OBJECTIVE: The objective of this study was to develop and evaluate predictive models based on a combination of T2-weighted images (T2WI) and different machine learning algorithms, and to explore the value of hepatic alveolar echinococcosis (HAE) activity assessment by magnetic resonance imaging (MRI) radiomics. MATERIALS AND METHODS: This retrospective study included 136 patients diagnosed with HAE at the First Affiliated Hospital of Xinjiang Medical University between 2012 and 2020. All subjects underwent MRI and positron emission tomography–computed tomography (PET-CT) before surgery. Taking the PET-CT examination results as the reference standard, patients were divided into active (90 cases) and inactive groups (46 cases). The volume of interest of the lesion was manually delineated on T2WI, and quantitative radiomics features were extracted. Synthetic Minority Oversampling Technology was used to balance the number of patients in the categories. To control for redundancy, the least absolute shrinkage and selection operator was used for feature screening after normalization, and ten optimal features were obtained based on correlation coefficient screening. Three machine learning classifiers were trained using five-fold cross-validation and their performance was compared to establish an optimal HAE activity assessment model. The performance of the classifier was evaluated by area under the receiver operating characteristics curve (AUC), sensitivity, specificity, and accuracy (ACC). The ten optimal features selected from each fold were combined using three machine learning algorithms: logistic regression, multilayer perceptron (MLP), and support vector machine, to establish an HAE activity prediction model. RESULTS: The three machine learning classifiers all showed good prediction performance with a mean AUC on the test set of more than 0.80, and the MLP showing the best performance (AUC = 0.830 ± 0.053, ACC = 0.817, sensitivity = 0.822, and specificity = 0.811). CONCLUSION: HAE activity can be accurately evaluated by a radiomics method using a combination of quantitative T2WI features and machine learning.

[ABSTRACT]  [FULL TEXT]  [PDF]  [Mobile Full text]  [EPub]

OBJECTIVES: The objective of this study is to evaluate the clinical profile of coronavirus disease-2019 (COVID-19) patients admitted to our hospital and to correlate their chest radiographic patterns with disease severity. MATERIALS AND METHODS: We retrospectively reviewed 500 patients with COVID-19 confirmed by reverse transcription-polymerase chain reaction who had abnormal baseline chest X-rays (CXRs) at the time of hospital admission. CXRs were characterized based on the site and nature of the lesions. Disease severity was determined using the Radiographic Assessment of Lung Edema (RALE) score. RESULTS: Significant associations were found between (1) the lesion site and patient outcome (P < 0.00001): patients with diffuse and basal infiltrates had high intensive care unit (ICU) admission rates (55.5% and 40%) and mortality rates (30.5% and 20%); (2) the nature of the lesion and patient outcome: patients with ground-glass opacities and consolidation had high mortality (20% and 18%, respectively); and (3) the RALE score and patient outcome: patients with a RALE score >15 had higher ICU admission and mortality rates. CONCLUSIONS: The CXR distribution patterns helped to triage patients and predict outcomes.

[ABSTRACT]  [FULL TEXT]  [PDF]  [Mobile Full text]  [EPub]

BACKGROUND: Brain abscess is a rare disease in childhood requiring prompt medical and/or surgical treatment. The objective was to discuss the head computed tomography and magnetic resonance imaging characteristics of children with brain abscess in congenital heart disease (CHD) group compared with the group without CHD, further comprehend the imaging findings, and assess the severity of brain abscess in children with CHD. MATERIALS AND METHODS: The radiological features of brain abscess in children hospitalized in Shanghai Children's Medical Center between September 2014 and September 2021 were retrospectively reviewed. RESULTS: Forty-four children (14 females, 30 males), with a median age of 8.6 years (range 2–15 years), were enrolled in this study. CHD was found in 22 cases. Twenty-one of these 22 patients were with cyanotic CHD. The remaining 22 cases were without CHD. Among the 22 patients with CHD, superficial abscesses of 2–5 cm in diameter are the main imaging findings, which was the same as in children without CHD. In our study, brain abscesses of CHD are usually single, while other brain abscesses are usually multiple in children without CHD. There was statistically significant difference in the number of abscesses between the group with CHD and the group without CHD (χ² = 6.3, P = 0.04). Compared with no CHD children, the nearest distance from the ventricular wall to the margin of brain abscess in children with CHD is often <7 mm. CONCLUSIONS: Brain abscesses in children with CHD have some special imaging characteristics. Superficial abscesses of 2–5 cm in diameter are the main imaging findings. Brain abscesses of CHD are usually single, while other brain abscesses are usually multiple in children without CHD. Compared with no CHD children, the nearest distance from the ventricular wall to the margin of brain abscess in children with CHD is shorter.

[ABSTRACT]  [FULL TEXT]  [PDF]  [Mobile Full text]  [EPub]

Mycoplasma pneumoniae pneumonia (MPP) is the most common type of childhood community-acquired pneumonia. MPP is generally mild and self-limiting, but a small percentage of patients still develop a refractory or severe clinical course. Imaging is an important tool for confirmed pneumonia, and it plays an important role in the diagnosis of MPP and assessment of the clinical course. However, imaging features of MPP reported in the literature vary in their patterns to distribution. A comprehensive and deep understanding of imaging findings of MPP in children is beneficial for an accurate diagnosis and guidance of its treatment.

[ABSTRACT]  [FULL TEXT]  [PDF]  [Mobile Full text]  [EPub]

Coronavirus disease (COVID-19) is highly infectious, has spread worldwide, and has a relatively high mortality rate. Early diagnosis and timely isolation are essential to control the spread of COVID-19. Computed tomography (CT) is considered to be an effective tool for the rapid diagnosis of COVID-19 and plays a key role in diagnosis, clinical course monitoring, and the evaluation of treatment outcomes. Artificial intelligence (AI) has emerged as a useful technology for early diagnosis, lesion quantification, and prognosis evaluation in patients with COVID-19. In this review, we discuss the role of CT in the diagnosis of COVID-19, typical CT manifestations of COVID-19 throughout the disease course, differential diagnoses, and the application of AI as a diagnostic and therapeutic tool in this patient population.

[ABSTRACT]  [FULL TEXT]  [PDF]  [Mobile Full text]  [EPub]

Emphysematous pyelonephritis (EPN) is a rare but serious infectious disease caused by anaerobic bacteria that is characterized by necrotizing renal parenchymal infection and has a high mortality rate. The initial management of EPN includes resuscitation, antibiotics, management of any diabetes, and minimally invasive procedures, such as percutaneous drainage. Surgery is required in the most severe cases. Here, we describe a male patient with type 2 diabetes who presented with fever and abdominal pain and was diagnosed with EPN based on laboratory results and a computed tomography (CT) scan showing distension of the left kidney, several small air bubbles in the perirenal space, and thickening of the perinephric fascia. A subcutaneous insulin infusion and meropenem were administered. Repeat CT imaging 13 days later showed gas and necrotic tissue in the renal parenchyma. Therefore, CT-guided renal puncture was performed, and Escherichia coli was cultured from the drained pus. Antibiotic treatment was continued for 20 days after admission when the patient's kidney function and clinical symptoms had significantly improved. Sixty days after diagnosis, left nephrectomy was performed. The perinephric abscess was under high tension, and the left kidney was very soft. Histopathological examination revealed severe inflammation and necrosis of the renal pelvis, with fibrosis. The patient recovered well after surgery and remained alive 5 months later. EPN is a very rare disease with a high mortality rate and is more likely to occur in patients with diabetes. Early diagnosis and treatment of the infection are extremely important. For patients in whom gas and/or necrosis continue to accumulate after conservative treatment, we recommend nephrectomy.

[ABSTRACT]  [FULL TEXT]  [PDF]  [Mobile Full text]  [EPub]

A middle-aged patient presented with epigastric pain, elevated white blood cell count, and mild anemia. Computed tomography (CT) revealed an exophytic gastric mass that mimicked a hepatic abscess or tumor, with intraperitoneal hemorrhage. The mass had heterogeneous texture, interstitial bleeding, and cystic change that showed delayed mild enhancement on multiphase CT. This was consistent with gastrointestinal stromal tumor that was confirmed by pathological and immunohistochemical analysis.

[ABSTRACT]  [FULL TEXT]  [PDF]  [Mobile Full text]  [EPub]