OBJECTIVE:To investigate the diagnostic meaning of MRI in intraocular tumors. METHODS:Forty-six cases of confirmed intraocular tumors,including choroidal melanoma(20 cases),retinoblastoma(18 cases),Coats disease(6 cases)and choroidal hemangioma(2 cases),were studied with MRI and compared with ultrasonography and CT. RESULTS:In making discoveries about intraocular tumors,there were no sighificant difference between MRI and B-ultrasonography or CT （P＞0.03，chi;2=1.0716）while there were highly statistic sighificance in dediding characters and position （P＜0.01，deceding character chi;2=29.8314，positionchi;2=13.659）of them. CONCLUSION:Among the examinations to find out about the position,character and secondary pathological insults of in traocular tumors MRI might be more available than CT and ultrasonography. （Chin J Ocul Fundus Dis,1997,13:93-95 ）
Focus on the inconsistency of the shape, location and size of brain glioma, a dual-channel 3-dimensional (3D) densely connected network is proposed to automatically segment brain glioma tumor on magnetic resonance images. Our method is based on a 3D convolutional neural network frame, and two convolution kernel sizes are adopted in each channel to extract multi-scale features in different scales of receptive fields. Then we construct two densely connected blocks in each pathway for feature learning and transmission. Finally, the concatenation of two pathway features was sent to classification layer to classify central region voxels to segment brain tumor automatically. We train and test our model on open brain tumor segmentation challenge dataset, and we also compared our results with other models. Experimental results show that our algorithm can segment different tumor lesions more accurately. It has important application value in the clinical diagnosis and treatment of brain tumor diseases.
ObjectiveAicardi and Goutières syndrome was first reported as a rare hereditary encephalopathy with white matter involvement in 1984. Typical clinical manifestations include severe mental motor development retardation or regression, pyramidal and extrapyramidal symptoms and signs, epilepsy, microcephaly and frostbite.MethodsTo collect a case of patient who presented with convulsions 14 days after birth without obvious inducement. The child was diagnosed as epilepsy in the local hospital and the symptoms improved after treatment with antiepileptic drugs. At 4 months, the child presented nods and clenched fists, and was diagnosed as infantile spasm. After Adrenocorticotrophic hormone and drug treatment, the symptoms gradually improved. Due to upper respiratory track infection, the child was aggravated at the age of 1 year and 2 months, and then diagnosed as Aicardi-Goutières syndrome by video EEG, skull MRI, fundus and gene screening.ResultsSurgery and treatment with antiepileptic drugs significantly improved the symptoms of the child, and the pathological biopsy of the brain tissue supported the previous diagnosis.ConclusionsThe report of this case will help to improve the clinician's diagnosis and treatment of Aicardi-Goutières syndrome.
This paper performs a comprehensive study on the computer-aided detection for the medical diagnosis with deep learning. Based on the region convolution neural network and the prior knowledge of target, this algorithm uses the region proposal network, the region of interest pooling strategy, introduces the multi-task loss function: classification loss, bounding box localization loss and object rotation loss, and optimizes it by end-to-end. For medical image it locates the target automatically, and provides the localization result for the next stage task of segmentation. For the detection of left ventricular in echocardiography, proposed additional landmarks such as mitral annulus, endocardial pad and apical position, were used to estimate the left ventricular posture effectively. In order to verify the robustness and effectiveness of the algorithm, the experimental data of ultrasonic and nuclear magnetic resonance images are selected. Experimental results show that the algorithm is fast, accurate and effective.
ObjectiveTo investigate the clinical value of magnetic resonance imaging (MRI) combined with ultrasound (US) contrasting with MRI in evaluating the pathological complete response (pCR) of breast cancer after neoadjuvant chemotherapy (NAC).MethodsThe imaging data of patients with primary invasive breast cancer who completed the surgical resection after NAC and met the inclusion criteria in the Breast Disease Diagnosis and Treatment Center of Affiliated Hospital of Qingdao University from December 2016 to December 2019 were collected retrospectively. These patients were evaluated by MRI and MRI combined with US examination respectively. The results of MRI alone and MRI combined with US were designed into imaging of complete remission (rCR) and imaging of non-complete remission (non-rCR). With results of postoperative pathology as the gold standard, the sensitivity, specificity, and positive predictive value (PPV) of MRI alone and MRI combined with US in predicting pCR of patients with rCR or non-rCR were calculated and which were further analyzed in the 4 subtypes of breast cancer (HR+/HER2+, HR+/HER2–, HR–/HER2+, and HR–/HER2– subtype).Results① According to the inclusion and exclusion criteria, a total of 146 patients with primary invasive breast cancer were included, including 34 cases of HR+/HER2+subtype, 63 cases of HR+/HER2– subtype, 23 cases of HR–/HER2+ subtype, and 26 cases of HR–/HER2– subtype. ② After NAC, 36 cases had a pCR, among which 9 cases (26.5%) were in HR+/HER2+ subtype, 10 cases (15.9%) were in HR+/HER2– subtype, 8 cases (34.8%) were in HR–/HER2+ subtype, and 9 cases (34.6%) were in HR–/HER2– subtype. ③ After NAC, 22 (78.6%) of the 28 patients evaluated by MRI alone achieved pCR, 17 (81.0%) of the 21 patients evaluated by MRI combined with US achieved pCR, and the PPV value of pCR evaluated by MRI alone and MRI combined with US was 78.6% and 81.0%, respectively. ④ Both MRI alone and MRI combined with US predicted NAC showed the highest PPV values in patients with HR–/HER2– subtype breast cancer (85.7% and 100%, respectively), and the lowest values in HR+/ HER2– subtype breast cancer (71.4% and 60.0%, respectively).ConclusionFor the overall patients with primary invasive breast cancer, MRI combined with US is superior to MRI alone in the evaluation of efficacy after NAC, and among the patients with different subtypes of breast cancer, except HR+/HER2– subtype, MRI combined with US is still more effective in predicting efficacy after NAC than MRI alone.
Urokinase plasminogen activator receptor (uPAR) is a membrane protein which is attached to the cellular external membrane. The uPAR expression can be observed both in tumor cells and in tumor-associated stromal cells. Thus, in the present study, the human amino-terminal fragment (hATF), as a targeting element to uPAR, is used to conjugate to the surface of superparamagnetic iron nanoparticle (SPIO). Flowcytometry was used to examine the uPAR expression in different tumor cell lines. The specificity of hATF-SPIO was verified by Prussian blue stain and cell phantom test. The imaging properties of hATF-SPIO were confirmed in vivo magnetic resonance imaging (MRI) of uPAR-elevated colon tumor. Finally, the distribution of hATF-SPIO in tumor tissue was confirmed by pathological staining. Results showed that the three cells in which we screened, presented different expression characteristics, i.e., Hela cells strongly expressed uPAR, HT29 cells moderately expressed uPAR, but Lovo cells didn't express uPAR. In vitro, after incubating with Hela cells, hATF-SPIO could specifically combined to and be subsequently internalized by uPAR positive cells, which could be observed via Prussian blue staining. Meanwhile T2WI signal intensity of Hela cells, after incubation with targeted probe, significantly decreased, and otherwise no obvious changes in Lovo cells both by Prussian blue staining and MRI scans. In vivo, hATF-SPIO could be systematically delivered to HT29 xenograft and accumulated in the tumor tissue which was confirmed by Prussian Blue stain compared to Lovo xenografts. Twenty-four hours after injection of targeting probe, the signal intensity of HT29 xenografts was lower than Lovo ones which was statistically significant. This targeting nanoparticles enabled not only in vitro specifically combining to uPAR positive cells but also in vivo imaging of uPAR moderately elevated colon cancer lesions.
ObjectiveTo explore the correlation between cognitive function and diffusion tensor imaging (DTI) in children with self-limited epilepsy with centrotemporal spikes (SelECTS). Methods A total of 28 children with SelECTS who visited our hospital from June 2020 to December 2022 were selected as the SelECTS group. An additional 28 healthy children of similar age and gender were selected as the control group. Cognitive function was assessed using the Wechsler Intelligence Scale for Children (WISC). The SelECTS group also underwent cranial DTI. The results of the WISC were then combined with DTI values for correlation analysis. Results Children in the SelECTS group exhibited varying degrees of cognitive deficits. Their full-scale IQ and verbal IQ were significantly lower than those of the control group (P<0.05). Specific cognitive domains, including classification, verbal comprehension, block design, knowledge, and comprehension, also showed significantly lower scores compared to the control group (P<0.05). DTI revealed significant microstructural changes in multiple regions of interest in the SelECTS group (P<0.05), and these changes were correlated with the results of several cognitive function tests. Conclusion Children with SelECTS have certain cognitive deficits. There is evidence of occult damage in brain white matter, and cognitive function is correlated with damage in specific brain regions.
The effect of deep brain stimulation (DBS) surgery treatment for Parkinson's disease is determined by the accuracy of the electrodes placement and localization. The subthalamic nuclei (STN) as the implant target is small and has no clear boundary on the images. In addition, the intra-operative magnetic resonance images (MRI) have such a low resolution that the artifacts of the electrodes impact the observation. The three-dimensional (3D) visualization of STN and other nuclei nearby is able to provide the surgeons with direct and accurate localizing information. In this study, pre- and intra-operative MRIs of the Parkinson's disease patients were used to realize the 3D visualization. After making a co-registration between the high-resolution pre-operative MRIs and the low-resolution intra-operative MRIs, we normalized the MRIs into a standard atlas space. We used a special threshold mask to search the lead trajectories in each axial slice. After checking the location of the electrode contacts with the coronal MRIs of the patients, we reconstructed the whole lead trajectories. Then the STN and other nuclei nearby in the standard atlas space were visualized with the grey images of the standard atlas, accomplishing the lead reconstruction and nerve nuclei visualization near STN of all patients. This study provides intuitive and quantitative information to identify the accuracy of the DBS electrode implantation, which could help decide the post-operative programming setting.
The medical magnetic resonance (MR) image reconstruction is one of the key technologies in the field of magnetic resonance imaging (MRI). The compressed sensing (CS) theory indicates that the image can be reconstructed accurately from highly undersampled measurements by using the sparsity of the MR image. However, how to improve the image reconstruction quality by employing more sparse priors of the image becomes a crucial issue for MRI. In this paper, an adaptive image reconstruction model fusing the double dictionary learning is proposed by exploiting sparse priors of the MR image in the image domain and transform domain. The double sparse model which combines synthesis sparse model with sparse transform model is applied to the CS MR image reconstruction according to the complementarity of synthesis sparse and sparse transform model. Making full use of the two sparse priors of the image under the synthesis dictionary and transform dictionary learning, the proposed model is tackled in stages by the iterative alternating minimization algorithm. The solution procedure needs to utilize the synthesis and transform K-singular value decomposition (K-SVD) algorithms. Compared with the existing MRI models, the experimental results show that the proposed model can more efficiently improve the quality of the image reconstruction, and has faster convergence speed and better robustness to noise.