Evaluation of Bony Femoral Morphological Parameters in Anterior Cruciate Ligament Injury Using Magnetic Resonance Imaging: A Retrospective Unmatched Case-Control Study.

Background Over time, there has been a noticeable increase in anterior cruciate ligament (ACL) injuries. The current imperative is to anticipate predisposing factors and proactively prevent ACL injuries. The occurrence of ACL injuries has been linked to diverse factors associated with the morphology of the distal femur. Objectives Through this study, we aim to compare the anatomic variables of distal femur morphology such as notch width (NW), bicondylar width (BW), notch entrance width (NEW), and notch width index (NWI) between patients with ACL injuries and non-injured patients using MRI. We also aim to make a comparison of these factors between male and female genders to assess the gender variability. Material and methods A retrospective case-control study was conducted amongst patients who underwent MRI Knee scan for clinical suspicion of internal derangement during the study period. We selected the first 125 individuals who were found to have ACL injury in the MRI scans and selected another 125 individuals who had an intact ACL in the scans, to serve as controls in the study. Demographic information was retrieved from the hospital's electronic records, and the assessment of NW, NWI, BW, and NEW was conducted through a review of MRI sequences. They were then compared between the cases and control groups, as well as between male and female genders. Results The ACL-injured group exhibited statistically significant reductions in NW and NWI. While 17.39 mm was the mean NW among cases, 17.86 was the mean value among controls. Similarly, the mean NWI was 0.25 among patients with ACL injuries and 0.27 among controls. Gender-based comparisons also revealed statistically significant differences in NW and NWI measurements, where females were reported to have comparatively lower measurements. The mean NW for males and females in the injured group were 18.26 mm and 15.40 mm, respectively, while it was 18.71 mm and 16.90 mm, respectively, in the control group. In the case of NEW, males in the injured group had a slightly higher value (21.33 mm) than the controls (20.65). Females on the other hand exhibited a lower mean value of NEW in ACL-injured group (18.51 mm) in comparison to the non-injured (18.79 mm). BW did not seem to show a significant difference between the two groups. Conclusions In the studied population, ACL injuries demonstrated a higher occurrence in individuals with a narrow femoral intercondylar NWI. If any of these characteristics are identified in an MRI, it may be helpful to identify individuals who are at a higher risk of developing ACL injuries and may thereby help in planning preventative strategies.

Validation of a Deep Learning Model for Detecting Chest Pathologies from Digital Chest Radiographs.

Purpose: Manual interpretation of chest radiographs is a challenging task and is prone to errors. An automated system capable of categorizing chest radiographs based on the pathologies identified could aid in the timely and efficient diagnosis of chest pathologies. Method: For this retrospective study, 4476 chest radiographs were collected between January and April 2021 from two tertiary care hospitals. Three expert radiologists established the ground truth, and all radiographs were analyzed using a deep-learning AI model to detect suspicious ROIs in the lungs, pleura, and cardiac regions. Three test readers (different from the radiologists who established the ground truth) independently reviewed all radiographs in two sessions (unaided and AI-aided mode) with a washout period of one month. Results: The model demonstrated an aggregate AUROC of 91.2% and a sensitivity of 88.4% in detecting suspicious ROIs in the lungs, pleura, and cardiac regions. These results outperform unaided human readers, who achieved an aggregate AUROC of 84.2% and sensitivity of 74.5% for the same task. When using AI, the aided readers obtained an aggregate AUROC of 87.9% and a sensitivity of 85.1%. The average time taken by the test readers to read a chest radiograph decreased by 21% (p < 0.01) when using AI. Conclusion: The model outperformed all three human readers and demonstrated high AUROC and sensitivity across two independent datasets. When compared to unaided interpretations, AI-aided interpretations were associated with significant improvements in reader performance and chest radiograph interpretation time.

Levels of Autonomous Radiology.

Radiology, being one of the younger disciplines of medicine with a history of just over a century, has witnessed tremendous technological advancements and has revolutionized the way we practice medicine today. In the last few decades, medical imaging modalities have generated seismic amounts of medical data. The development and adoption of artificial intelligence applications using this data will lead to the next phase of evolution in radiology. It will include automating laborious manual tasks such as annotations, report generation, etc, along with the initial radiological assessment of patients and imaging features to aid radiologists in their diagnostic and treatment planning workflow. We propose a level-wise classification for the progression of automation in radiology, explaining artificial intelligence assistance at each level with the corresponding challenges and solutions. We hope that such discussions can help us address challenges in a structured way and take the necessary steps to ensure the smooth adoption of new technologies in radiology.

A deep learning approach for automated diagnosis of pulmonary embolism on computed tomographic pulmonary angiography.

BACKGROUND: Computed tomographic pulmonary angiography (CTPA) is the diagnostic standard for confirming pulmonary embolism (PE). Since PE is a life-threatening condition, early diagnosis and treatment are critical to avoid PE-associated morbidity and mortality. However, PE remains subject to misdiagnosis. METHODS: We retrospectively identified 251 CTPAs performed at a tertiary care hospital between January 2018 to January 2021. The scans were classified as positive (n = 55) and negative (n = 196) for PE based on the annotations made by board-certified radiologists. A fully anonymized CT slice served as input for the detection of PE by the 2D segmentation model comprising U-Net architecture with Xception encoder. The diagnostic performance of the model was calculated at both the scan and the slice levels. RESULTS: The model correctly identified 44 out of 55 scans as positive for PE and 146 out of 196 scans as negative for PE with a sensitivity of 0.80 [95% CI 0.68, 0.89], a specificity of 0.74 [95% CI 0.68, 0.80], and an accuracy of 0.76 [95% CI 0.70, 0.81]. On slice level, 4817 out of 5183 slices were marked as positive for the presence of emboli with a specificity of 0.89 [95% CI 0.88, 0.89], a sensitivity of 0.93 [95% CI 0.92, 0.94], and an accuracy of 0.89 [95% CI 0.887, 0.890]. The model also achieved an AUROC of 0.85 [0.78, 0.90] and 0.94 [0.936, 0.941] at scan level and slice level, respectively for the detection of PE. CONCLUSION: The development of an AI model and its use for the identification of pulmonary embolism will support healthcare workers by reducing the rate of missed findings and minimizing the time required to screen the scans.

Utility of selected sequence MRI imaging of the axial skeleton in the diagnosis of axial spondyloarthritis.

Background: Whole body MRI has been used to evaluate inflammatory lesions associated with axial spondyloarthritis (SpA). These sequences are extensive, time consuming and add to the cost of the investigation. We aimed to determine the utility of selected sequence MRI imaging of the axial skeleton including spine, pelvis and sacroiliac (SI) joints to identify features of (SpA). Methods: A retrospective study was conducted on 76 patients diagnosed with SpA that underwent a selective sequence MRI imaging of the axial skeleton. The MRI were reported by two musculoskeletal trained radiologists were reviewed. The MRI sequences included whole spine sequences of sagittal STIR (short tau inversion recovery), T1 weighted and T2 weighted sequences. Coronal STIR and T1 weighted images were studied for SI joints and pelvis. The MRI were assessed based on the guidelines outlined by the Assessment of SpondyloArthritis International Society (ASAS) for features of spondylitis, spondylodiscitis, enthesitis, synovitis, capsulitis, bone marrow edema, fatty marrow replacement, erosions and bony ankylosis. Inflammatory lesions were documented in the spine, sacroiliac, facet, hip and costovertebral joints. Results: The mean scan duration was 28 min. SI joint involvement was noted in 74 (97.3%) of patients. The other most prevalent findings were spondylitis in 44 (57.8%) patients, costovertebral joint involvement in 31 (40.7%), facet joint lesions in 32 (42.1%), spondylodiscitis in 21 (27.6%), enthesitis in 13 (17.1%), hip lesions in 16 (21%) and ankylosis in 10 (13.1%). Conclusions: This selective sequence imaging of the pelvis and spine was able to identify typical lesions of SpA in a shorter time period. Fifty-five percent patients had lesions in the posterior elements including facet joints and costovertbral joints that would be missed on traditional SI joint imaging.

Observer performance evaluation of the feasibility of a deep learning model to detect cardiomegaly on chest radiographs.

Background: Cardiothoracic ratio (CTR) is the ratio of the diameter of the heart to the diameter of the thorax. An abnormal CTR (>0.55) is often an indicator of an underlying pathological condition. The accurate prediction of an abnormal CTR chest X-rays (CXRs) aids in the early diagnosis of clinical conditions. Purpose: We propose a deep learning (DL)-based model for automatic CTR calculation to assist radiologists with rapid diagnosis of cardiomegaly and thus optimise the radiology flow. Material and Methods: The study population included 1012 posteroanterior CXRs from a single institution. The Attention U-Net DL architecture was used for the automatic calculation of CTR. An observer performance test was conducted to assess the radiologist's performance in diagnosing cardiomegaly with and without artificial intelligence assistance. Results: U-Net model exhibited a sensitivity of 0.80 [95% CI: 0.75, 0.85], specificity >99%, precision of 0.99 [95% CI: 0.98, 1], and a F1 score of 0.88 [95% CI: 0.85, 0.91]. Furthermore, the sensitivity of the reviewing radiologist in identifying cardiomegaly increased from 40.50% to 88.4% when aided by the AI-generated CTR. Conclusion: Our segmentation-based AI model demonstrated high specificity (>99%) and sensitivity (80%) for CTR calculation. The performance of the radiologist on the observer performance test improved significantly with provision of AI assistance. A DL-based segmentation model for rapid quantification of CTR can therefore have significant potential to be used in clinical workflows by reducing radiologists' burden and alerting to an abnormal enlarged heart early on.

High resolution T2∗ mapping in assessment of knee articular cartilage on 3T MRI.

OBJECTIVE: To evaluate the spectrum of T2∗ values in healthy cartilage of young asymptomatic adults on high resolution 3T MRI. METHODS: A total of 50 asymptomatic adult volunteers with age ranging from 18 to 35 years were enrolled for the study with the purpose of assessing T2∗ values in healthy cartilage without any degenerative changes. The articular cartilage was assessed on two sections, one each through the medial and lateral compartments. The cartilage was segmented into 8 regions through the tibio-femoral and patella-femoral joints. Further post processing was done using multiple ROI placement to delineate ROI areas for calculation of full thickness and zonal (superficial and deep) T2∗ values. Thus, a total of 1200 ROI areas (50 volunteers, 8 segments, and 3 areas for each segment) were assessed. RESULTS: The results revealed a superior bulk T2∗ value of 29.2 ± 3.6 ms from the posterior medial femoral cartilage and 26.1 ± 3.1 ms from the patellar region. Intermediate values were obtained from posterior lateral femoral cartilage, central femoral cartilage, and trochlea. The tibial plateau cartilage had the lowest values - 19.6 ± 2.6 ms for the medial tibial plateau and 20.6 ± 2.8 ms for lateral tibial plateau. The study demonstrated substantial regional physiological variation existing in the T2∗ values across various regions of the knee joint, which could be attributed to varying amounts of shearing forces across the joint. No significant differences were noted in bulk T2∗ values between the two genders, with only the trochlear segment revealing significantly increased values in males (p = 0.007). All the cartilage segments revealed significantly increased T2∗ values in the superficial zone as compared to the deep zone. CONCLUSION: There is a significant regional difference in the bulk T2∗ values of articular cartilage in a normal physiological state across various joint segments. A zonal gradient with increasing values from the deep to the superficial zone also exists. These findings can prove invaluable in assessing changes in T2∗ values occurring in diseased/degenerative cartilage.

Evaluating the Association Between Comorbidities and COVID-19 Severity Scoring on Chest CT Examinations Between the Two Waves of COVID-19: An Imaging Study Using Artificial Intelligence.

Background Coronavirus disease 2019 (COVID-19) has accounted for over 352 million cases and five million deaths globally. Although it affects populations across all nations, developing or transitional, of all genders and ages, the extent of the specific involvement is not very well known. This study aimed to analyze and determine how different were the first and second waves of the COVID-19 pandemic by assessing computed tomography severity scores (CT-SS). Methodology This was a retrospective, cross-sectional, observational study performed at a tertiary care Institution. We included 301 patients who underwent CT of the chest between June and October 2020 and 1,001 patients who underwent CT of the chest between February and April 2021. All included patients were symptomatic and were confirmed to be COVID-19 positive. We compared the CT-SS between the two datasets. In addition, we analyzed the distribution of CT-SS concerning age, comorbidities, and gender, as well as their differences between the two waves of COVID-19. Analysis was performed using the SPSS version 22 (IBM Corp., Armonk, NY, USA). The artificial intelligence platform U-net architecture with Xception encoder was used in the analysis. Results The study data revealed that while the mean CT-SS did not differ statistically between the two waves of COVID-19, the age group most affected in the second wave was almost a decade younger. While overall the disease had a predilection toward affecting males, our findings showed that females were more afflicted in the second wave of COVID-19 compared to the first wave. In particular, the disease had an increased severity in cases with comorbidities such as hypertension, diabetes mellitus, bronchial asthma, and tuberculosis. Conclusions This assessment demonstrated no significant difference in radiological severity score between the two waves of COVID-19. The secondary objective revealed that the two waves showed demographical differences. Hence, we iterate that no demographical subset of the population should be considered low risk as the disease manifestation was heterogeneous.

Simplified Guide to MR Elastography in Early Detection of Hepatic Fibrosis with Case Reports: The New Norm in Assessing Liver Health.

The current unhealthy diets and sedentary lifestyle have led to increase in the prevalence of diabetes and metabolic syndrome globally. Fatty liver is a common occurrence in metabolic syndrome. The liver health is often ignored due to delayed warning signs. Fatty changes of the liver is one of the common findings in ultrasonography. Ultrasound does not detect fibrosis except when cirrhosis is developed. Early stages of fibrosis are asymptomatic with no significant laboratory or preliminary imaging findings. With fibrosis, the elasticity of the liver is reduced and becomes stiffer. Over the years, many techniques have developed to assess the stiffness of the liver, starting from palpation, ultrasonography, and recently developed magnetic resonance elastography (MRE). In this article, we have tried to simplify the concepts of MRE to detect fibrosis and present few case reports. The basic steps involved in generating elastograms and interpretation with some insight on how to incorporate it into the clinical workflow are discussed. MRE is superior to various other available techniques and even offers certain advantages over biopsy. MRE is FDA approved for liver fibrosis since 2009, yet it is hardly used in the Indian setting. MRE is a safe and noninvasive technique to evaluate a large volume of the liver and can be a new norm for the evaluation of fatty liver. Magnetic resonance imaging (MRI)-based elastography techniques hold an exciting future in providing mechanical properties of tissues in various organs like spleen, brain, kidney, and heart.

Real-world analysis of artificial intelligence in musculoskeletal trauma.

Musculoskeletal trauma accounts for a large percentage of emergency room visits and is amongst the top causes of unscheduled patient visits to the emergency room. Musculoskeletal trauma results in expenditure of billions of dollars and protracted losses of quality-adjusted life years. New and innovative methods are needed to minimise the impact by ensuring quick and accurate assessment. However, each of the currently utilised radiological procedures, such as radiography, ultrasonography, computed tomography, and magnetic resonance imaging, has resulted in implosion of medical imaging data. Deep learning, a recent advancement in artificial intelligence, has demonstrated the potential to analyse medical images with sensitivity and specificity at par with experts. In this review article, we intend to summarise and showcase the various developments which have occurred in the dynamic field of artificial intelligence and machine learning and how their applicability to different aspects of imaging in trauma can be explored to improvise our existing reporting systems and improvise on patient outcomes.

Key Technology Considerations in Developing and Deploying Machine Learning Models in Clinical Radiology Practice.

The use of machine learning to develop intelligent software tools for the interpretation of radiology images has gained widespread attention in recent years. The development, deployment, and eventual adoption of these models in clinical practice, however, remains fraught with challenges. In this paper, we propose a list of key considerations that machine learning researchers must recognize and address to make their models accurate, robust, and usable in practice. We discuss insufficient training data, decentralized data sets, high cost of annotations, ambiguous ground truth, imbalance in class representation, asymmetric misclassification costs, relevant performance metrics, generalization of models to unseen data sets, model decay, adversarial attacks, explainability, fairness and bias, and clinical validation. We describe each consideration and identify the techniques used to address it. Although these techniques have been discussed in prior research, by freshly examining them in the context of medical imaging and compiling them in the form of a laundry list, we hope to make them more accessible to researchers, software developers, radiologists, and other stakeholders.

Evaluation of Changes in Condylar Cartilage Thickness Using MRI and Ultrasound Imaging in Patients Treated by Mandibular Advancement With Myofunctional Appliance: An In-Vivo Pilot Study.

Introduction The changes occurring due to growth modulation of the condylar cartilage act as an important mechanism for mandibular advancement using myofunctional appliance therapy. So this study aims to evaluate the condylar cartilage thickness by using MRI and USG in patients undergoing myofunctional appliance therapy for mandibular advancement with the null hypothesis being that there are no changes seen in the thickness of condylar cartilage in growing children. Materials and methods A prospective evaluation of samples having skeletal Class-II malocclusion ranging between cervical vertebral maturation index (CVMI) stage II and III, requiring twin block functional therapy was performed. Ten patients were selected randomly who underwent MRI and USG in the open and close positions for the evaluation of condylar cartilage thickness and the dimensional changes in the width of the right and left condyle in mm at T0 and T1. Result There was no statistically significant difference between the values interpreted by MRI or USG imaging when compared at T0 and T1 and in the open and closed mouth on the left and right sides. At T0, the mean thickness noted was 0.49 mm and 0.48 mm during opening and closing on the left side and 0.52 mm in both positions on the right side. At T1, the mean thickness noted was 0.8 and 0.79mm during opening and closing on the left side, whereas it was 0.81 mm in both positions on the right side. Conclusion The condylar cartilage thickness increases significantly after twin block therapy suggestive of mandibular growth in skeletal class II malocclusion. It can be inferred that both MRI and USG carry equal diagnostic interpretation, as there was no statistically significant difference between the two imaging modalities.

Double trouble!!! An unusual presentation of cervical cord herniation and medial end clavicle non-union in a single patient.

Non-union of medial end clavicle is rare. Though traditionally they have been treated with conservative methods, surgery at initial presentation should be considered in these patients since conservative treatment can lead to non-union. Herniation of spinal cord, though rare, is seen in the thoracic region but can also occur in the cervical region post-traumatically as seen in our patient.

Parosteal Osteosarcoma of the Distal Radius Mimicking an Osteochondroma-A Diagnostic Misadventure.

We present a case of a parosteal osteosarcoma mimicking an osteochondroma with atypical clinical features, radiographic findings, and histological examination. This report serves to exemplify the importance of recognizing the similarities between these 2 entities and other peculiar features that will help to differentiate between sessile osteochondromas and parosteal osteosarcomas, to prevent misdiagnosis.

A peek into the future of radiology using big data applications.

Big data is extremely large amount of data which is available in the radiology department. Big data is identified by four Vs - Volume, Velocity, Variety, and Veracity. By applying different algorithmic tools and converting raw data to transformed data in such large datasets, there is a possibility of understanding and using radiology data for gaining new knowledge and insights. Big data analytics consists of 6Cs - Connection, Cloud, Cyber, Content, Community, and Customization. The global technological prowess and per-capita capacity to save digital information has roughly doubled every 40 months since the 1980's. By using big data, the planning and implementation of radiological procedures in radiology departments can be given a great boost. Potential applications of big data in the future are scheduling of scans, creating patient-specific personalized scanning protocols, radiologist decision support, emergency reporting, virtual quality assurance for the radiologist, etc. Targeted use of big data applications can be done for images by supporting the analytic process. Screening software tools designed on big data can be used to highlight a region of interest, such as subtle changes in parenchymal density, solitary pulmonary nodule, or focal hepatic lesions, by plotting its multidimensional anatomy. Following this, we can run more complex applications such as three-dimensional multi planar reconstructions (MPR), volumetric rendering (VR), and curved planar reconstruction, which consume higher system resources on targeted data subsets rather than querying the complete cross-sectional imaging dataset. This pre-emptive selection of dataset can substantially reduce the system requirements such as system memory, server load and provide prompt results. However, a word of caution, "big data should not become "dump data" due to inadequate and poor analysis and non-structured improperly stored data. In the near future, big data can ring in the era of personalized and individualized healthcare.

Data mining in radiology.

Data mining facilitates the study of radiology data in various dimensions. It converts large patient image and text datasets into useful information that helps in improving patient care and provides informative reports. Data mining technology analyzes data within the Radiology Information System and Hospital Information System using specialized software which assesses relationships and agreement in available information. By using similar data analysis tools, radiologists can make informed decisions and predict the future outcome of a particular imaging finding. Data, information and knowledge are the components of data mining. Classes, Clusters, Associations, Sequential patterns, Classification, Prediction and Decision tree are the various types of data mining. Data mining has the potential to make delivery of health care affordable and ensure that the best imaging practices are followed. It is a tool for academic research. Data mining is considered to be ethically neutral, however concerns regarding privacy and legality exists which need to be addressed to ensure success of data mining.

Cloud Computing for radiologists.

Cloud computing is a concept wherein a computer grid is created using the Internet with the sole purpose of utilizing shared resources such as computer software, hardware, on a pay-per-use model. Using Cloud computing, radiology users can efficiently manage multimodality imaging units by using the latest software and hardware without paying huge upfront costs. Cloud computing systems usually work on public, private, hybrid, or community models. Using the various components of a Cloud, such as applications, client, infrastructure, storage, services, and processing power, Cloud computing can help imaging units rapidly scale and descale operations and avoid huge spending on maintenance of costly applications and storage. Cloud computing allows flexibility in imaging. It sets free radiology from the confines of a hospital and creates a virtual mobile office. The downsides to Cloud computing involve security and privacy issues which need to be addressed to ensure the success of Cloud computing in the future.

Fine needle aspiration cytology diagnosis of paravertebral extraosseus Ewing's sarcoma.

Extraskeletal Ewing's sarcoma (EES) is a rare tumor. Paravertebral Ewing's sarcoma requires more extensive therapy as compared to Ewing's sarcoma of bone. Fine needle aspiration cytology (FNAC) plays an important role in the early diagnosis of these cases. We present a case where paravertebral extraosseous Ewing's sarcoma was diagnosed on FNAC in a 19-year-old girl.