Since the outbreak of coronavirus disease 2019 (COVID-19), chest computed tomography (CT) has been an important imaging modality in the diagnosis, treatment and follow-up of patients with COVID-19,but meanwhile the risk of cross-infection between the staff and patients in Department of Radiology is increasing. Shelter CT is specifically used for the examination of patients with suspected or confirmed COVID-19 to reduce the infection risk. Based on practical work experience, the management and prevention measures for COVID-19 in shelter CT are discussed from the aspects of the installation, function division and examination procedures of shelter CT, patient examination route, the staff management and infection prevention for radiology technologists, and the disinfection of CT equipments and object surface.

In grating-based phase contrast imaging,the phase stepping technique is commonly utilized for data acquisition and signal retrieval from acquired intensity data.However,the algorithm efficiency with respect to the dark-field retrieval has yet to be sufficiently evaluated.Herein the algorithm efficiency of dark-field retrieval based on Cramér-Rao lower bound is evaluated.The theoretical analysis and numerical results demonstrates that fully efficient algorithm is currently available only for 3-step phase stepping technique,and other techniques with more phase steps are all sub-optimal.Quantitatively,the dependence of the algorithm efficiency on the phase step number and the visibility is investigated.It is shown that the phase stepping technique can nearly approach its theoretical optimal efficiency in the case of a low visibility.With a phase step greater than 5,the algorithm efficiency is only 77.4%in the case of a high visibility.The study can provide some reference for signal-to-noise ratio improvement and potential dose optimization in X-ray and neutron grating-based dark-field imaging.

Objective To analyze the individualized CTV-to-PTV margin dose and positioning errors in radiotherapy for esophageal cancer for improving the treatment accuracy while meeting dose requirements.Methods Fifty-four esophageal cancer patients admitted to Sun Yat-sen University Cancer Center at Huangpu District from June 2021 to June 2022 were enrolled.All of the patients underwent CBCT scans in each fraction,and a total of 1283 CBCT images were collected.The image registration between CBCT image before radiotherapy and planning CT image was carried out to obtain errors in vertical(VRT),longitudinal(LNG),lateral(LAT),Roll,Pitch,and YAW directions.The mean values of six-dimensional positioning errors in the first 5 fractions were calculated,and the results were compared with the total fractional errors using the single sample t-test method for determining the differences.The CTV-to-PTV margin was calculated with the formula(margin=2.5∑+0.7δ),and the calculated margins were divided into 5 groups:Group A(5 mm expansion in all directions),Group B(7.9 mm expansion in LAT direction,and 5 mm expansion in other directions),Group C(11.03 mm expansion in LNG direction,and 5 mm expansion in the other directions),Group D(6.36 mm expansion in VRT direction,and 5 mm expansion in the other directions),and Group E(7.9 mm expansion in LAT direction,11.03 mm expansion in LNG direction,and 6.36 mm expansion in VRT direction).Simulation planning was conducted for 10 patients.Results The proportions of differences between the mean values of six-dimensional errors in the first 5 fractions and the total fractional errors in 54 patients were analyzed.There was no significant difference in 192 out of the 324 directions in 54 patients,accounting for 59.26%(P>0.05).Among them,the LAT,LNG,VRT,Pitch,Roll and YAW directions accounted for 64.81%,57.41%,51.85%,64.81%,57.41%and 59.26%of the total cases.The calculated CTV-to-PTV margin was 7.90,11.03 and 6.36 mm in LAT,LNG and VRT directions.The statistical analysis showed that the differences in the coverage rates of organs-at-risk and target areas among the 5 groups of CTV-to-PTV margins were trivial(P>0.05).Conclusion Using the positioning errors in the first 5 fractions of radiotherapy for esophageal cancer to predict subsequent positioning errors is feasible.The reasonable individualized margin in radiotherapy for esophageal cancer can reduce the inter-fractional off-target rate without increasing the dose delivered to organs-at-risk.The study provides a reference for the target volume margin of esophageal cancer and an important basis for precision treatment.

Objective To analyze the target volume margins and positioning errors in the radiotherapy for nasopharyngeal carcinoma(NPC)using the cone-beam computed tomography(CBCT)of Halcyon linear accelerator for providing a reference for the margin from clinical target volume to planning target volume(CTV-to-PTV margin)in the radiotherapy for NPC using Halcyon linear accelerator,hence improving treatment precision and effectiveness.Methods A total of 117 NPC patients who received volumetric modulated arc therapy using Halcyon linear accelerator from May 2020 to June 2022 in Jinshazhou Hospital of Guangzhou University of Chinese Medicine were enrolled.The 3861 CBCT images collected from the patients were matched with the CT images to obtain the correction values of the treatment couch in lateral(Lat),longitudinal(Lng)and vertical(Vrt)directions for positioning error analysis.The CTV-to-PTV margin was obtained by the equation(margin =2.5∑+0.7δ).Results The positioning errors in the radiotherapy for NPC using Halcyon linear accelerator were 0.10(0.00,0.10)cm,0.10(0.00,0.20)cm and 0.20(0.10,0.30)cm in Lat,Lng and Vrt directions,respectively.The CTV-to-PTV margins in Lat,Lng and Vrt directions were 0.12,0.12 and 0.09 cm,respectively.Conclusion Low positioning errors can be achieved for NPC patients undergoing image-guided treatment using Halcyon linear accelerator.

Objective To develop an auto-segmentation model based on no new U-net for delineating high-risk clinical target volume(HR-CTV)and organs-at-risk(OAR)in CT-guided brachytherapy of cervical cancer,and to explore its clinical value.Methods The CT images of 63 patients with locally advanced cervical cancer who had completed image-guided brachytherapy were collected.The HR-CTV and OAR including bladder,rectum and sigmoid colon were delineated manually by a senior oncologist,and the results were taken as the gold standard.The automatic and manual segmentation results were compared,and Dice similarity coefficient was used to evaluate HR-CTV and OAR auto-segmentation accuracies.Results The Dice similarity coefficients of HR-CTV,bladder,rectum,and sigmoid colon were 0.903±0.015,0.948±0.011,0.903±0.008,and 0.803±0.024,respectively.Conclusion The established model can realize the accurate segmentations of HR-CTV,bladder,rectum and sigmoid colon,but the oncologist still needs to scrupulously check the results.

Eyelid tumor is a serious eye disease that leads to vision loss or even blindness.The similarity between benign and malignant characteristics makes it difficult for ophthalmologists lacking clinical experience to distinguish between them.To address the problem,a method(ResNet101_CBAM)based on two-stage target localization using fully convolutional one-stage object detection(FCOS)and residual network incorporating a dual attention mechanism is proposed to realize the automatic diagnosis of benign and malignant eyelid tumors.FCOS is used to automatically localize the overall contour of the orbit,removing the background and surrounding noise,and then finely localize the tumor lesion inside the orbit.The obtained lesion region is input into ResNet101_CBAM for the automatic diagnosis of benign and malignant eyelid tumors.The experimental results show that the average precision of the target localization algorithm for tumor lesion is 0.821,and that compared with ResNet101,ResNet101_CBAM improves the sensitivity and accuracy in eyelid tumor classification by 4.7%and 3.0%,respectively,indicating that the proposed model has superior performances in the automatic diagnosis of benign and malignant eyelid tumors.

In view of numerous subtle features in fundus disease images,small sample sizes,and difficulties in diagnosis,both deep learning and medical imaging technologies are used to develop a fundus disease diagnosis model that integrates multi-scale features and hybrid domain attention mechanism.Resnet50 network is taken as the baseline network,and it is modified in the study.The method uses parallel multi-branch architecture to extract the features of fundus diseases under different receptive fields for effectively improving the feature extraction ability and computational efficiency,and adopts hybrid domain attention mechanism to select information that is more critical to the current task for effectively enhancing the classification performance.The test on ODIR dataset shows that the proposed method has a diagnostic accuracy of 93.2%for different fundus diseases,which is 5.2%higher than the baseline network,demonstrating a good diagnostic performance.

A fall detection algorithm for community healthcare is proposed to avoid the secondary injury caused by untimely treatment when the elder living alone falls in the community.The algorithm has two branches,namely 2D convolution and 3D convolution,which allow it can extract spatial and temporal features simultaneously.The dense connections added in the 3D branch enhance the ability to extract temporal features;the residual blocks in the 2D branch are redesigned to improve the ability of spatial feature extraction;and a non-local attention mechanism is introduced to the branch fusion for better feature fusion.The algorithm also takes scene information into consideration,and it is supervised by SIoU loss function and the combined loss function to realize fall detection.The experiment on the expanded public URFD dataset reveals that the proposed method has a detection accuracy of 98.3%,which verifies its performance and robustness for fall detection.

Objective To construct a COVID-19 CT image classification model based on lightweight RG DenseNet.Methods A RG-DenseNet model was constructed by adding channel and spatial attention modules to DenseNet121 for minimizing the interference of irrelevant features,and replacing Bottleneck module in DenseNet with pre-activated RG beneck2 module for reducing model parameters while maintaining accuracy as much as possible.The model performance was verified with 3-category classification experiments on the COVIDx CT-2A dataset.Results RG-DenseNet had an accuracy,precision,recall rate,specificity,and F1-score of 98.93%,98.70%,98.97%,99.48%,and 98.83%,respectively.Conclusion Compared with the original model DenseNet121,RG-DenseNet reduces the number of parameters and the computational complexity by 92.7%,while maintaining an accuracy reduction of only 0.01%,demonstrating a significant lightweight effect and high practical application value.

Deep learning can automatically learn representative features from image data for clinical analysis,such as glioma staging/grading,prediction of molecular marker status,differentiation of tumor pseudoprogression from true progression,and survival prediction.Radiomics aims to quantitatively describe tumors based on imaging features extracted from routine medical images,and it can capture small changes in tissues and lesions,such as heterogeneity within tumor volume,tumor shape,and their changes over time during serial imaging.Image analysis technology based on radiomics and deep learning can simplify and automate the diagnosis and treatment of glioma,with high accuracy.The review gives a brief introduction of radiomics methods and deep learning technologies,and then summarizes the application of radiomics methods and deep learning technologies in the diagnosis and treatment of glioma in recent years,expecting to provide a preoperative basis for the treatment scheme selection for glioma patients.