neural networks (DNNs) have recently showed remarkable performance in various computer vision tasks, including classification and segmentation of medical images. Deep ensembles (an aggregated prediction of multiple...
Deep neural networks (DNNs) have recently showed remarkable performance in various computer vision tasks, including classification and segmentation of medical images. Deep ensembles (an aggregated prediction of multiple DNNs) were shown to improve a DNN's performance in various classification tasks. Here we explore how deep ensembles perform in the image segmentation task, in particular, organ segmentations in CT (Computed Tomography) images. Ensembles of V-Nets were trained to segment multiple organs using several in-house and publicly available clinical studies. The ensembles segmentations were tested on images from a different set of studies, and the effects of ensemble size as well as other ensemble parameters were explored for various organs. Compared to single models, Deep Ensembles significantly improved the average segmentation accuracy, especially for those organs where the accuracy was lower. More importantly, Deep Ensembles strongly reduced occasional "catastrophic" segmentation failures characteristic of single models and variability of the segmentation accuracy from image to image. To quantify this we defined the "high risk images": images for which at least one model produced an outlier metric (performed in the lower 5% percentile). These images comprised about 12% of the test images across all organs. Ensembles performed without outliers for 68%-100% of the "high risk images" depending on the performance metric used.
Utility of Artificial Intelligence for Real-Time Anatomical Landmark Identification in Ultrasound-Guided Thoracic Paravertebral Block.
Zhao Y, Zheng S, Cai N … +5 more , Zhang Q, Zhong H, Zhou Y, Zhang B, Wang G
J Digit Imaging · 2023 Oct · PMID 37291383 · Full text
Thoracic paravertebral block (TPVB) is a common method of inducing perioperative analgesia in thoracic and abdominal surgery. Identifying anatomical structures in ultrasound images is very important especially for inexpe...
Thoracic paravertebral block (TPVB) is a common method of inducing perioperative analgesia in thoracic and abdominal surgery. Identifying anatomical structures in ultrasound images is very important especially for inexperienced anesthesiologists who are unfamiliar with the anatomy. Therefore, our aim was to develop an artificial neural network (ANN) to automatically identify (in real-time) anatomical structures in ultrasound images of TPVB. This study is a retrospective study using ultrasound scans (both video and standard still images) that we acquired. We marked the contours of the paravertebral space (PVS), lung, and bone in the TPVB ultrasound image. Based on the labeled ultrasound images, we used the U-net framework to train and create an ANN that enabled real-time identification of important anatomical structures in ultrasound images. A total of 742 ultrasound images were acquired and labeled in this study. In this ANN, the Intersection over Union (IoU) and Dice similarity coefficient (DSC or Dice coefficient) of the paravertebral space (PVS) were 0.75 and 0.86, respectively, the IoU and DSC of the lung were 0.85 and 0.92, respectively, and the IoU and DSC of the bone were 0.69 and 0.83, respectively. The accuracies of the PVS, lung, and bone were 91.7%, 95.4%, and 74.3%, respectively. For tenfold cross validation, the median interquartile range for PVS IoU and DSC was 0.773 and 0.87, respectively. There was no significant difference in the scores for the PVS, lung, and bone between the two anesthesiologists. We developed an ANN for the real-time automatic identification of thoracic paravertebral anatomy. The performance of the ANN was highly satisfactory. We conclude that AI has good prospects for use in TPVB. Clinical registration number: ChiCTR2200058470 (URL: http://www.chictr.org.cn/showproj.aspx?proj=152839 ; registration date: 2022-04-09).
Using Deep Learning to Detect the Presence and Location of Hemoperitoneum on the Focused Assessment with Sonography in Trauma (FAST) Examination in Adults.
Leo MM, Potter IY, Zahiri M … +3 more , Vaziri A, Jung CF, Feldman JA
J Digit Imaging · 2023 Oct · PMID 37286904 · Full text
Abdominal ultrasonography has become an integral component of the evaluation of trauma patients. Internal hemorrhage can be rapidly diagnosed by finding free fluid with point-of-care ultrasound (POCUS) and expedite decis...
Abdominal ultrasonography has become an integral component of the evaluation of trauma patients. Internal hemorrhage can be rapidly diagnosed by finding free fluid with point-of-care ultrasound (POCUS) and expedite decisions to perform lifesaving interventions. However, the widespread clinical application of ultrasound is limited by the expertise required for image interpretation. This study aimed to develop a deep learning algorithm to identify the presence and location of hemoperitoneum on POCUS to assist novice clinicians in accurate interpretation of the Focused Assessment with Sonography in Trauma (FAST) exam. We analyzed right upper quadrant (RUQ) FAST exams obtained from 94 adult patients (44 confirmed hemoperitoneum) using the YoloV3 object detection algorithm. Exams were partitioned via fivefold stratified sampling for training, validation, and hold-out testing. We assessed each exam image-by-image using YoloV3 and determined hemoperitoneum presence for the exam using the detection with highest confidence score. We determined the detection threshold as the score that maximizes the geometric mean of sensitivity and specificity over the validation set. The algorithm had 95% sensitivity, 94% specificity, 95% accuracy, and 97% AUC over the test set, significantly outperforming three recent methods. The algorithm also exhibited strength in localization, while the detected box sizes varied with a 56% IOU averaged over positive cases. Image processing demonstrated only 57-ms latency, which is adequate for real-time use at the bedside. These results suggest that a deep learning algorithm can rapidly and accurately identify the presence and location of free fluid in the RUQ of the FAST exam in adult patients with hemoperitoneum.
Visual Image Annotation for Bowel Obstruction: Repeatability and Agreement with Manual Annotation and Neural Networks.
Murphy PM
J Digit Imaging · 2023 Oct · PMID 37278918 · Full text
Bowel obstruction is a common cause of acute abdominal pain. The development of algorithms for automated detection and characterization of bowel obstruction on CT has been limited by the effort required for manual annota...
Bowel obstruction is a common cause of acute abdominal pain. The development of algorithms for automated detection and characterization of bowel obstruction on CT has been limited by the effort required for manual annotation. Visual image annotation with an eye tracking device may mitigate that limitation. The purpose of this study is to assess the agreement between visual and manual annotations for bowel segmentation and diameter measurement, and to assess agreement with convolutional neural networks (CNNs) trained using that data. Sixty CT scans of 50 patients with bowel obstruction from March to June 2022 were retrospectively included and partitioned into training and test data sets. An eye tracking device was used to record 3-dimensional coordinates within the scans, while a radiologist cast their gaze at the centerline of the bowel, and adjusted the size of a superimposed ROI to approximate the diameter of the bowel. For each scan, 59.4 ± 15.1 segments, 847.9 ± 228.1 gaze locations, and 5.8 ± 1.2 m of bowel were recorded. 2d and 3d CNNs were trained using this data to predict bowel segmentation and diameter maps from the CT scans. For comparisons between two repetitions of visual annotation, CNN predictions, and manual annotations, Dice scores for bowel segmentation ranged from 0.69 ± 0.17 to 0.81 ± 0.04 and intraclass correlations [95% CI] for diameter measurement ranged from 0.672 [0.490-0.782] to 0.940 [0.933-0.947]. Thus, visual image annotation is a promising technique for training CNNs to perform bowel segmentation and diameter measurement in CT scans of patients with bowel obstruction.
Development and Validation of Multi-Omics Thymoma Risk Classification Model Based on Transfer Learning.
Liu W, Wang W, Zhang H … +3 more , Guo M, Xu Y, Liu X
J Digit Imaging · 2023 Oct · PMID 37268842 · Full text
The paper aims to develop prediction model that integrates clinical, radiomics, and deep features using transfer learning to stratifying between high and low risk of thymoma. Our study enrolled 150 patients with thymoma...
The paper aims to develop prediction model that integrates clinical, radiomics, and deep features using transfer learning to stratifying between high and low risk of thymoma. Our study enrolled 150 patients with thymoma (76 low-risk and 74 high-risk) who underwent surgical resection and pathologically confirmed in Shengjing Hospital of China Medical University from January 2018 to December 2020. The training cohort consisted of 120 patients (80%) and the test cohort consisted of 30 patients (20%). The 2590 radiomics and 192 deep features from non-enhanced, arterial, and venous phase CT images were extracted and ANOVA, Pearson correlation coefficient, PCA, and LASSO were used to select the most significant features. A fusion model that integrated clinical, radiomics, and deep features was developed with SVM classifiers to predict the risk level of thymoma, and accuracy, sensitivity, specificity, ROC curves, and AUC were applied to evaluate the classification model. In both the training and test cohorts, the fusion model demonstrated better performance in stratifying high and low risk of thymoma. It had AUCs of 0.99 and 0.95, and an accuracy of 0.93 and 0.83, respectively. This was compared to the clinical model (AUCs of 0.70 and 0.51, accuracy of 0.68 and 0.47), the radiomics model (AUCs of 0.97 and 0.82, accuracy of 0.93 and 0.80), and the deep model (AUCs of 0.94 and 0.85, accuracy of 0.88 and 0.80). The fusion model integrating clinical, radiomics and deep features based on transfer learning was efficient for noninvasively stratifying high risk and low risk of thymoma. The models could help to determine surgery strategy for thymoma cancer.
Automatic Image Segmentation and Grading Diagnosis of Sacroiliitis Associated with AS Using a Deep Convolutional Neural Network on CT Images.
Zhang K, Luo G, Li W … +10 more , Zhu Y, Pan J, Li X, Liu C, Liang J, Zhan Y, Zheng J, Li S, Cai W, Hong G
J Digit Imaging · 2023 Oct · PMID 37268841 · Full text
Ankylosing spondylitis (AS) is a chronic inflammatory disease that causes inflammatory low back pain and may even limit activity. The grading diagnosis of sacroiliitis on imaging plays a central role in diagnosing AS. Ho...
Ankylosing spondylitis (AS) is a chronic inflammatory disease that causes inflammatory low back pain and may even limit activity. The grading diagnosis of sacroiliitis on imaging plays a central role in diagnosing AS. However, the grading diagnosis of sacroiliitis on computed tomography (CT) images is viewer-dependent and may vary between radiologists and medical institutions. In this study, we aimed to develop a fully automatic method to segment sacroiliac joint (SIJ) and further grading diagnose sacroiliitis associated with AS on CT. We studied 435 CT examinations from patients with AS and control at two hospitals. No-new-UNet (nnU-Net) was used to segment the SIJ, and a 3D convolutional neural network (CNN) was used to grade sacroiliitis with a three-class method, using the grading results of three veteran musculoskeletal radiologists as the ground truth. We defined grades 0-I as class 0, grade II as class 1, and grades III-IV as class 2 according to modified New York criteria. nnU-Net segmentation of SIJ achieved Dice, Jaccard, and relative volume difference (RVD) coefficients of 0.915, 0.851, and 0.040 with the validation set, respectively, and 0.889, 0.812, and 0.098 with the test set, respectively. The areas under the curves (AUCs) of classes 0, 1, and 2 using the 3D CNN were 0.91, 0.80, and 0.96 with the validation set, respectively, and 0.94, 0.82, and 0.93 with the test set, respectively. 3D CNN was superior to the junior and senior radiologists in the grading of class 1 for the validation set and inferior to expert for the test set (P < 0.05). The fully automatic method constructed in this study based on a convolutional neural network could be used for SIJ segmentation and then accurately grading and diagnosis of sacroiliitis associated with AS on CT images, especially for class 0 and class 2. The method for class 1 was less effective but still more accurate than that of the senior radiologist.
An AI-Based Image Quality Control Framework for Knee Radiographs.
Sun H, Wang W, He F … +10 more , Wang D, Liu X, Xu S, Zhao B, Li Q, Wang X, Jiang Q, Zhang R, Liu S, Xiao Y
J Digit Imaging · 2023 Oct · PMID 37268840 · Full text
Image quality control (QC) is crucial for the accurate diagnosis of knee diseases using radiographs. However, the manual QC process is subjective, labor intensive, and time-consuming. In this study, we aimed to develop a...
Image quality control (QC) is crucial for the accurate diagnosis of knee diseases using radiographs. However, the manual QC process is subjective, labor intensive, and time-consuming. In this study, we aimed to develop an artificial intelligence (AI) model to automate the QC procedure typically performed by clinicians. We proposed an AI-based fully automatic QC model for knee radiographs using high-resolution net (HR-Net) to identify predefined key points in images. We then performed geometric calculations to transform the identified key points into three QC criteria, namely, anteroposterior (AP)/lateral (LAT) overlap ratios and LAT flexion angle. The proposed model was trained and validated using 2212 knee plain radiographs from 1208 patients and an additional 1572 knee radiographs from 753 patients collected from six external centers for further external validation. For the internal validation cohort, the proposed AI model and clinicians showed high intraclass consistency coefficients (ICCs) for AP/LAT fibular head overlap and LAT knee flexion angle of 0.952, 0.895, and 0.993, respectively. For the external validation cohort, the ICCs were also high, with values of 0.934, 0.856, and 0.991, respectively. There were no significant differences between the AI model and clinicians in any of the three QC criteria, and the AI model required significantly less measurement time than clinicians. The experimental results demonstrated that the AI model performed comparably to clinicians and required less time. Therefore, the proposed AI-based model has great potential as a convenient tool for clinical practice by automating the QC procedure for knee radiographs.
Enhancement of Non-Linear Deep Learning Model by Adjusting Confounding Variables for Bone Age Estimation in Pediatric Hand X-rays.
Kim KD, Kyung S, Jang M … +4 more , Ji S, Lee DH, Yoon HM, Kim N
J Digit Imaging · 2023 Oct · PMID 37268839 · Full text
In medicine, confounding variables in a generalized linear model are often adjusted; however, these variables have not yet been exploited in a non-linear deep learning model. Sex plays important role in bone age estimati...
In medicine, confounding variables in a generalized linear model are often adjusted; however, these variables have not yet been exploited in a non-linear deep learning model. Sex plays important role in bone age estimation, and non-linear deep learning model reported their performances comparable to human experts. Therefore, we investigate the properties of using confounding variables in a non-linear deep learning model for bone age estimation in pediatric hand X-rays. The RSNA Pediatric Bone Age Challenge (2017) dataset is used to train deep learning models. The RSNA test dataset is used for internal validation, and 227 pediatric hand X-ray images with bone age, chronological age, and sex information from Asan Medical Center (AMC) for external validation. U-Net based autoencoder, U-Net multi-task learning (MTL), and auxiliary-accelerated MTL (AA-MTL) models are chosen. Bone age estimations adjusted by input, output prediction, and without adjusting the confounding variables are compared. Additionally, ablation studies for model size, auxiliary task hierarchy, and multiple tasks are conducted. Correlation and Bland-Altman plots between ground truth and model-predicted bone ages are evaluated. Averaged saliency maps based on image registration are superimposed on representative images according to puberty stage. In the RSNA test dataset, adjusting by input shows the best performances regardless of model size, with mean average errors (MAEs) of 5.740, 5.478, and 5.434 months for the U-Net backbone, U-Net MTL, and AA-MTL models, respectively. However, in the AMC dataset, the AA-MTL model that adjusts the confounding variable by prediction shows the best performance with an MAE of 8.190 months, whereas the other models show the best performances by adjusting the confounding variables by input. Ablation studies of task hierarchy reveal no significant differences in the results of the RSNA dataset. However, predicting the confounding variable in the second encoder layer and estimating bone age in the bottleneck layer shows the best performance in the AMC dataset. Ablations studies of multiple tasks reveal that leveraging confounding variables plays an important role regardless of multiple tasks. To estimate bone age in pediatric X-rays, the clinical setting and balance between model size, task hierarchy, and confounding adjustment method play important roles in performance and generalizability; therefore, proper adjusting methods of confounding variables to train deep learning-based models are required for improved models.
An Effective Approach to Improve the Automatic Segmentation and Classification Accuracy of Brain Metastasis by Combining Multi-phase Delay Enhanced MR Images.
Chen M, Guo Y, Wang P … +6 more , Chen Q, Bai L, Wang S, Su Y, Wang L, Gong G
J Digit Imaging · 2023 Aug · PMID 37259008 · Full text
The objective of this study is to analyse the diffusion rule of the contrast media in multi-phase delayed enhanced magnetic resonance (MR) T1 images using radiomics and to construct an automatic classification and segmen...
The objective of this study is to analyse the diffusion rule of the contrast media in multi-phase delayed enhanced magnetic resonance (MR) T1 images using radiomics and to construct an automatic classification and segmentation model of brain metastases (BM) based on support vector machine (SVM) and Dpn-UNet. A total of 189 BM patients with 1047 metastases were enrolled. Contrast-enhanced MR images were obtained at 1, 3, 5, 10, 18, and 20 min following contrast medium injection. The tumour target volume was delineated, and the radiomics features were extracted and analysed. BM segmentation and classification models in the MR images with different enhancement phases were constructed using Dpn-UNet and SVM, and differences in the BM segmentation and classification models with different enhancement times were compared. (1) The signal intensity for BM decreased with time delay and peaked at 3 min. (2) Among the 144 optimal radiomics features, 22 showed strong correlation with time (highest R-value = 0.82), while 41 showed strong correlation with volume (highest R-value = 0.99). (3) The average dice similarity coefficients of both the training and test sets were the highest at 10 min for the automatic segmentation of BM, reaching 0.92 and 0.82, respectively. (4) The areas under the curve (AUCs) for the classification of BM pathology type applying single-phase MRI was the highest at 10 min, reaching 0.674. The AUC for the classification of BM by applying the six-phase image combination was the highest, reaching 0.9596, and improved by 42.3% compared with that by applying single-phase images at 10 min. The dynamic changes of contrast media diffusion in BM can be reflected by multi-phase delayed enhancement based on radiomics, which can more objectively reflect the pathological types and significantly improve the accuracy of BM segmentation and classification.
Weakly Supervised Breast Lesion Detection in Dynamic Contrast-Enhanced MRI.
Sun R, Wei C, Jiang Z … +3 more , Huang G, Xie Y, Nie S
J Digit Imaging · 2023 Aug · PMID 37253896 · Full text
Currently, obtaining accurate medical annotations requires high labor and time effort, which largely limits the development of supervised learning-based tumor detection tasks. In this work, we investigated a weakly super...
Currently, obtaining accurate medical annotations requires high labor and time effort, which largely limits the development of supervised learning-based tumor detection tasks. In this work, we investigated a weakly supervised learning model for detecting breast lesions in dynamic contrast-enhanced MRI (DCE-MRI) with only image-level labels. Two hundred fifty-four normal and 398 abnormal cases with pathologically confirmed lesions were retrospectively enrolled into the breast dataset, which was divided into the training set (80%), validation set (10%), and testing set (10%) at the patient level. First, the second image series S2 after the injection of a contrast agent was acquired from the 3.0-T, T1-weighted dynamic enhanced MR imaging sequences. Second, a feature pyramid network (FPN) with convolutional block attention module (CBAM) was proposed to extract multi-scale feature maps of the modified classification network VGG16. Then, initial location information was obtained from the heatmaps generated using the layer class activation mapping algorithm (Layer-CAM). Finally, the detection results of breast lesion were refined by the conditional random field (CRF). Accuracy, sensitivity, specificity, and area under the receiver operating characteristic (ROC) curve (AUC) were utilized for evaluation of image-level classification. Average precision (AP) was estimated for breast lesion localization. Delong's test was used to compare the AUCs of different models for significance. The proposed model was effective with accuracy of 95.2%, sensitivity of 91.6%, specificity of 99.2%, and AUC of 0.986. The AP for breast lesion detection was 84.1% using weakly supervised learning. Weakly supervised learning based on FPN combined with Layer-CAM facilitated automatic detection of breast lesion.
About
- Frequency
- Sun
- Papers found
- 200
- RSS feed
- Subscribe