Medical Imaging Studies Research Articles

LUMBAR SPINAL STENOSIS: UNDERSTANDING SYMPTOMS, CAUSES, DIAGNOSIS, AND TREATMENT OPTIONS

Lumbar spinal stenosis is a degenerative condition characterized by the narrowing of the spinal canal in the lower back, leading to compression of the spinal cord and nerves. It primarily affects older individuals and is associated with age-related changes in the spine, such as osteoarthritis and disc degeneration. The condition presents with symptoms such as neurogenic claudication, radiculopathy, and loss of balance. Diagnosis involves medical history, physical examination, and imaging studies. Treatment options include non-surgical approaches like medication, physical therapy, and injections, as well as surgical interventions when conservative measures fail. A multidisciplinary approach is essential for effective management of lumbar spinal stenosis, improving symptoms and enhancing the quality of life for affected individuals.

Deformable medical image registration with global-local transformation network and region similarity constraint.

Deformable medical image registration can achieve fast and accurate alignment between two images, enabling medical professionals to analyze images of different subjects in a unified anatomical space. As such, it plays an important role in many medical image studies. Current deep learning (DL)-based approaches for image registration directly learn spatial transformation from one image to another, relying on a convolutional neural network and ground truth or similarity metrics. However, these methods only use a global similarity energy function to evaluate the similarity of a pair of images, which ignores the similarity of regions of interest (ROIs) within the images. This can limit the accuracy of the image registration and affect the analysis of specific ROIs. Additionally, DL-based methods often estimate global spatial transformations of images directly, without considering local spatial transformations of ROIs within the images. To address this issue, we propose a novel global-local transformation network with a region similarity constraint that maximizes the similarity of ROIs within the images and estimates both global and local spatial transformations simultaneously. Experiments conducted on four public 3D MRI datasets demonstrate that the proposed method achieves the highest registration performance in terms of accuracy and generalization compared to other state-of-the-art methods.

Do physicians remember cases? Implications for longitudinal designs in medical research and competency assessment

Background In pathology and other specialties of diagnostic medicine, longitudinal studies and competency assessments often involve physicians interpreting the same images multiple times. In these designs, a washout period is used to reduce the chances that later interpretations are influenced by prior exposure. Objective/s The present study examines whether a washout period between 9 and 39 months is sufficient to prevent three effects of prior exposure when pathologists review digital breast tissue biopsies and render diagnostic decisions: faster case review durations, higher confidence, and lower perceived difficulty. Methods In a longitudinal breast pathology study, 48 resident pathologists reviewed a mix of five novel and five repeated digital whole slide images during Phase 2, occurring 9–39 months after an initial Phase 1 review. Importantly, cases that were repeated for some participants in Phase 2 were novel for other participants in Phase 2. We statistically tested for differences in participants’ case review duration, self-reported confidence, and self-reported difficulty in Phase 2 based on whether the case was novel or repeated. Results No statistically significant difference in review time, confidence, or difficulty as a function of whether the case was repeated or novel in a Phase 2 review occurring 9-39 months after initial viewing; this same result was found in a subset of participants with a shorter (9-14-months) washout. Conclusion These results provide evidence to support the efficacy of at least a 9-months washout period in the design of longitudinal medical imaging and informatics studies to ensure no detectable effect of initial exposure on participant’s subsequent case review.

Diagnosis of pituitary abscess and treatment via transsphenoidal surgery: experience from 15 cases

ObjectivePituitary abscess is an often misdiagnosed, rare clinical disorder. To improve diagnostic accuracy and the efficacy of surgical and antibiotic therapy for patients with pituitary abscess, herein, we retrospectively reviewed 15 patients who presented with pituitary abscesses from 2005 to 2022. DesignRetrospective study. PatientsFifteen patients underwent transsphenoidal surgery and received antibiotic treatment. MeasurementsComplete details regarding medical history, clinical manifestations, laboratory examinations, imaging studies, and treatment strategies were obtained for all patients. ResultsMost patients presented with hypopituitarism and headaches, while some presented with fever, visual disturbances, and diabetes insipidus (DI). Abscesses showed significant annular enhancement post gadolinium injection. In most patients, pituitary abscess can be cured via microscopic or endoscopic drainage of the abscess followed by antibiotic treatment. Complete cure of pituitary abscess was observed in nine patients, with six cases of prolonged hypopituitarism and only one case of recurrence. Long-term hormone replacement therapy was effective in the postoperative management of hypopituitarism. ConclusionsThe typical manifestations of pituitary abscess include hypopituitarism and headaches; the presence of an enhanced ring at the edge of the mass on contrast-enhanced magnetic resonance images (MRI) is highly suggestive of pituitary abscess. We recommend antibiotic treatment for 4–6 weeks postoperatively, based on the results of bacterial cultures or metagenomic next-generation sequencing (mNGS).

Utilizing Triage Data for Medical Imaging Studies in the Emergency Department.

The use of radiological images is widespread in the emergency department (ED) as physicians commonly rely on them during initial evaluations to confirm diagnoses, contributing to prolonged waiting times. This study aimed to determine the relationship between commonly gathered triage data and the need for radiological imaging. Data were collected from electronic charts that contained routinely collected hospital data at the time of triage in the King Abdulaziz Medical City (KAMC) in Riyadh ED. The binary logistic regression results demonstrated a statistically significant relationship between age and radiological imaging ordered in the ED. Each one-unit increase in age corresponded to a 0.983-fold increase in the likelihood of ordering radiological imaging (odds ratio: 0.983, 95% confidence interval: 0.972-0.995, p = 0.004). In contrast, hypertension, diabetes, and heart failure were independent predictors of the need for radiological imaging in the ED (p >0.05). Patient data that are immediately available during ED triage can be used to predict the need for radiological imaging during ED visits. Such models can identify patients who may require radiological imaging during ED visits and expedite patient disposition.

Machine learning and deep learning techniques for breast cancer diagnosis and classification: a comprehensive review of medical imaging studies.

Breast cancer is a major public health concern, and early diagnosis and classification are critical for effective treatment. Machine learning and deep learning techniques have shown great promise in the classification and diagnosis of breast cancer. In this review, we examine studies that have used these techniques for breast cancer classification and diagnosis, focusing on five groups of medical images: mammography, ultrasound, MRI, histology, and thermography. We discuss the use of five popular machine learning techniques, including Nearest Neighbor, SVM, Naive Bayesian Network, DT, and ANN, as well as deep learning architectures and convolutional neural networks. Our review finds that machine learning and deep learning techniques have achieved high accuracy rates in breast cancer classification and diagnosis across various medical imaging modalities. Furthermore, these techniques have the potential to improve clinical decision-making and ultimately lead to better patient outcomes.

Abdominal imaging associates body composition with COVID-19 severity.

The main drivers of COVID-19 disease severity and the impact of COVID-19 on long-term health after recovery are yet to be fully understood. Medical imaging studies investigating COVID-19 to date have mostly been limited to small datasets and post-hoc analyses of severe cases. The UK Biobank recruited recovered SARS-CoV-2 positive individuals (n = 967) and matched controls (n = 913) who were extensively imaged prior to the pandemic and underwent follow-up scanning. In this study, we investigated longitudinal changes in body composition, as well as the associations of pre-pandemic image-derived phenotypes with COVID-19 severity. Our longitudinal analysis, in a population of mostly mild cases, associated a decrease in lung volume with SARS-CoV-2 positivity. We also observed that increased visceral adipose tissue and liver fat, and reduced muscle volume, prior to COVID-19, were associated with COVID-19 disease severity. Finally, we trained a machine classifier with demographic, anthropometric and imaging traits, and showed that visceral fat, liver fat and muscle volume have prognostic value for COVID-19 disease severity beyond the standard demographic and anthropometric measurements. This combination of image-derived phenotypes from abdominal MRI scans and ensemble learning to predict risk may have future clinical utility in identifying populations at-risk for a severe COVID-19 outcome.

Ultrasound for Breast Cancer Screening in Resource-Limited Settings: Current Practice and Future Directions.

Breast cancer (BC) is the most prevalent cancer among women globally. Cancer screening can reduce mortality and improve women's health. In developed countries, mammography (MAM) has been primarily utilized for population-based BC screening for several decades. However, it is usually unavailable in low-resource settings due to the lack of equipment, personnel, and time necessary to conduct and interpret the examinations. Ultrasound (US) with high detection sensitivity for women of younger ages and with dense breasts has become a supplement to MAM for breast examination. Some guidelines suggest using US as the primary screening tool in certain settings where MAM is unavailable and infeasible, but global recommendations have not yet reached a unanimous consensus. With the development of smart devices and artificial intelligence (AI) in medical imaging, clinical applications and preclinical studies have shown the potential of US combined with AI in BC screening. Nevertheless, there are few comprehensive reviews focused on the role of US in screening BC in underserved conditions, especially in technological, economical, and global perspectives. This work presents the benefits, limitations, advances, and future directions of BC screening with technology-assisted and resource-appropriate strategies, which may be helpful to implement screening initiatives in resource-limited countries.

A Review on Diabetic Retinopathy Detection Methods

Diabetes mellitus frequently results in diabetic retinopathy (DR), which results in lesions on the eye that impair vision. Blindness may result if it is not caught in time. Unfortunately, there is no cure for DR; treatment only preserves eyesight. Early diagnosis and treatment of DR can greatly lower the chance of vision loss. In contrast to computer-aided diagnosis methods, the manual diagnosis of DR retinal fundus images by ophthalmologists is time-, effort-, and cost-consuming as well as prone to error. Recently, deep learning has become one of the most common techniques that has improved performance in many areas, especially medical image analysis and classification. Convolutional neural networks are more frequently employed as a deep learning technique in the study of medical images, and they are very successful. In this paper, state-of-the-art techniques for deep learning-based detection and classification of DR color fundus images have been reviewed.

Protocol for Multi-modality MEdical imaging sTudy bAsed on KaiLuan Study (META-KLS): rationale, design and database building

IntroductionMulti-modality medical imaging study, especially brain MRI, greatly facilitates the research on subclinical brain disease. However, there is still a lack of such studies with a wider age span of...

Detection of abnormal extraocular muscles in small datasets of computed tomography images using a three-dimensional variational autoencoder

We sought to establish an unsupervised algorithm with a three–dimensional (3D) variational autoencoder model (VAE) for the detection of abnormal extraocular muscles in small datasets of orbital computed tomography (CT) images. 334 CT images of normal orbits and 96 of abnormal orbits diagnosed as thyroid eye disease were used for training and validation; 24 normal and 11 abnormal orbits were used for the test. A 3D VAE was developed and trained. All images were preprocessed to emphasize extraocular muscles and to suppress background noise (e.g., high signal intensity from bones). The optimal cut-off value was identified through receiver operating characteristic (ROC) curve analysis. The ability of the model to detect muscles of abnormal size was assessed by visualization. The model achieved a sensitivity of 79.2%, specificity of 72.7%, accuracy of 77.1%, F1-score of 0.667, and AUROC of 0.801. Abnormal CT images correctly identified by the model showed differences in the reconstruction of extraocular muscles. The proposed model showed potential to detect abnormalities in extraocular muscles using a small dataset, similar to the diagnostic approach used by physicians. Unsupervised learning could serve as an alternative detection method for medical imaging studies in which annotation is difficult or impossible to perform.

The usefulness of laparoscopic surgery for inguinal endometriosis

BackgroundInguinal endometriosis is a rare clinical disease with an unclear etiology and pathogenesis, and its diagnosis requires accurate medical history-taking and histological examination. However, surgical treatment for the condition has not yet been standardized. This report presents two cases of inguinal endometriosis.Case presentationThe first patient was a 36-year-old woman who complained of pain and swelling in her right inguinal region. Physical examination revealed a soft, tender right inguinal mass. The size of the mass repeatedly increased and decreased during menstruation and did not show swelling with abdominal pressure. Magnetic resonance imaging showed a 3.5 × 2.5 cm mass with high intensity on T2-weighted imaging in the right inguinal canal, and no communication was found between the lesion site and the abdominal cavity. We diagnosed this case as inguinal endometriosis and managed it using an anterior approach and laparoscopic observation. The second patient was a 51-year-old woman who presented with an intermittently painful mass in her right inguinal region. The mass tended to increase in size, with worsening pain before menstruation. Abdominal computed tomography revealed a 2 × 2 cm cystic mass in the right inguinal region. We made a diagnosis of inguinal ectopic endometriosis and decided to operate via the totally extraperitoneal (TEP) method for excision plus transabdominal observation. The postoperative course in both cases was uneventful with no recurrence.ConclusionsInguinal endometriosis is a rare entity that should be suspected in patients with cyclical symptoms of inguinal pain and swelling that correlate with their menstrual cycle, which might otherwise be attributed to inguinal hernia. It is crucial to make a preoperative diagnosis based on a careful medical review, physical examination, and imaging studies, and to make an appropriate surgical plan. Particularly, in the case of ectopic inguinal endometriosis involving the canal of Nuck, laparoscopic observation is useful for the intraoperative diagnosis of inguinal endometriosis to help rule out the involvement of other abdominal sites. However, it is important to select and modify the surgical technique to avoid rupturing the endometrisis mass and prevent postoperative recurrence.

Estimación forense de la edad basada en el estudio de la epífisis proximal de la clavícula. Actualización y factores de interferencia

Forensic Age Estimation in migrant minors without family reference is a developing field within forensic sciences. The international and national action guides recommend for this expert reports the practice of multiplanar medical imaging studies, more precisely computed tomography, of the proximal end of the clavicle. This type of test is reserved for those cases in which a possible majority of the supposedly minor cannot be ruled out by other methods. This diagnostic test has peculiarities derived from its relative complexity of interpretation and the relatively high frequency of appearance of developmental anomalies that make it difficult to apply it as a test of expert value. This review aims to be an update of the state of the art in this field of expertise with a special focus on its expert value and the impact generated by the aforementioned anomalies in its practical application.

РАЗРАБОТКА КОМПЬЮТЕРНЫХ АЛГОРИТМОВ МОДЕЛИРОВАНИЯ СИНТЕТИЧЕСКИХ ИЗОБРАЖЕНИЙ НА ОСНОВЕ ГЛУБОКИХ ГЕНЕРАТИВНЫХ МОДЕЛЕЙ

Deep neural networks have made significant breakthroughs in the field of medical image analysis. However, due to their high data requirements, small datasets in medical imaging tasks can hinder their capabilities. Synthetic data generation is a promising alternative to augment training datasets and enable medical imaging studies on a larger scale. Recently, deep generative models have attracted the attention of the computer vision community because they allow the generation of photorealistic synthetic images. In this paper, we investigate the potential use of deep generative models and develop computer algorithms to generate highresolution synthetic images. The obtained results of computer modelling confirmed the high efficiency and advantages of diffusion models in image synthesis problems.

Early detection of gastric cancer via high-resolution terahertz imaging system.

Terahertz (THz) wave has demonstrated a good prospect in recent years, but the resolution is still one of the problems that restrict the application of THz technology in medical imaging. Paraffin-embedded samples are mostly used in THz medical imaging studies, which are thicker and significantly different from the current gold standard slice pathological examination in sample preparation. In addition, THz absorption in different layers of normal and cancerous tissues also remains to be further explored. In this study, we constructed a high-resolution THz imaging system to scan non-tumorous adjacent tissue slices and gastric cancer (GC) tissue slices. In this system, a THz quantum cascade laser emitted a pulsed 3THz signal and the transmitted THz wave was received by a THz detector implemented in a 65nm CMOS process. The slice thickness was only 20μm, which was close to that of the medical pathology examination. We successfully found THz transmittance differences between different layers of normal gastric tissues based on THz images, and the resolution could reach 60μm for the first time. The results indicated that submucosa had a lower THz transmittance than that of mucosa and muscular layer in non-tumorous adjacent tissue. However, in GC tissue, THz transmittance of mucosa and submucosa was similar, caused by the decreased transmittance of mucosa, where the cancer occurs. Therefore, we suppose that the similar terahertz transmittance between gastric mucosa and submucosa may indicate the appearance of cancerization. The images obtained from our THz imaging system were clearer than those observed with naked eyes, and can be directly compared with microscopic images. This is the first application of THz imaging technology to identify non-tumorous adjacent tissue and GC tissue based on the difference in THz wave absorption between different layers in the tissue. Our present work not only demonstrated the potential of THz imaging to promote early diagnosis of GC, but also suggested a new direction for the identification of normal and cancerous tissues by analyzing differences in THz transmittance between different layers of tissue.

Multi-Task Learning with High-Dimensional Noisy Images

Recent medical imaging studies have given rise to distinct but inter-related datasets corresponding to multiple experimental tasks or longitudinal visits. Standard scalar-on-image regression models that fit each dataset separately are not equipped to leverage information across inter-related images, and existing multi-task learning approaches are compromised by the inability to account for the noise that is often observed in images. We propose a novel joint scalar-on-image regression framework involving wavelet-based image representations with grouped penalties that are designed to pool information across inter-related images for joint learning, and which explicitly accounts for noise in high-dimensional images via a projection-based approach. In the presence of nonconvexity arising due to noisy images, we derive nonasymptotic error bounds under nonconvex as well as convex grouped penalties, even when the number of voxels increases exponentially with sample size. A projected gradient descent algorithm is used for computation, which is shown to approximate the optimal solution via well-defined nonasymptotic optimization error bounds under noisy images. Extensive simulations and application to a motivating longitudinal Alzheimer’s disease study illustrate significantly improved predictive ability and greater power to detect true signals, that are simply missed by existing methods without noise correction due to the attenuation to null phenomenon. Supplementary materials for this article are available online.

Abstract 13419: Racial Differences in Patients With Spontaneous Coronary Artery Dissection: A Report of the ISCAD Registry

Introduction: Spontaneous coronary artery dissection (SCAD) is a nonatherosclerotic cause of myocardial infarction. The demographics and clinical differences based on race in patients with SCAD are not well understood. Methods: Demographics and clinical variables were obtained from the prospective, multicenter iSCAD Registry. Characteristics of study participants (including demographics, past medical history, timeline, imaging studies, follow-up visits, SCAD presentation, and treatment) were compared between white and non-white patients. Race and ethnicity was self-reported by a patient questionnaire. Results: A total of 505 patients were included in the analysis. Among them, 427 (84.6%) were characterized as white and 78 (15.4%) were characterized as non-white. White patients were 49.5 years and non-white patients were 46.8 years at the time of the initial SCAD presentation (p=0.0435). White patients were more likely to be married (77% vs 51%, p<0.0001) and were more likely to have a college degree (42 vs 28%, p<0.0001). White patients were more likely to present to the hospital in less than 24 hours (85% vs 75%, p=0.07) and have an angiogram in less than 24 hours (78% vs 58%, p=0.004). White patients were more likely to present with STEMI (30% vs 11%, p=0.004) and were more likely to receive revascularization (25% vs 11%, p=0.0147). A multivariable analysis explored the variables associated with angiography greater than 24 hours. STEMI vs NSTEMI [OR 0.090 (0.035 to 0.231), p<0.0001] and white vs non-white [OR 0.376 (0.188 to 0.755), p=0.0059] remained significant. White patients were more likely to have fibromuscular dysplasia (FMD), (43% vs 30%, p=0.0336). There was no significant difference in the number of non-coronary vascular imaging studies between the two groups. Conclusions: The number of non-white patients enrolled in the iSCAD registry is low. White patients with SCAD were more likely to present with STEMI and were more likely to have a diagnosis of FMD. Further research is needed to study associations between race and differences in presentation, comorbidities, and treatment in patients with SCAD.

Identifying Calcaneal Anatomical Regions of Interest (ROI) for Quantitative Ultrasound Application in Subadults

Whilst quantitative ultrasound can be reliably used to assess bone health in adults, the fixed location of the transducers in current devices may result in inaccurate and unreliable measurements in bone assessment in children due to the variation in foot size during growth. To improve positioning for paediatric assessment, Jaworski et al. (1995) created an anatomical method to identify the region of interest (ROI), however, there have been no medical imaging studies to confirm that the Jaworski method results in consistent placement of the transducer on the centre of the calcaneal body to avoid edge artefacts. In this study, computed tomography scans of the tarsus were collected from 498 individuals (258 females; 240 males) aged 2 to 20 years and used to create three novel anatomical methods to identify ROI on the calcaneus using palpable landmarks. In addition, the established Jaworski method was applied to the same scans and compared to our novel methods. The maximum ROI significantly increased with age with males having significantly greater diameters, supporting the recommendation that ½ inch diameter transducers should be used on individuals younger than 7 years of age. We identified that 79% of the ‘Jaworski points’ lied anterosuperior to the ROI centre point identified in this study, with 10% of the points lying outside the ROI. Of the three novel methods, only the calcaneal insertion method demonstrated small measurement variance between individuals of the same age in each sex and is therefore the preferred method for ultrasound clinical application.

Creation of a Multi-vendor Image-Sharing Solution.

Transferring medical imaging studies from one institution to another is a common occurrence in today's medical practice. For the past two decades, radiology departments have relied on physical media (compact discs and digital video disks) and human couriers to accomplish image transfer. This mode of transfer is slow, prone to failure, and reliant on outdated technology. To address these shortcomings, multiple image-sharing vendors have created electronic, cloud-based solutions. While these solutions solve multiple problems, a new problem has been introduced: it is difficult to send or receive images across image-sharing platforms. In this work, we describe how we have developed a solution to share images across multiple vendor platforms.

Automated analysis of low-field brain MRI in cerebral malaria.

A central challenge of medical imaging studies is to extract biomarkers that characterize disease pathology or outcomes. Modern automated approaches have found tremendous success in high-resolution, high-quality magnetic resonance images. These methods, however, may not translate to low-resolution images acquired on magnetic resonance imaging (MRI) scanners with lower magnetic field strength. In low-resource settings where low-field scanners are more common and there is a shortage of radiologists to manually interpret MRI scans, it is critical to develop automated methods that can augment or replace manual interpretation, while accommodating reduced image quality. We present a fully automated framework for translating radiological diagnostic criteria into image-based biomarkers, inspired by a project in which children with cerebral malaria (CM) were imaged using low-field 0.35 Tesla MRI. We integrate multiatlas label fusion, which leverages high-resolution images from another sample as prior spatial information, with parametric Gaussian hidden Markov models based on image intensities, to create a robust method for determining ventricular cerebrospinal fluid volume. We also propose normalized image intensity and texture measurements to determine the loss of gray-to-white matter tissue differentiation and sulcal effacement. These integrated biomarkers have excellent classification performance for determining severe brain swelling due toCM.