Current Medical Imaging - Volume 20, Issue 1, 2024

Ingestion of gastrointestinal foreign bodies (FB) is a common clinical problem worldwide. Approximately 10–20% of FBs require an endoscopic procedure for removal, and < 1% require surgery.

An 89-year-old male with Alzheimer's disease was hospitalized because of abdominal pain, abdominal distention, vomiting for three days, and cessation of bowel movements for six days. Abdominal computed tomography (CT) scan showed a small intestinal obstruction and an atypical FB in the small intestine. A pill and remaining plastic casing were removed from the small intestine during surgery. FB is a square with four sharp acute angles at its edge. The patient was discharged after two weeks of treatment, and no recurrence or complications were observed during the 6-month follow-up.

Atypical intestinal FBs may cause misdiagnosis and easily lead to serious complications. Therefore, an appropriate radiological examination, such as CT, is necessary for unexplained intestinal obstruction. Symptomatic intestinal FBs should be actively removed to avoid serious complications.

Dural sinus arteriovenous malformation is an uncommon intracranial vascular malformation. The affected cases may suffer from severe neurological injury. Prenatal ultrasound has been used to diagnose fetal intracranial vascular abnormality, but prenatal three-dimensional (3D) ultrasound presents a very rare anomaly; an arteriovenous malformation of the dural sinus has not been reported.

This study aimed to emphasize the diagnostic value of 3D ultrasound in the fetus with dural sinus arteriovenous malformation.

A 38-year-old woman was referred for targeted fetal ultrasonography at 37 weeks of gestation due to an ultrasound that showed a cystic lesion in the posterior cranial fossa. The fetus demonstrated obvious dilatation of the torcular herophili, bilateral transverse sinuses, and bilateral sigmoid sinuses, appearing as a novel bull's horn sign on 3D ultrasound. After birth, cerebral angiography confirmed the diagnosis of dural arteriovenous fistula (DAVF) in the occipital sinus region.

3D ultrasound is an appealing method for prenatal diagnosis of dural sinus arteriovenous malformation.

Brain tumours represent a diagnostic challenge, especially in the imaging area, where the differentiation of normal and pathologic tissues should be precise. The use of up-to-date machine learning techniques would be of great help in terms of brain tumor identification accuracy from MRI data.

This research paper aims to check the efficiency of a federated learning method that joins two classifiers, such as convolutional neural networks (CNNs) and random forests (R.F.F.), with dual U-Net segmentation for federated learning. This procedure benefits the image identification task on preprocessed MRI scan pictures that have already been categorized.

In addition to using a variety of datasets, federated learning was utilized to train the CNN-RF model while taking data privacy into account. The processed MRI images with Median, Gaussian, and Wiener filters are used to filter out the noise level and make the feature extraction process easy and efficient. The surgical part used a dual U-Net layout, and the performance assessment was based on precision, recall, F1-score, and accuracy.

The model achieved excellent classification performance on local datasets as CRPs were high, from 91.28% to 95.52% for macro, micro, and weighted averages. Throughout the process of federated averaging, the collective model outperformed by reaching 97% accuracy compared to those of 99%, which were subjected to different clients. The correctness of how data is used helps the federated averaging method convert individual model insights into a consistent global model while keeping all personal data private.

The combined structure of the federated learning framework, CNN-RF hybrid model, and dual U-Net segmentation is a robust and privacy-preserving approach for identifying MRI images from brain tumors. The results of the present study exhibited that the technique is promising in improving the quality of brain tumor categorization and provides a pathway for practical utilization in clinical settings.

In this study, we harnessed three cutting-edge algorithms' capabilities to refine the elbow fracture prediction process through X-ray image analysis. Employing the YOLOv8 (You only look once) algorithm, we first identified Regions of Interest (ROI) within the X-ray images, significantly augmenting fracture prediction accuracy.

Subsequently, we integrated and compared the ResNet, the SeResNet (Squeeze-and-Excitation Residual Network) ViT (Vision Transformer) algorithms to refine our predictive capabilities. Furthermore, to ensure optimal precision, we implemented a series of meticulous refinements. This included recalibrating ROI regions to enable finer-grained identification of diagnostically significant areas within the X-ray images. Additionally, advanced image enhancement techniques were applied to optimize the X-ray images' visual quality and structural clarity.

These methodological enhancements synergistically contributed to a substantial improvement in the overall accuracy of our fracture predictions. The dataset utilized for training, testing & validation, and comprehensive evaluation exclusively comprised elbow X-ray images, where predicting the fracture with three algorithms: Resnet50; accuracy 0.97, precision 1, recall 0.95, SeResnet50; accuracy 0.97, precision 1, recall 0.95 & ViT-B-16 with high accuracy of 0.99, precision same as the other two algorithms, with a recall of 0.95.

This approach has the potential to increase the precision of diagnoses, lessen the burden of radiologists, easily integrate into current medical imaging systems, and assist clinical decision-making, all of which could lead to better patient care and health outcomes overall.

Guizhi Shaoyao Zhimu Decoction can be used in the treatment of rheumatoid arthritis, but there is scarce literature on using ultrasound combined with electrophysiology to evaluate the efficacy of this traditional Chinese medicine.

This study aimed to explore the clinical effect of Guizhi Shaoyao Zhimu decoction on cold-dampness arthralgia rheumatoid arthritis (RA) by ultrasound and electrophysiological examination.

A total of 64 patients with rheumatoid arthritis were randomly divided into two groups, with 32 cases in each group. The control group was treated with conventional western medicine, and the experimental group was treated with Guizhi Shaoyao Zhimu Decoction in addition to conventional western medicine. After 4 weeks of treatment, traditional Chinese medicine (TCM) symptom scores, erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), 28 joint disease range of motion score (DAS28), ultrasonic score, and electrophysiological examination results were observed.

There were significant differences in TCM syndrome scores, ESR, CRP, DAS28, and ultrasound scores in the two groups before and after treatment (P<0.05). Compared between the two groups after treatment, there were statistically significant differences in TCM syndrome scores, ESR, CRP, DAS28, and ultrasound scores (P<0.05). The motor nerve conduction velocity (MNCV), sensory nerve conduction velocity (SNCV), and action potential (AP) of the median nerve and ulnar nerve in the experimental group were significantly increased compared with the control group (P<0.05).

Guizhi Shaoyao Zhimu Decoction combined with conventional western medicine has a significant effect on cold-dampness arthralgia rheumatoid arthritis, and ultrasound and electrophysiological examination can be used to evaluate its curative effect.

The maxillary sinuses are air-filled cavities which vary in size and shape. Sinus radiography has been widely used in the determination of the gender of the individual, especially in forensic investigation for human identification and sexing of individuals. The advanced radiographic techniques like cone beam computed tomography (CBCT), especially the axial and coronal sections, have been considered as a subtle concept in forensic odontology. Aim: The current study aimed in evaluating the parameters of the maxillary sinus using CBCT and to identify its implication in gender determination.

Current study consists of 50 patients who were divided into two groups, group I consisted of 25 males and group II consisted of 25 females, where maxillary sinus dimensions like maximum length (anteroposteriorly), maximum width (mediolaterally) and maximum height (superioinferiorly) were evaluated using CBCT scans in axial and coronal sections respectively.

Shapiro-Wilk test was used to determine the normality and Independent t-test was used to compare the two groups, followed by predictive analysis. Maxillary sinus, right length (p<0.001), right width (p<0.001), right height (p<0.001), left length (p<0.001), left width (p<0.001), left height (p<0.001). Right and left maxillary sinus parameters were different between males and females, with statistical significance indicating the presence of sexual dimorphism.

In this study, maxillary sinus parameters like length, width and height in CBCT were significantly different between males and females. Maxillary sinus can be a useful gender predictor in the forensic identification of the individual.

This study intends to provide a novel Invasive Weed Optimization (IWO) algorithm for the detection of Polycystic Ovary Syndrome (PCOS) from ultrasound ovarian images. PCOS is an intricate anarchy described by hyperandrogenemia and irregular menstruation. Indian women are increasingly finding reproductive disorders, namely PCOS.

The women having PCOS grow more small follicles in their ovaries. The radiologists take a look into women's ovaries by use of ultrasound scanning equipment to manually count the number of follicles and their size for fertility treatment. These may lead to error diagnosis.

This paper proposed an automatic follicle detection system for identifying PCOS in the ovary using IWO. The performance of IWO is improved in Modified Invasive Weed Optimization (MIWO). This algorithm imitates the biological weeds' behavior. The MIWO is employed to obtain the optimal threshold by maximizing the between-class variance of the modified Otsu method. The efficiency of the proposed method has been compared with the well-known optimization technique called Particle Swarm Optimization (PSO) and with IWO.

Experimental results proved that the MIWO finds an optimal threshold higher than that of IWO and PSO.

This study aimed to establish a model based on Multi-detector Computed Tomography (MDCT), ¹⁸F-fluorodeoxyglucose Positron Emission Tomography/Computed Tomography (¹⁸F-FDG PET/CT), and clinical features for predicting different growth patterns of preoperative Invasive Adenocarcinoma (IAC).

This retrospective study included 357 patients diagnosed with IAC who underwent surgical treatment. According to pathological subtypes, IAC was classified into low-risk growth patterns (lepidic, acinar) and high-risk growth patterns (papillary, micropapillary, and solid). The clinical features of patients, preoperative MDCT, and ¹⁸F-FDG PET imaging characteristics were collected. Logistic regression analysis was used to determine the independent risk factors for the high-risk growth pattern of IAC and construct models for predicting the high-/low-risk growth patterns of IAC. Receiver operating characteristics and calibration curves were plotted and Decision Curve Analysis (DCA) was performed to evaluate the performance and clinical benefits of the models, respectively.

Gender, tumor location, size, spiculation, and SUVavg were independent risk factors for high-risk growth patterns of IAC. The PET/CT imaging-clinical characteristics combined model could well identify high-/low-risk growth patterns of IAC (AUC=0.789), which outperformed the CT model (AUC=0.689, p=0.0012), PET model (AUC=0.742, p=0.0022), and clinical model (AUC=0.607, p<0.0001). The calibration curve indicated good coherence between all model predictions and actual observations in both training and test sets (p>0.05). DCA revealed the highest clinical benefit of PET/CT imaging-clinical characteristics combined model in identifying the high-risk growth pattern of IAC.

The PET/CT imaging-clinical model based on multiphase MDCT features, ¹⁸F-FDG PET features, and clinical characteristics could predict the high-risk growth pattern of IAC preoperatively, aiding clinicians in deciding personalized treatment strategies.

Kikuchi-Fujimoto Disease (KFD) is a rare condition, distinguished by its hallmark presentation of regional lymphadenopathy in young adult females. While initially observed to exclusively affect cervical lymph nodes in females under 40 years old, KFD is now known to impact individuals of any age or gender and manifest with adenopathy in various anatomical sites. Nonspecific imaging findings for KFD include enlarged lymph nodes, often exhibiting abnormal morphology.

In this study, we present the case of a 49 year old asymptomatic woman, in whom several enlarged left axillary lymph nodes were incidentally noted during routine mammography. The diagnosis of KFD was determined via ultrasound-guided core needle biopsy. Histological examination of the biopsied lymph node revealed necrotizing lymphadenitis, consistent with KFD.

The uncommon and broad presentation of KFD highlights the significance of acquiring tissue samples to distinguish this condition from resembling malignancies or autoimmune disorders.

The use of Magnetic Resonance Imaging (MRI) and radiomics improves the management of Prostate Cancer (PCa) and helps in differentiating between clinically insignificant and significant PCa. This study has explored the diagnostic value of radiomic analysis based on functional parameter maps from monoexponential and diffusion kurtosis models in MRI for differentiating between clinically insignificant and significant PCa.

In total, 105 PCa cases, including 38 clinically insignificant and 67 clinically significant PCa cases, were retrospectively analyzed. The patients were randomly divided into training and testing sets in a ratio of 7:3. Univariate and multivariate logistic regression analyses were performed, and 1,352 radiomic features were extracted from ADC, MD, and MK images. Clinical, radiomic, and clinical–radiomic models were developed and compared using receiver operating characteristic curve analysis, decision curve analysis, and calibration curves.

Clinical variables, such as TPSA, PI-RADS, and FPSA, were identified as independent risk factors for differentiating between clinically insignificant and significant PCa. In radiomics, three features were identified as highly weighted indicators. The clinical–radiomic model based on the clinical and radiomic features demonstrated the highest predictive efficacy for clinically insignificant and significant PCa, with area under the curve values of 0.940 and 0.861 in the training and test sets, respectively.

The predictive model constructed from clinical and radiomic features exhibited substantial diagnostic differentiation capabilities for clinically insignificant and significant PCa. The clinical–radiomic model displayed the highest predictive performance, promising significant contributions to future clinical treatment and assessment of PCa.

Three-dimensional (3D) whole-heart magnetic resonance imaging (MRI) is an excellent tool to check the heart anatomy of patients with congenital and acquired heart disease. However, most 3D whole-heart MRI acquisitions take a long time to perform, and the sequence used is susceptible to banding artifacts.

To validate an unsupervised neural network that can reduce acquisition time and improve image quality for 3D whole-heart MRI by super-resolving low-resolution images.

The results of the super-resolution neural network (SRNN) were compared with bilinear interpolation, a state-of-the-art method known as AdapSR, and the ground truth high-resolution images qualitatively and quantitatively. Thirty pediatric patients with varying congenital and acquired heart diseases were included in this study. Results from the SRNN without a ground truth image were compared qualitatively with the contrast-enhanced whole-heart images. Signal-to-noise ratio (SNR) was used to quantitatively compare each of the methods and the high-resolution ground truth.

As confirmed by both the quantitative and qualitative results, the SRNN improves image quality. Furthermore, because it only requires a low-resolution acquisition, the use of the SRNN reduces acquisition time.

The SRNN lessens noise and eliminates artifacts while maintaining correct anatomical structure in the images.

The permanent canine usually has a single root and a single root canal. A one-rooted canine with two canals or a canine with two roots and two separate canals may also occur at a lower incidence in the permanent dentition. However, bilateral symmetrical mandibular canines with two roots and two separate canals are less common.

This study reported a lower incidence case of bilateral symmetrical mandibular canines with two roots and two separate canals, which was found based on a CBCT examinaton. The patient visited our department and was consulted for orthodontic treatment due to the irregularity of her lower anterior teeth. As the abnormal root morphology of bilateral mandibular canines greatly increased the difficulty of orthodontic treatment, the patient finally gave up orthodontic treatment after communication.

This case report provides supplementary data to better understand the complexities of the root canal system of canines.

Multimodal medical image fusion techniques play an important role in clinical diagnosis and treatment planning. The process of combining multimodal images involves several challenges depending on the type of modality, transformation techniques, and mapping of structural and metabolic information.

Artifacts can form during data acquisition, such as minor movement of patients, or data pre-processing, registration, and normalization. Unlike single-modality images, the detection of an artifact is a more challenging task in complementary fused multimodal images. Many medical image fusion techniques have been developed by various researchers, but not many have tested the robustness of their fusion approaches. The main objective of this study is to identify and classify the noise and artifacts present in the fused MRI-SPECT brain images using transfer learning by fine-tuned CNN networks. Deep neural network-based techniques are capable of detecting minor amounts of noise in images. In this study, three pre-trained convolutional neural network-based models (ResNet50, DenseNet 169, and InceptionV3) were used to detect artifacts and various noises including Gaussian, Speckle, Random, and mixed noises present in fused MRI -SPECT brain image datasets using transfer learning.

The five-fold stratified cross-validation (SCV) is used to evaluate the performance of networks. The obtained performance results for the pre-trained DenseNet169 model for various folds were greater as compared with the rest of the models; the former had an average accuracy of five-fold of 93.8±5.8%, 98%±3.9%, 97.8±1.64%, and 93.8±5.8%, whereas InceptionNetV3 had a value of 90.6±9.8%, 98.8±1.6%, 91.4±9.74%, and 90.6±9.8%, and ResNet50 had a value of 75.8±21%.84.8±7.6%, 73.8±22%, and 75.8±21% for Gaussian, speckle, random and mixed noise, respectively.

Based on the performance results obtained, the pre-trained DenseNet 169 model provides the highest accuracy among the other four used models.

Immune checkpoint inhibitor-related pneumonitis (ICI-P) is a fatal adverse event of immunotherapy. However, there is a lack of methods to identify patients who have a high risk of developing ICI-P in immunotherapy.

We aim at predicting the individualized risk of developing ICI-P by computed tomography (CT) images and deep learning to assist in personalized immunotherapy planning.

We first explored the prognostic value of the commonly used clinical factors. Moreover, we proposed a novel whole-lung analysis deep learning (DL) model, which is constructed using a combination of Densely Connected Convolutional Networks (DenseNet) and Feature Pyramid Networks (FPN). This DL model mines global lung information from CT images for predicting the risk of developing ICI-P, and it is fully automated and does not require manually annotating images. Finally, 157 patients were collected and randomly divided into training and testing sets for performance evaluation.

In the testing set, the clinical model achieved an Area Under the Curve (AUC) of 0.710 and accuracy of 0.625. By mining global lung information, the DL model achieved AUC=0.780 and accuracy=0.729 in the testing set, where the DL score revealed a significant difference between ICI-P and non-ICI-P patients. Through deep learning visualization technique, we found that many areas outside of tumor (e.g., pleural retraction, pleural effusion, and the abnormalities in vessels) are important for predicting the risk of developing ICI-P in immunotherapy.

The whole-lung analysis DL model provides an easy-to-use method for identifying patients at high risk of developing ICI-P by CT images, which is important for individualized treatment planning in immunotherapy. The performance improvement over the clinical model indicates that mining whole-lung information in CT images is effective for prognostic prediction in immunotherapy.

Applications of artificial intelligence (AI) in medical ultrasound have rapidly grown in recent years. Therefore, it is necessary to identify and visualize global research trends and hotspots of AI in medical ultrasound to provide guidance for further exploitation.

This study aims to highlight the global research trends and hotspots of the top 100 most-cited papers related to AI in medical ultrasound by combining quantitative and visualization methods.

Articles on AI in medical ultrasound were selected from the WoSCC database and ranked by citation count. After identifying the 100 most-cited papers, we conducted a quantitative and visualized analysis of bibliometric characteristics, including leading research countries, prominent institutions, key authors and journals, author clusters and collaborations, and keyword co-occurrence network analysis.

The top 100 highly cited papers from the WoSCC database were published between 1999 and 2021, with total citations ranging from 91 to 1580. The most cited article was published in IEEE Transactions on Medical Imaging. The top three most prolific countries/regions were the United States, mainland China, and the United Kingdom. The most published institutions and journals were Idaho University and IEEE Transactions on Medical Imaging. Twelve authors published more than four papers, with Suri, JS being the most productive author. The most studied topics were “ultrasound”, “computer-aided diagnosis”, and “segmentation”. Ultrasonography of Superficial Organs was the main site that was studied the most.

This study provides comprehensive insights into the characteristics of AI in medical ultrasound through quantitative and visualized analysis of the most highly cited literature. It serves as a valuable reference for the development and applications of AI, fostering potential collaborations within this domain.

This study aimed to verify whether the Liver-to-portal Ratio (LPR) can assess the adequacy of the Hepatobiliary Phase (HBP) for patients with different liver functions.

A total of 125 patients were prospectively enrolled in the study and graded into the non-cirrhosis group (45), Child-Pugh A group (40), and Child-Pugh B/C group (40). The LPR on HBP was calculated after eight HBPs were obtained within 5-40 minutes. The adequate HBP was determined according to the 2016 European Society of Gastrointestinal and Abdominal Radiology (ESGAR) consensus statement. The differences in LPR and lesions’ conspicuity between 10-min HBP and adequate HBP were analyzed by paired t-test and Wilcoxon signed-rank test, respectively. The chi-square test was used to test the difference in proportion with LPR larger than 1.462 between 10-min HBP and adequate HBP.

The differences in LPR and lesions’ conspicuity between 10-min HBP and adequate HBP were significant in Child-Pugh A and Child-Pugh B/C groups (P < 0.05), except for the non-cirrhosis group (P > 0.05). The differences in proportion with LPR larger than 1.462 between 10-min HBP and adequate HBP were not statistically significant in all groups (all P > 0.05).

The adequate HBP obtained according to the 2016 ESGAR consensus statement could provide larger LPR and better lesions’ conspicuity than 10-min HBP, especially for cirrhotic patients; however, the efficacy of using an LPR cutoff of 1.462 as the standard of the adequate HBP may be compromised in patients with cirrhosis.

Deep neural networks (DNNs) have made significant contributions to diagnosing pneumonia from chest X-ray imaging. However, certain aspects of diagnosis and planning can be further enhanced through the implementation of a Quantum Deep Neural Network (QDNN). Therefore, we introduced a technique that integrates neural networks with quantum algorithms named the ZFNet-quantum neural network for detecting pneumonia using 5863 X-ray scans with binary cases.

The hybrid model efficiently pre-processes complex and high-dimensional data by extracting significant features from the ZFNet model. These significant features are given to the quantum circuit algorithm and further embedded into a quantum device. The parameterized quantum circuit algorithm using qubits, superposition theorem, and entanglement phenomena generates 4 features from 4098 features extracted from images via a deep transfer learning model. Moreover, to validate the outcome measures of the proposed technique, we used various PennyLane quantum devices to detect pneumonia and normal control images. By using the Adam optimizer, which exploits an adaptive learning rate that is fixed to 10⁻⁶ and six layers of a quantum circuit composed of quantum gates, the proposed model achieves an accuracy of 96.5%, corresponding to 25 epochs.

The integrated ZFNet-quantum learning network outperforms the deep transfer learning network in terms of testing accuracy, as the accuracy gained by the Convolutional Neural Network (CNN) is 94%. Therefore, we use a hybrid classical-quantum model to detect pneumonia in which a variational quantum algorithm enhances the outcomes of a ZFNet transfer learning method.

This approach is an efficient and automated method for detecting pneumonia and could significantly enhance outcome measures related to the speed and accuracy of the network in the clinical and healthcare sectors.

Challenging HR conditions, such as elevated Heart Rate (HR) and Heart Rate Variability (HRV), are major contributors to motion artifacts in Coronary Computed Tomography Angiography (CCTA). This study aims to assess the impact of a deep learning-based motion correction algorithm (MCA) on motion artifacts in patients with challenging HR conditions, focusing on image quality and diagnostic performance of CCTA.

This retrospective study included 240 patients (mean HR: 88.1 ± 14.5 bpm; mean HRV: 32.6 ± 45.5 bpm) who underwent CCTA between June, 2020 and December, 2020. CCTA images were reconstructed with and without the MCA. The signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) were measured to assess objective image quality. Subjective image quality was evaluated by two radiologists using a 5-point scale regarding vessel visualization, diagnostic confidence, and overall image quality. Moreover, all vessels with scores ≥ 3 were considered clinically interpretable. The diagnostic performance of CCTA with and without MCA for detecting significant stenosis (≥ 50%) was assessed in 34 patients at both per-vessel and per-patient levels, using invasive coronary angiography as the reference standard.

The MCA significantly improved subjective image quality, increasing the vessel interpretability from 89.9% (CI: 0.88-0.92) to 98.8% (CI: 0.98-0.99) (p < 0.001). The use of MCA resulted in significantly higher diagnostic performance in both patient-based (AUC: 0.83 vs. 0.58, p = 0.04) and vessel-based (AUC: 0.92 vs. 0.81, p < 0.001) analyses, with the vessel-based accuracy notably increased from 79.4% (CI: 0.72-0.86) to 91.2% (CI: 0.85-0.95) (p = 0.01). There were no significant differences in objective image quality between the two reconstructions. The mean effective dose in this study was 2.8 ± 1.1 mSv.

The use of MCA allows for obtaining high-quality CCTA images and superior diagnostic performance with low radiation exposure in patients with elevated HR and HRV.