Automated assessment of pelvic radiographs using deep learning: A reliable diagnostic tool for pelvic malalignment.

Pelvic malalignment leads to general imbalance and adversely affects leg length. Timely and accurate diagnosis of pelvic alignment in patients is crucial to prevent additional complications arising from delayed treatment. Currently, doctors typically assess pelvic alignment either manually or through radiography. This study aimed to develop and assess the validity of a deep learning-based system for automatically measuring 10 radiographic parameters necessary for diagnosing pelvic displacement using standing anteroposterior pelvic X-rays. Between March 2016 and June 2021, pelvic radiographs from 1215 patients were collected. After applying specific selection criteria, 550 pelvic radiographs were chosen for analysis. These data were utilized to develop a deep learning-based system capable of automatically measuring radiographic parameters relevant to pelvic displacement diagnosis. The system's diagnostic accuracy was evaluated by comparing automatically measured values with those assessed by a clinician using 200 radiographs selected from the initial 550. The results indicated that the system exhibited high reliability, accuracy, and reproducibility, with a Pearson correlation coefficient of ≥0.9, an intra-class correlation coefficient of ≥0.9, a mean absolute error of ≤1 cm, mean square error of ≤1 cm, and root mean square error of ≤1 cm. Moreover, the system's measurement time for a single radiograph was found to be 18 to 20 times faster than that required by a clinician for manual inspection. In conclusion, our proposed deep learning-based system effectively utilizes standing anteroposterior pelvic radiographs to precisely and consistently measure radiographic parameters essential for diagnosing pelvic displacement.

Transformers in medical image segmentation: a narrative review.

Background and Objective: Transformers, which have been widely recognized as state-of-the-art tools in natural language processing (NLP), have also come to be recognized for their value in computer vision tasks. With this increasing popularity, they have also been extensively researched in the more complex medical imaging domain. The associated developments have resulted in transformers being on par with sought-after convolution neural networks, particularly for medical image segmentation. Methods combining both types of networks have proven to be especially successful in capturing local and global contexts, thereby significantly boosting their performances in various segmentation problems. Motivated by this success, we have attempted to survey the consequential research focused on innovative transformer networks, specifically those designed to cater to medical image segmentation in an efficient manner. Methods: Databases like Google Scholar, arxiv, ResearchGate, Microsoft Academic, and Semantic Scholar have been utilized to find recent developments in this field. Specifically, research in the English language from 2021 to 2023 was considered. Key Content and Findings: In this survey, we look into the different types of architectures and attention mechanisms that uniquely improve performance and the structures that are in place to handle complex medical data. Through this survey, we summarize the popular and unconventional transformer-based research as seen through different key angles and analyze quantitatively the strategies that have proven more advanced. Conclusions: We have also attempted to discern existing gaps and challenges within current research, notably highlighting the deficiency of annotated medical data for precise deep learning model training. Furthermore, potential future directions for enhancing transformers' utility in healthcare are outlined, encompassing strategies such as transfer learning and exploiting foundation models for specialized medical image segmentation.

Ensemble deep-learning networks for automated osteoarthritis grading in knee X-ray images.

The Kellgren-Lawrence (KL) grading system is a scoring system for classifying the severity of knee osteoarthritis using X-ray images, and it is the standard X-ray-based grading system for diagnosing knee osteoarthritis. However, KL grading depends on the clinician's subjective assessment. Moreover, the accuracy varies significantly depending on the clinician's experience and can be particularly low. Therefore, in this study, we developed an ensemble network that can predict a consistent and accurate KL grade for knee osteoarthritis severity using a deep learning approach. We trained individual models on knee X-ray datasets using the most suitable image size for each model in an ensemble network rather than using datasets with a single image size. We then built the ensemble network using these models to overcome the instability of single models and further improve accuracy. We conducted various experiments using a dataset of 8260 images from the Osteoarthritis Initiative open dataset. The proposed ensemble network exhibited the best performance, achieving an accuracy of 76.93% and an F1-score of 0.7665. The Grad-CAM visualization technique was used to further evaluate the focus of the model. The results demonstrated that the proposed ensemble network outperforms existing techniques that have performed well in KL grade classification. Moreover, the proposed model focuses on the joint space around the knee to extract the imaging features required for KL grade classification, revealing its high potential for diagnosing knee osteoarthritis.

A deep learning approach for fully automated measurements of lower extremity alignment in radiographic images.

During clinical evaluation of patients and planning orthopedic treatments, the periodic assessment of lower limb alignment is critical. Currently, physicians use physical tools and radiographs to directly observe limb alignment. However, this process is manual, time consuming, and prone to human error. To this end, a deep-learning (DL)-based system was developed to automatically, rapidly, and accurately detect lower limb alignment by using anteroposterior standing X-ray medical imaging data of lower limbs. For this study, leg radiographs of non-overlapping 770 patients were collected from January 2016 to August 2020. To precisely detect necessary landmarks, a DL model was implemented stepwise. A radiologist compared the final calculated measurements with the observations in terms of the concordance correlation coefficient (CCC), Pearson correlation coefficient (PCC), and intraclass correlation coefficient (ICC). Based on the results and 250 frontal lower limb radiographs obtained from 250 patients, the system measurements for 16 indicators revealed superior reliability (CCC, PCC, and ICC ≤ 0.9; mean absolute error, mean square error, and root mean square error ≥ 0.9) for clinical observations. Furthermore, the average measurement speed was approximately 12 s. In conclusion, the analysis of anteroposterior standing X-ray medical imaging data by the DL-based lower limb alignment diagnostic support system produces measurement results similar to those obtained by radiologists.

Unsupervised anomaly detection for posteroanterior chest X-rays using multiresolution patch-based self-supervised learning.

The demand for anomaly detection, which involves the identification of abnormal samples, has continued to increase in various domains. In particular, with increases in the data volume of medical imaging, the demand for automated screening systems has also risen. Consequently, in actual clinical practice, radiologists can focus only on diagnosing patients with abnormal findings. In this study, we propose an unsupervised anomaly detection method for posteroanterior chest X-rays (CXR) using multiresolution patch-based self-supervised learning. The core aspect of our approach is to leverage patch images of different sizes for training and testing to recognize diverse anomalies characterized by unknown shapes and scales. In addition, self-supervised contrastive learning is applied to learn the generalized and robust features of the patches. The performance of the proposed method is evaluated using posteroanterior CXR images from a public dataset for training and testing. The results show that the proposed method is superior to state-of-the-art anomaly detection methods. In addition, unlike single-resolution patch-based methods, the proposed method consistently exhibits a good overall performance regardless of the evaluation criteria used for comparison, thus demonstrating the effectiveness of using multiresolution patch-based features. Overall, the results of this study validate the effectiveness of multiresolution patch-based self-supervised learning for detecting anomalies in CXR images.

Dual-Stage Deeply Supervised Attention-Based Convolutional Neural Networks for Mandibular Canal Segmentation in CBCT Scans.

Accurate segmentation of mandibular canals in lower jaws is important in dental implantology. Medical experts manually determine the implant position and dimensions from 3D CT images to avoid damaging the mandibular nerve inside the canal. In this paper, we propose a novel dual-stage deep learning-based scheme for the automatic segmentation of the mandibular canal. In particular, we first enhance the CBCT scans by employing the novel histogram-based dynamic windowing scheme, which improves the visibility of mandibular canals. After enhancement, we designed 3D deeply supervised attention UNet architecture for localizing the Volumes Of Interest (VOIs), which contain the mandibular canals (i.e., left and right canals). Finally, we employed the Multi-Scale input Residual UNet (MSiR-UNet) architecture to segment the mandibular canals using VOIs accurately. The proposed method has been rigorously evaluated on 500 and 15 CBCT scans from our dataset and from the public dataset, respectively. The results demonstrate that our technique improves the existing performance of mandibular canal segmentation to a clinically acceptable range. Moreover, it is robust against the types of CBCT scans in terms of field of view.

Evaluation of Automated Measurement of Hair Density Using Deep Neural Networks.

Recently, deep learning has been employed in medical image analysis for several clinical imaging methods, such as X-ray, computed tomography, magnetic resonance imaging, and pathological tissue imaging, and excellent performance has been reported. With the development of these methods, deep learning technologies have rapidly evolved in the healthcare industry related to hair loss. Hair density measurement (HDM) is a process used for detecting the severity of hair loss by counting the number of hairs present in the occipital donor region for transplantation. HDM is a typical object detection and classification problem that could benefit from deep learning. This study analyzed the accuracy of HDM by applying deep learning technology for object detection and reports the feasibility of automating HDM. The dataset for training and evaluation comprised 4492 enlarged hair scalp RGB images obtained from male hair-loss patients and the corresponding annotation data that contained the location information of the hair follicles present in the image and follicle-type information according to the number of hairs. EfficientDet, YOLOv4, and DetectoRS were used as object detection algorithms for performance comparison. The experimental results indicated that YOLOv4 had the best performance, with a mean average precision of 58.67.

Computer-aided automatic measurement of leg length on full leg radiographs.

OBJECTIVES: To develop and evaluate a deep learning (DL)-based system for measuring leg length on full leg radiographs of diverse patients, including those with orthopedic hardware implanted for surgical treatment. METHODS: This study retrospectively assessed 2767 X-ray scanograms of 2767 patients who did or did not have orthopedic hardware implanted between January 2016 and December 2019. A cascaded DL model was developed to localize the relevant landmarks on the pelvis, knees, and ankles required for measuring leg length. Statistical analysis was performed using the correlation coefficient analysis and Bland-Altman plots to assess the agreement between the reference standard and DL-calculated lengths. RESULTS: Testing data comprised 400 radiographs from 400 patients. Of these radiographs, 100 were from patients with orthopedic hardware implanted in their pelvis, knees, or ankles. For all testing data, leg lengths derived from the DL-based measurement system, with or without internal fixation devices, showed excellent agreement with the reference standard (femoral length, r = 0.99 (P < .001); root mean square error (RMSE) = 0.17 cm; mean difference, - 0.01 ± 0.17 cm; 95% limit of agreement (LoA), - 0.35 to 0.34; tibial length, r = 0.99 (P < .001); RMSE = 0.17 cm; mean difference, - 0.02 ± 0.17 cm, 95% LoA, - 0.34 to 0.31; and full leg length, r = 1.0 (P < .001); RMSE = 0.19 cm; mean difference, 0.05 ± 0.18 cm; 95% LoA, - 0.31 to 0.40). The mean time for leg length measurement for each patient using the DL-based system was 8.68 ± 0.18 s. CONCLUSION: The DL-based leg length measurement system could provide similar performance to radiologists in terms of accuracy and reliability for a diverse group of patients.

Evaluation of the feasibility of explainable computer-aided detection of cardiomegaly on chest radiographs using deep learning.

We examined the feasibility of explainable computer-aided detection of cardiomegaly in routine clinical practice using segmentation-based methods. Overall, 793 retrospectively acquired posterior-anterior (PA) chest X-ray images (CXRs) of 793 patients were used to train deep learning (DL) models for lung and heart segmentation. The training dataset included PA CXRs from two public datasets and in-house PA CXRs. Two fully automated segmentation-based methods using state-of-the-art DL models for lung and heart segmentation were developed. The diagnostic performance was assessed and the reliability of the automatic cardiothoracic ratio (CTR) calculation was determined using the mean absolute error and paired t-test. The effects of thoracic pathological conditions on performance were assessed using subgroup analysis. One thousand PA CXRs of 1000 patients (480 men, 520 women; mean age 63 ± 23 years) were included. The CTR values derived from the DL models and diagnostic performance exhibited excellent agreement with reference standards for the whole test dataset. Performance of segmentation-based methods differed based on thoracic conditions. When tested using CXRs with lesions obscuring heart borders, the performance was lower than that for other thoracic pathological findings. Thus, segmentation-based methods using DL could detect cardiomegaly; however, the feasibility of computer-aided detection of cardiomegaly without human intervention was limited.

Automated Bone Age Assessment Using Artificial Intelligence: The Future of Bone Age Assessment.

Bone age assessments are a complicated and lengthy process, which are prone to inter- and intra-observer variabilities. Despite the great demand for fully automated systems, developing an accurate and robust bone age assessment solution has remained challenging. The rapidly evolving deep learning technology has shown promising results in automated bone age assessment. In this review article, we will provide information regarding the history of automated bone age assessments, discuss the current status, and present a literature review, as well as the future directions of artificial intelligence-based bone age assessments.

Automatic Lung Segmentation on Chest X-rays Using Self-Attention Deep Neural Network.

Accurate identification of the boundaries of organs or abnormal objects (e.g., tumors) in medical images is important in surgical planning and in the diagnosis and prognosis of diseases. In this study, we propose a deep learning-based method to segment lung areas in chest X-rays. The novel aspect of the proposed method is the self-attention module, where the outputs of the channel and spatial attention modules are combined to generate attention maps, with each highlighting those regions of feature maps that correspond to "what" and "where" to attend in the learning process, respectively. Thereafter, the attention maps are multiplied element-wise with the input feature map, and the intermediate results are added to the input feature map again for residual learning. Using X-ray images collected from public datasets for training and evaluation, we applied the proposed attention modules to U-Net for segmentation of lung areas and conducted experiments while changing the locations of the attention modules in the baseline network. The experimental results showed that our method achieved comparable or better performance than the existing medical image segmentation networks in terms of Dice score when the proposed attention modules were placed in lower layers of both the contracting and expanding paths of U-Net.

Evaluation of the clinical efficacy of a TW3-based fully automated bone age assessment system using deep neural networks.

PURPOSE: The aim of this study was to evaluate the clinical efficacy of a Tanner-Whitehouse 3 (TW3)-based fully automated bone age assessment system on hand-wrist radiographs of Korean children and adolescents. MATERIALS AND METHODS: Hand-wrist radiographs of 80 subjects (40 boys and 40 girls, 7-15 years of age) were collected. The clinical efficacy was evaluated by comparing the bone ages that were determined using the system with those from the reference standard produced by 2 oral and maxillofacial radiologists. Comparisons were conducted using the paired t-test and simple regression analysis. RESULTS: The bone ages estimated with this bone age assessment system were not significantly different from those obtained with the reference standard (P>0.05) and satisfied the equivalence criterion of 0.6 years within the 95% confidence interval (- 0.07 to 0.22), demonstrating excellent performance of the system. Similarly, in the comparisons of gender subgroups, no significant difference in bone age between the values produced by the system and the reference standard was observed (P>0.05 for both boys and girls). The determination coefficients obtained via regression analysis were 0.962, 0.945, and 0.952 for boys, girls, and overall, respectively (P=0.000); hence, the radiologist-determined bone ages and the system-determined bone ages were strongly correlated. CONCLUSION: This TW3-based system can be effectively used for bone age assessment based on hand-wrist radiographs of Korean children and adolescents.

Volumetric lung nodule segmentation using adaptive ROI with multi-view residual learning.

Accurate quantification of pulmonary nodules can greatly assist the early diagnosis of lung cancer, enhancing patient survival possibilities. A number of nodule segmentation techniques, which either rely on a radiologist-provided 3-D volume of interest (VOI) or use the constant region of interests (ROIs) for all the slices, are proposed; however, these techniques can only investigate the presence of nodule voxels within the given VOI. Such approaches restrain the solutions to freely investigate the nodule presence outside the given VOI and also include the redundant structures (non-nodule) into VOI, which limits the segmentation accuracy. In this work, a novel semi-automated approach for 3-D segmentation of lung nodule in computerized tomography scans, has been proposed. The technique is segregated into two stages. In the first stage, a 2-D ROI containing the nodule is provided as an input to perform a patch-wise exploration along the axial axis using a novel adaptive ROI algorithm. This strategy enables the dynamic selection of the ROI in the surrounding slices to investigate the presence of nodules using a Deep Residual U-Net architecture. This stage provides the initial estimation of the nodule utilized to extract the VOI. In the second stage, the extracted VOI is further explored along the coronal and sagittal axes, in patchwise fashion, with Residual U-Nets. All the estimated masks are then fed into a consensus module to produce a final volumetric segmentation of the nodule. The algorithm is rigorously evaluated on LIDC-IDRI dataset, which is the largest publicly available dataset. The proposed approach achieved the average dice score of 87.5%, which is significantly higher than the existing state-of-the-art techniques.

Retrospective Motion Correction in Multishot MRI using Generative Adversarial Network.

Multishot Magnetic Resonance Imaging (MRI) is a promising data acquisition technique that can produce a high-resolution image with relatively less data acquisition time than the standard spin echo. The downside of multishot MRI is that it is very sensitive to subject motion and even small levels of motion during the scan can produce artifacts in the final magnetic resonance (MR) image, which may result in a misdiagnosis. Numerous efforts have focused on addressing this issue; however, all of these proposals are limited in terms of how much motion they can correct and require excessive computational time. In this paper, we propose a novel generative adversarial network (GAN)-based conjugate gradient SENSE (CG-SENSE) reconstruction framework for motion correction in multishot MRI. First CG-SENSE reconstruction is employed to reconstruct an image from the motion-corrupted k-space data and then the GAN-based proposed framework is applied to correct the motion artifacts. The proposed method has been rigorously evaluated on synthetically corrupted data on varying degrees of motion, numbers of shots, and encoding trajectories. Our analyses (both quantitative as well as qualitative/visual analysis) establish that the proposed method is robust and reduces several-fold the computational time reported by the current state-of-the-art technique.