ABSTRACT
Introduction: At the end of the year 2019 a novel virus named SARS-CoV-2 caused Coronavirus Disease 19 (COVID-19), manifesting as lung infection which can result in severe pneumonia. The gold standard for diagnosis of the virus is the detection of viral RNA through reverse transcriptase PCR (RT-PCR). Materials and Methods: This study is a retrospective analysis of diagnostic yield of chest imaging modalities in the diagnosis of COVID-19 in patients admitted to the Imam Khomeini hospital, Tehran, Iran during march 2020 until July 2020. Results: 204 hospitalized patients with a mean age of 58.5 years, diagnosed with COVID-19 were enrolled in this study, who had their disease confirmed by PCR. About 20% of patients had normal O2 Saturation (above 93%) and 80% had low O2 Saturation. Also 70% of patients were hospitalized to the ICU. Among investigated patients, 87.6% had abnormal findings in their CXR. Also, 97.1% of patients had abnormal CT-Scan. In this study, the sensitivity of the CXR in the diagnosis of COVID-19 was 87.5% (CI 95%, 83 to 91) and the sensitivity of the chest CT-Scan was 97.1% (CI 95%, 94.8 to 99). Discussion: Utilizing CXR as a first-line imaging modality is recommended in many countries and clinical settings and chest CT-Scan is mainly reserved for other additional roles. This study reveals a CXR sensitivity of 87.5% which is in accordance with the recent literature (69-90%).

ABSTRACT
Neglected ruptures of the patellar tendon are rare and severe injuries. Their diagnosis is strongly suspected based on clinical examination, but confirmation is provided through imaging, particularly magnetic resonance imaging (MRI), which is equally important for surgical planning. Here, we report a case of a patient with a neglected rupture of the patellar tendon to highlight the role of MRI in diagnosis. Surgical repair was performed using the semitendinosus tendon and wire framing.

ABSTRACT
The integration of Artificial Intelligence (AI) in radiology, especially for chest computed tomography (CT) scan analysis, marked a significant advancement in medical diagnostics, aiming to improve patient care and streamline the workflow for radiologists. This review article examined the role of current AI technologies, including machine learning (ML), deep learning (DL), convolutional neural networks (CNN), and radiomics, in enhancing the detection and characterisation of lung diseases. These technologies are instrumental in identifying complex patterns within imaging data and constructing more informed decisions regarding disease severity, progression, and potential treatment options. Deep learning and CNN have demonstrated effectiveness in analysing the intricate details present in chest CT scans, offering a high degree of accuracy. Radiomics complements these methods by extracting quantitative features from medical images, providing deeper insights into disease characteristics that are not visible through standard imaging techniques. The application of AI has shown promise in improving the diagnosis and management of interstitial lung diseases and lung cancers, contributing to the development of personalised treatment plans. However, this review also highlights limitations, such as small sample sizes in studies, which may impact the generalisability of AI applications in this field. Despite these challenges, the ongoing incorporation of AI into radiological practices is anticipated to significantly enhance the accuracy and efficiency of lung disease diagnostics, setting a foundation for future research and improvements in clinical practice.

ABSTRACT
The increasing use of computed tomography (CT) in injury management is a growing concern as ionizing radiation (IR) is associated with cancer-related risks especially in young persons, who are most affected by injuries. This study aimed to assess ionizing radiation exposure from head CT in patients with injuries and to propose CT dose age-specific diagnostic reference levels (DRLs). Data was extracted from two prospectively collected CT registries over a period of 5 years at two community-based university-affiliated hospitals from 2019 to 2024. The linear relationship between CT dose (dose-length product; DLP) and age was assessed using Pearson’s correlation whilst linear regression was used to determine the strength of the relationship. The 75th percentiles and 95% confidence intervals (CIs) of the DLP were determined. Eligible initial head CT scans for patients with injuries were 1,155. There were 685 (59.31%) males (sex ratio of 1.5:1) and overall median age of 33 years (interquartile range: 21 to 46). Road traffic injuries were the source of injury in 1090 individuals (94.37%; 95% CI: 92.88 - 95.63%), followed by falls (50 cases; 4.33%; 95% CI: 3.23 - 5.67%) and assaults (10 cases; 0.87%; 95% CI: 0.42 – 1.59). The median DLP was 1,062 mGy.cm (range: 264 to 1,954 mGy.cm). There was a positive linear relationship between the DLP and the age of the patients (Pearson’s rho = 0.38, p<0.001). DRLs were comparable to international values. Measures to curb the rising incidence of injuries and the continuous implementation of CT optimization techniques will reduce IR exposure.

ABSTRACT
Bronchogenic cancers account for an overwhelming majority of all malignancies responsible for Pancoast syndrome. However, the spectrum of non-bronchogenic malignant causes is broad and although less frequent, they are characterized by late diagnosis at advanced disease stage with a poor prognosis. The clinical presentation of Pancoast tumors can be polymorphic based on the etiology and anatomic extent, leading to considerable delays in the request of appropriate investigative techniques. Chest radiography may be helpful in first-line imaging but the cross-sectional modalities are more sensitive for diagnosis, staging and follow-up of treatment. Early reporting of symptoms and increased clinical vigilance could lead to an early diagnosis and improved 5-year survival. We report the case of a 40-year-old male with worsening left shoulder and arm pain over the course of a few months and a left neck swelling.