The diagnostic value of CT-guided puncture biopsy combined with serum gamma-glutamyltransferase and abnormal prothrombin in serum alpha-fetoprotein negative primary liver cancer / 中国医师进修杂志

Objective:To investigate the diagnostic value of CT-guided puncture biopsy combined with serum gamma-glutamyltransferase (GGT) and abnormal prothrombin (PIVKA-Ⅱ) in serum alpha-fetoprotein(AFP) negative primary liver cancer (PHC).Methods:Eighty patients with AFP negative PHC treatment in Fuyang Women and Children′s Hospital from January 2018 to March 2021 were selected as AFP negative PHC group, and another 85 patients diagnosed with benign liver tumor during the same period were selected as the control group retrospectively. The patients of the two groups underwent CT-guided biopsy and the levels of GGT and PIVKA-Ⅱ were detected. The single diagnostic value and combined diagnostic value of AFP negative PHC were analyzed by receiver operating characteristic (ROC) curve.Results:Seventy-five of the 80 patients in the AFP negative PHC group were confirmed as liver malignant lesions by biopsy, with a coincidence of 93.75%, and 84 of the 85 patients in the control group were confirmed as liver benign lesions by biopsy, with a coincidence of 98.82%. The levels of AFP, GGT and PIVKA-Ⅱ in AFP negative PHC group were significantly higher than those in the control group: (175.67 ± 39.58) μg/L vs. (18.74 ± 7.42) μg/L, (1 245.37 ± 255.41) U/L vs. (486.63 ± 89.05) U/L, (385.49 ± 30.27) AU/L vs. (25.84 ± 7.66) AU/L, there were statistical differences ( P<0.05). Spearman correlation analysis showed that serum AFP was positively correlated with GGT and PIVKA-Ⅱ ( r = 0.858 and 0.429, P<0.05). The results of ROC curve showed that the area under curve of CT-guided biopsy combined with GGT and PIVKA-Ⅱ in the diagnosis of AFP negative PHC was 0.877, the sensitivity was 91.19%, the specificity was 87.34%. Conclusions:CT-guided biopsy combined with GGT and PIVKA-Ⅱ detection of AFP negative PHC can effectively improve the diagnostic value.

Application of a light-weighted convolutional neural network for automatic recognition of coal workers' pneumoconiosis in the early stage / 中华劳动卫生职业病杂志

Objective: To construct and verify a light-weighted convolutional neural network (CNN), and explore its application value for screening the early stage (subcategory 0/1 and stage Ⅰ of pneumoconiosis) of coal workers' pneumoconiosis (CWP) from digital chest radiography (DR) . Methods: A total of 1225 DR images of coal workers who were examined at an Occupational Disease Prevention and Control Institute in Anhui Province from October 2018 to March 2021 were retrospectively collected. All DR images were collectively diagnosed by 3 radiologists with diagnostic qualifications and gave diagnostic results. There were 692 DR images with small opacity profusion 0/- or 0/0 and 533 DR images with small opacity profusion 0/1 to stage Ⅲ of pneumoconiosis. The original chest radiographs were preprocessed differently to generate four datasets, namely 16-bit grayscale original image set (Origin16), 8-bit grayscale original image set (Origin 8), 16-bit grayscale histogram equalized image set (HE16) and 8-bit grayscale histogram equalized image set (HE8). The light-weighted CNN, ShuffleNet, was applied to train the generated prediction model on the four datasets separately. The performance of the four models for pneumoconiosis prediction was evaluated on a test set containing 130 DR images using measures such as the receiver operating characteristic (ROC) curve, accuracy, sensitivity, specificity, and Youden index. The Kappa consistency test was used to compare the agreement between the model predictions and the physician diagnosed pneumoconiosis results. Results: Origin16 model achieved the highest ROC area under the curve (AUC=0.958), accuracy (92.3%), specificity (92.9%), and Youden index (0.8452) for predicting pneumoconiosis, with a sensitivity of 91.7%. And the highest consistency between identification and physician diagnosis was observed for Origin16 model (Kappa value was 0.845, 95%CI: 0.753-0.937, P<0.001). HE16 model had the highest sensitivity (98.3%) . Conclusion: The light-weighted CNN ShuffleNet model can efficiently identify the early stages of CWP, and its application in the early screening of CWP can effectively improve physicians' work efficiency.

Application of deep learning to the differenciation of the invasion depth in colorectal adenomas / 中华消化内镜杂志

Objective:To evaluate deep learning for differentiating invasion depth of colorectal adenomas under image enhanced endoscopy (IEE).Methods:A total of 13 246 IEE images from 3 714 lesions acquired from November 2016 to June 2021 were retrospectively collected in Renmin Hospital of Wuhan University, Shenzhen Hospital of Southern Medical University and the First Hospital of Yichang to construct a deep learning model to differentiate submucosal deep invasion and non-submucosal deep invasion lesions of colorectal adenomas. The performance of the deep learning model was validated in an independent test and an external test. The full test was used to compare the diagnostic performance between 5 endoscopists and the deep learning model. A total of 35 videos were collected from January to June 2021 in Renmin Hospital of Wuhan University to validate the diagnostic performance of the endoscopists with the assistance of deep learning model.Results:The accuracy and Youden index of the deep learning model in image test set were 93.08% (821/882) and 0.86, which were better than those of endoscopists [the highest were 91.72% (809/882) and 0.78]. In video test set, the accuracy and Youden index of the model were 97.14% (34/35) and 0.94. With the assistance of the model, the accuracy of endoscopists was significantly improved [the highest was 97.14% (34/35)].Conclusion:The deep learning model obtained in this study could identify submucosal lesions with deep invasion accurately for colorectal adenomas, and could improve the diagnostic accuracy of endoscopists.

Research on lightweight model of intelligent-assisted diagnosis of common fundus diseases based on fundus color photography / 中华眼底病杂志

Objective:To observe the diagnostic value of six classification intelligent auxiliary diagnosis lightweight model for common fundus diseases based on fundus color photography.Methods:A applied research. A dataset of 2 400 color fundus images from Nanjing Medical University Eye Hospital and Zhejiang Mathematical Medical Society Smart Eye Database was collected, which was desensitized and labeled by a fundus specialist. Of these, 400 each were for diabetic retinopathy, glaucoma, retinal vein occlusion, high myopia, age-related macular degeneration, and normal fundus. The parameters obtained from the classical classification models VGGNet16, ResNet50, DenseNet121 and lightweight classification models MobileNet3, ShuffleNet2, GhostNet trained on the ImageNet dataset were migrated to the six-classified common fundus disease intelligent aid diagnostic model using a migration learning approach during training as initialization parameters for training to obtain the latest model. 1 315 color fundus images of clinical patients were used as the test set. Evaluation metrics included sensitivity, specificity, accuracy, F1-Score and agreement of diagnostic tests (Kappa value); comparison of subject working characteristic curves as well as area under the curve values for different models.Result:Compared with the classical classification model, the storage size and number of parameters of the three lightweight classification models were significantly reduced, with ShuffleNetV2 having an average recognition time per sheet 438.08 ms faster than the classical classification model VGGNet16. All 3 lightweight classification models had Accuracy > 80.0%; Kappa values > 70.0% with significant agreement; sensitivity, specificity, and F1-Score for the diagnosis of normal fundus images were ≥ 98.0%; Macro-F1 was 78.2%, 79.4%, and 81.5%, respectively.Conclusion:The intelligent assisted diagnosis of common fundus diseases based on fundus color photography is a lightweight model with high recognition accuracy and speed; the storage size and number of parameters are significantly reduced compared with the classical classification model.

Infiltration of tumor-associated macrophages and its association with prognosis of patients with diffuse large B-cell lymphoma based on gene expression profile data / 白血病·淋巴瘤

Objective:To analyze the infiltration of tumor-associated macrophages and their subtypes, and to investigate their association with prognosis of patients with diffuse large B-cell lymphoma (DLBCL) based on the gene chip expression database.Methods:The data were retrieved from microarray (Affymetrix U133 plus 2.0) database (No:GSE10846) of DLBCL patients in PubMed gene expression omnibus (GEO). The database included 414 DLBCL patients, among which 306 cases had complete clinical, cell of origin phenotype (COO subtyping), treatment and follow-up information. The data analysis was performed on the online computer program which could identify the cell-type (CIBERSORT) by estimating relative percentage of RNA transcripts. From the returned result file, the percentage of immune cells including macrophages subtypes of all cases in all identifiable immune cells in the microenvironment was identified in GSE10846 database. Taking the median percentage of macrophages subsets in all types of immune cells as cut-off value; ≥ cut-off value was high infiltration and < cut-off value was low infiltration. The median value of gene RNA expression level of myc, bcl-2, programmed death ligand-1 (PD-L1) and programmed death ligand-2 (PD-L2) of 414 DLBCL patients in the GSE10846 database was treated as the cut-off value; ≥ cut-off value was the high expression and < cut-off value was the low expression. The correlation of the expression levels of all subsets and total macrophages with clinical factors, gene expression, survival was analyzed; Cox proportional hazard model was used to make multivariate analysis of the prognosis for DLBCL patients. surv_cutpoint function of surv_miner package in R 4.0.4 software was used for the optimal cut-off value of the percentage of macrophages subsets in all immune cells in the microenvironment; the result less than the optimal cut-off value was statistically low infiltration and the result greater than or equal to the optimal cut-off value was statistically high infiltration.Results:CIBERSORT analysis showed that M0 macrophages [15.00% (0-44.41%)], M1 macrophages [7.46% (0-23.00%)] and M2 macrophages [6.28% (0-43.35%)] in the tumor microenvironment were identified in all 414 DLBCL cases. Among 306 patients with complete clinical and follow-up data, there were 155 cases (50.7%), 152 cases (49.7%), 156 cases (51.0%), 152 cases (49.7%), respectively in high infiltration patients with M0, M1, M2 and total macrophages; the high infiltration of M0 macrophages was correlated with COO subtyping germinal center B-cell (GCB) type and the high expression of PD-L1 gene, the absence of myc and bcl-2 double high expression at RNA level (R-DEL) (all P < 0.05); the high infiltration of M1 macrophages was correlated with female, the high expression of PD-L1 gene and PD-L2 gene (all P < 0.05); the high infiltration of M2 macrophages was correlated with COO subtyping GCB type, the high expression of PD-L2 gene (all P < 0.05); the high infiltration of total macrophages was correlated with female, COO subtyping GCB type, the high expression of PD-L1 gene and PD-L2 gene, the absence of R-DEL (all P < 0.05).The high expression of PD-L1 gene was associated with high infiltration of M0, M1 and total macrophages (all P < 0.01), and high PD-L2 gene expression was correlated with high infiltration of M1, M2 and total macrophages (all P < 0.01). The overall survival (OS) of M0 macrophage high infiltration group was better than that of the lower infiltration group ( P = 0.002); the OS of M2 macrophage low infiltration group was better than that of the high infiltration group ( P = 0.019). The OS of R-DEL group was worse than that of R-DEL absent group ( P = 0.001). The patients with low international prognostic index (IPI) score (0-2), COO subtyping GCB type, and treatment with rituximab had better OS (all P < 0.01). Multivariate Cox regression analysis showed that 60 years or above, COO subtyping non-GCB type, treatment without rituximab, M0 macrophage low infiltration, M2 macrophage high infiltration were all independent adverse prognostic factors for OS of DLBCL patients (all P < 0.05). The optimal cut-off value for M0 macrophages was 4.3%, and the optimal cut-off value for M2 macrophages was 4.8%, and the OS in the group with statistically low infiltration of M0 macrophage was worse ( P < 0.001), and so was the OS in the group with statistically high infiltration of M2 macrophage ( P = 0.001). Conclusions:Tumor-associated macrophage is confirmed as the most abundant immune cells in the tumor microenvironment of DLBCL. Patients with high infiltration of M2 macrophage have poor prognosis, while high infiltration of M0 macrophage indicates a better prognosis.

Misconceptions in the health technology industry that are delaying the translation of artificial intelligence technology into relevant clinical applications / Conceitos enviesados na indústria de tecnologia da saúde que retardam a tradução da inteligência artificial em ferramentas clínicas relevantes

Abstract There is great optimism that artificial intelligence (AI), as it disrupts the medical world, will provide considerable improvements in all areas of health care, from diagnosis to treatment. In addition, there is considerable evidence that AI algorithms have surpassed human performance in various tasks, such as analyzing medical images, as well as correlating symptoms and biomarkers with the diagnosis and prognosis of diseases. However, the mismatch between the performance of AI-based software and its clinical usefulness is still a major obstacle to its widespread acceptance and use by the medical community. In this article, three fundamental concepts observed in the health technology industry are highlighted as possible causative factors for this gap and might serve as a starting point for further evaluation of the structure of AI companies and of the status quo.

Resumo Há uma grande expectativa de que a inteligência artificial (IA), ao transformar a medicina, determine melhoras relevantes em todas as áreas da assistência médica, desde o diagnóstico até o tratamento. Simultaneamente, há evidências de que algoritmos baseados em IA já ultrapassaram o desempenho do ser humano em diversas atividades, como, por exemplo, na análise de imagens médicas ou na associação entre sintomas e biomarcadores com o diagnóstico e prognóstico de doenças. No entanto, a defasagem entre o potencial de desempenho das ferramentas ou aplicativos médicos que utilizam IA e sua relevância clínica prejudica bastante a utilização em larga escala desses programas de computadores. Neste artigo, três conceitos básicos da indústria de tecnologia da saúde são sugeridos como possíveis fatores causais para essa dissincronia entre desempenho e utilidade. Tal discussão pode servir como ponto de partida para uma avaliação mais profunda sobre a estrutura e status quo da indústria médica tecnológica atual.

Proposal for computer-aided diagnosis based on ultrasound images of the kidney: is it possible to compare shades of gray among such images? / Proposição de um diagnóstico auxiliado por computador de imagem renal ultrassonográfica: é possível comparar tons de cinza das imagens ultrassonográficas renais?

Abstract Objective: To compare ultrasound images of the kidney obtained, randomly or in a controlled manner (standardizing the physical aspects of the ultrasound system), by various professionals and with different devices. Materials and Methods: We evaluated a total of 919 images of kidneys, obtained by five professionals using two types of ultrasound systems, in 24 patients. The images were categorized into four types, by how they were acquired and processed. We compared the gray-scale median and different gray-scale ranges representative of virtual histological tissues. Results: There were statistically significant differences among the five professionals, regardless of the type of ultrasound system employed, in terms of the gray-scale medians for the images obtained (p < 2.2e-16). Analyzing the four categories of images-a totally random image (without any standardization); a standardized image (with fixed values for gain, time gain control, and dynamic range); a normalized version of the random image; and a normalized version of the standardized image-we determined that the random image, even after normalization, differed quite significantly among the professionals (p = 0.006098). The analysis of the normalized version of the standardized image did not differ significantly among the professionals (p = 0.7319). Conclusion: Our findings indicate that a gray-scale analysis of ultrasound images of the kidney performs better when the image acquisition process is standardized and the images undergo a process of normalization.

Resumo Objetivo: Comparar imagens renais ultrassonográficas obtidas de maneira aleatória e controlada (padronizando fatores físicos do aparelho de ultrassom) por diferentes profissionais e aparelhos. Materiais e Métodos: Foram obtidos quatro tipos de imagens, de acordo com sua aquisição e processamento por cinco profissionais e dois tipos de aparelhos de ultrassonografia, em 24 pacientes, totalizando 919 imagens. Comparamos a mediana de escala de cinza e diferentes intervalos de tons de cinza representantes de tecidos histológicos virtuais. Resultados: As medianas de escala de cinza de imagens renais obtidas por dois tipos de aparelhos foram estatisticamente diferentes (p < 2.2e-16). Analisando os quatro tipos de imagens, partindo de uma totalmente aleatória (sem qualquer padronização), uma padronizada (fixado o ganho, time gain control e dynamic range), e essas duas passando por um processo de normatização, obteve-se que a imagem aleatória é totalmente diversa entre os profissionais (p = 0,006098), mesmo passando pelo processo de normatização. A imagem padronizada, após passar pelo processo de normatização, apresentou resultados equivalentes, não possuindo diferença estatística (p = 0,7319). Conclusão: Constatou-se que na análise de tons de cinza deve-se usar um mesmo tipo de máquina e uma imagem em que sejam padronizados aspectos físicos, passando por um processo de normatização/padronização.

Research progress of CT texture analysis in the diagnosis of renal tumors / 中华全科医师杂志

The classification and differentiation of renal tumors are related to the choice of therapeutic modalities and prognosis of patients. CT texture analysis is an objective and quantitative assessment of tumor heterogeneity based on the distribution and relationship characteristics of pixel or voxel gray levels in images, which can make up for the subjective limitations of traditional image visual analysis methods for diagnosis. In this article, CT texture analysis and its application in the diagnosis of renal tumors are reviewed and the limitations are also discussed.

Diagnosis value of MRI in acute rotator interval injury / 中国骨伤

OBJECTIVE@#To compare clinical application of 1.5 T MRI in acute rotator interval injury.@*METHODS@#Totally 160 patients with acute rotator cuff tear by clinical diagnosis were retrospectively analyzed by MRI examination and arthroscopy from March 2016 to February 2019, including 122 males and 38 females, aged from 22 to 71 years old with an average of (42.35±3.48) years old. Based on the results of arthroscopy as the gold standard, the shape and signal changes of rotator cuff, rotator interval, peripheral bursa, bone and soft tissue were observed by MRI on axial, oblique coronal and oblique sagittal imagese.@*RESULTS@#The direct MRI signs of acute rotator interval injury displayed thickening, diminution, distortion, interruption of the coracohumeral ligament and superior glenohumeral ligament complex with highsignal intensity on fat-suppression by proton weighted sequence. The indirect MRI signs displayed rotator cuff, peripheral bone and soft tissue injury. The consistency of the results between the two methods was quite satisfactory (Kappa=0.85), and the concordance rate of the two methods has statistically significant (@*CONCLUSION@#MRI could clearly display acute rotator interval, and could accurately diagnose acute rotator interval injury, which provide more accurate imaging basis for clinical diagnosis and treatment.

Breast elastography: diagnostic performance of computer-aided diagnosis software and interobserver agreement / Elastografia mamária: desempenho diagnóstico de um software de sistema de diagnóstico assistido por computador e concordância interobservador

Abstract Objective: To determine the best cutoff value for classifying breast masses by ultrasound elastography, using dedicated software for strain elastography, and to determine the level of interobserver agreement. Materials and Methods: We enrolled 83 patients with 83 breast masses identified on ultrasound and referred for biopsy. After B-mode ultrasound examination, the lesions were manually segmented by three radiologists with varying degrees of experience in breast imaging, designated reader 1 (R1, with 15 years), reader 2 (R2, with 2 years), and reader 3 (R3, with 8 years). Elastography was performed automatically on the best image with computer-aided diagnosis (CAD) software. Cutoff values of 70%, 75%, 80%, and 90% of hard areas were applied for determining the performance of the CAD software. The best cutoff value for the most experienced radiologists was then compared with the visual assessment. Interobserver agreement for the best cutoff value was determined, as were the interclass correlation coefficient and concordance among the radiologists for the areas segmented. Results: The best cutoff value of the proportion of hard area within a breast mass, for experienced radiologists, was found to be 75%. At a cutoff value of 75%, the interobserver agreement was excellent between R1 and R2, as well as between R1 and R3, and good between R2 and R3. The interclass concordance coefficient among the three radiologists was 0.950. When assessing the segmented areas by size, we found that the level of agreement was higher among the more experienced radiologists. Conclusion: The best cutoff value for a quantitative CAD system to classify breast masses was 75%.

Resumo Objetivo: Determinar o melhor valor de corte para classificar os nódulos mamários pela elastografia por ultrassom, usando um software dedicado para elastografia por deformação, e determinar o nível de concordância interobservadores. Materiais e Métodos: Foram incluídos no estudo 83 pacientes com 83 massas mamárias identificadas no ultrassom e encaminhados para biópsia. Após o exame ultrassonográfico no modo B, as lesões foram manualmente segmentadas por três radiologistas com diferentes graus de experiência em imagem da mama: leitor 1 (R1, com 15 anos de experiência), leitor 2 (R2, com 2 anos de experiência) e leitor 3 (R3, com 8 anos de experiência). A classificação pela elastografia foi realizada automaticamente com base na melhor imagem com o software diagnóstico auxiliado por computador (DAC). Valores de corte de 70%, 75%, 80% e 90% das áreas duras foram aplicados para determinar o desempenho do software DAC. O melhor valor de corte para os radiologistas foi comparado com a avaliação visual. A concordância interobservadores para o melhor valor de corte foi determinada, assim como o coeficiente de correlação interclasses e a concordância entre os radiologistas para as áreas segmentadas. Resultados: O melhor valor de corte da proporção de área dura dentro de um nódulo mamário foi de 75% para os radiologistas mais experientes. Com um valor de corte de 75%, a concordância interobservadores foi excelente entre R1 e R2 e entre R1 e R3, e boa entre R2 e R3. O coeficiente de concordância interclasses entre os três radiologistas foi de 0,950. Ao avaliar as áreas segmentadas por tamanho, constatamos que o nível de concordância foi maior entre os radiologistas mais experientes. Conclusão: O melhor valor de corte para um sistema quantitativo de DAC para classificar as massas mamárias foi de 75%.

Quantitative anaysis of echogenicity of ≤4 cm solid renal cell carcinoma with gray-scale ultrasound / 中国医学影像技术

Objective: To investigate the echo characteristics and the diagnostic efficacy of ≤ 4 cm solid renal cortical tumors with computer software. Methods: The gray-scale ultrasonic images of 205 solid renal cortical tumors (167 malignant tumors and 38 benign tumors) in 202 patients were retrospectively analyzed. The echogenicity of the tumors were quantitatively analyzed with computer software, and relative unit grayscale values were measured. The graysacle value of the tumor was compared with that of the renal cortex and sinus. The echogenicity of tumors were further semi-quantitatively classified as obvious hyperechoic, slightly hyperechoic or hypoechoic. Results: The relative unit grayscale value and the echo distribution had significantly different between benign and malignant tumors (both P<0.05). Slightly hyper echogenicity was the major appearance of malignant renal tumors (147/167, 88.02%),while some malignant tumors showed hypoechoic (18/167, 10.78%). The majority of benign renal tumors were obvious hyperechoic (24/38, 63.16%) and some of them represented as slightly hyperechoic (12/38, 31.58%). There were statistically significant differences of echo distributions between benign and malignant tumors. Taken slightly hyperechoic or hypoechoic as standards for differentiating magligant from benign tumors, the sensitivity was 98.80%, but the specificity was 63.16%. Taken the relative unit gray value <1.51 as standard for differentiating magligant and benign ones, the sensitivity was 75.45%, while the specificity was 73.68%. Conclusion: The degree of tumor echo can be accurately quantified with computer software. Most ≤ 4 cm solid renal cell carcinomas show slightly high echogenicity.

Texture features of hepatic parenchyma based on high-frequency ultrasonic images in quantitative evaluation of liver function in cirrhosis patients / 中国医学影像技术

Objective: To explore the application value of quantitative evaluation on liver function in patients with hepatitis B cirrhosis based on texture features of liver parenchyma from high frequency ultrasound images. Methods: Two-dimensional high-frequency ultrasonograms of liver parenchyma in 95 patients with different levels of liver function (37 cases in Child-Pugh A group, 33 in Child-Pugh B group and 25 cases in Child-Pugh C group) and 21 healthy volunteers (control group) were collected. Based on the idea of machine learning, the defect detection method of computer technology was used to locate the defect in liver parenchyma images, and the defect maps of liver parenchyma were obtained. Then three parameters, i.e. the maximum number per unit area of defect maps x(D), the mean value of defect maps mean (D) and the entropy of defect maps ε(D) were extracted. The differences of three parameters among different groups were compared. Results: There were significant differences of x(D), mean (D) and ε(D) among 4 groups (all P<0.001). The value of x(D) and mean (D) in Child-Pugh A group, B group and C group were higher than those in control group (all P<0.05). The value of mean (D) in Child-Pugh A group and Child-Pugh B group were higher than that in control group (both P<0.05). The value of mena (D) and ε(D) in Child-Pugh B group were higher than those in Child-Pugh C group (both P<0.05). Conclusion: Texture features analysis of liver parenchyma based on high-frequency ultrasonic images can accurately assess liver cirrhosis with Child-Pugh A and Child-Pugh B, while the evaluation of liver cirrhosis with Child-Pugh C still needs to combine with other features of the liver.

Preliminary clinical application of fluorescence microscopic imaging and computer-aided diagnosis system in the diagnosis of superficial cutaneous fungal infections / 中华皮肤科杂志

Objective@#To evaluate the accuracy of automated fluorescence microscopic imaging and computer-aided diagnosis system (AFMICADS) in the auxiliary diagnosis of superficial cutaneous fungal infections.@*Methods@#Totally, 106 outpatients and inpatients with suspected superficial fungal infections were enrolled from clinical departments of Union Hospital, Tongji Medical College, Huazhong University of Science and Technology between July 2018 and September 2018. A total of 126 specimens were collected, including 83 skin scales and 43 nail parings. Each specimen was divided into 3 groups to be examined by conventional fungal microscopy, culture with modified Sabouraud dextrose agar and fluorescence microscopy (artificial fluorescence microscopy and AFMICADS-based fluorescence microscopy) respectively. A positive result was defined as that conventional fungal microscopy and/or fungal culture was positive. Consistency rate, sensitivity and specificity of the 3 microscopic methods were calculated. Statistical analysis was carried out with SPSS 10.0 software by using McNemar test and Kappa test for analyzing difference in the positive rate, as well as consistency, between the 3 microscopic methods and the positive standard, and by using efficiency test for comparing the consistency rate among the 3 microscopic methods.@*Results@#Of 126 specimens, 124 (98.4%) were positive for artificial fluorescence microscopy, and 123 (97.6%) for AFMICADS-based fluorescence microscopy. Both positive rates of the above 2 microscopic methods were significantly higher than the positive rate of the positive standard (77.8%, both P < 0.001) . The sensitivity, specificity and consistency rate of AFMICADS-based fluorescence microscopy were 100%, 10.7% and 80.2% respectively, and those of artificial fluorescence microscopy were 100%, 7.1% and 79.4% respectively. Additionally, no significant difference in the consistency was observed between the AFMICADS-based and artificial fluorescence microscopy (P > 0.05) .@*Conclusion@#The accuracy of AFMICADS-based fluorescence microscopy in the diagnosis of superficial cutaneous fungal infections is similar to that of artificial fluorescence microscopy.

Correlation of quantitative parameters of dynamic enhanced magnetic resonance imaging with Dukes stage, lymph node metastasis and tumor differentiation degree of rectal cancer / 中国医师杂志

Objective@#To explore the correlation between quantitative parameters of dynamic enhanced magnetic resonance imaging (DCE-MRI) and Dukes stage, lymph node metastasis, tumor differentiation degree and molecular biological indicators [Ki67 and human epidermal growth factor receptor 2 (CerbB-2)] of rectal cancer.@*Methods@#This study was a prospective study. From October 2014 to October 2017, 168 cases of rectal cancer patients were selected as the research objects. DCE-MRI was performed preoperatively to obtain the quantitative parameter values of region of interest (ROI) [apparent diffusion coefficient (ADC)mean, Ktrans, Ve, and Kep] in tumor site. The expression of Ki67 and CerbB-2 were detected by immunohistochemical. Correlations of DCE-MRI quantitative parameter values and rectal cancer Dukes staging, lymph node metastasis, tumor differentiation degree to Ki67 and CerbB-2 expression level were analysised.@*Results@#With the increase of tumor differentiation, ADCmean was increasing, while Ktrans and Ve showed a downward trend, with significant difference (P<0.05). Among them, Ktrans and Ve in the high-differentiation group and the medium-differentiation group were significantly lower than those in the low-differentiation group (P<0.05), while there was no statistical significance between the high-differentiation group and the medium-differentiation group (P>0.05). The ADCmean was statistically significant between different groups (P<0.05), while Kep was not statistically significant (P>0.05). The ADCmean of lymph node metastasis group was significantly lower than that of no lymph node metastasis group (P<0.001), while Ktrans, Ve and Kep were significantly higher than that of no lymph node metastasis group (P<0.001). With the increase of Dukes staging in rectal cancer, ADCmean was decreasing and Ktrans was increasing, with statistically significant difference (P<0.001); among which the ADCmean of Dukes A and B was significantly higher than that of Dukes C and D (P<0.05), and Ktrans was significantly lower than that of Dukes C and D (P<0.05). Ktrans and Kep were increased with the increased expression levels of Ki67 and CerbB-2, and the difference between Ktrans and Kep was statistically significant (P<0.05). The ADCmean decreased with the increase of Ki67 and CerbB-2 expression level, and the difference was statistically significant (P<0.05).@*Conclusions@#DCE-MRI quantitatively participates can reflect the biological behavior of rectal cancer, and can be used to evaluate the prognosis of tumors and guide the clinical selection of more appropriate treatment options.

Artificial intelligence-based bone age assessment using deep learning of characteristic regions in digital hand radiograph / 中华放射学杂志

Objective@#To detect the feasibility and efficiency of bone age(BA) artificial intelligence(AI) estimation based on deep learning features from traditional regions of interest(ROI) in hand digital radiographs(DR).@*Methods@#BA dataset of left hand DR with 11 858 subjects aged from 0 to 18 years in Children′s Hospital of Shanghai were split to training(80.0%) and validation (20.0%) set in this study. An improved regression convolutional neural networks and extreme gradient boosting decision tree method were utilized for the BA analysis based on traditional ROIs in the images. Another set of BA data with 1 229 subjects also in the hospital was adopted for test. Mean average precision(mAP) and mean absolute error(MAE) were used to assess model accuracy of detection and BA prediction, respectively.@*Results@#The mAP of ROIs detection of the model was 0.91,and MAE of all male and female subjects was 0.461 and 0.431 years respectively in validation and test sets. The difference less than 1 year in test accounted for 90.07% between BA assessment of the model and of the peadiatric radiologists, with an accuracy rate of 96.67%.The difference over 1 year was 9.03% (with underestimation of 6.43% and overestimation of 2.60%), in which corresponding age data was of being less in training set or sesamoid nearby adductor pollicis or fusion of epiphysis appeared in test set.@*Conclusion@#An AI model based on deep learning of traditional ROIs′ features in hand DR images is initially achieved to automatically predict BA rapidly and effectively, yet it still needs further optimization.

Artificial intelligence?based bone age assessment using deep learning of characteristic regions in digital hand radiograph / 中华放射学杂志

s] Objective To detect the feasibility and efficiency of bone age(BA) artificial intelligence(AI) estimation based on deep learning features from traditional regions of interest(ROI) in hand digital radiographs(DR). Methods BA dataset of left hand DR with 11 858 subjects aged from 0 to 18 years in Children′s Hospital of Shanghai were split to training(80.0%) and validation (20.0%) set in this study. An improved regression convolutional neural networks and extreme gradient boosting decision tree method were utilized for the BA analysis based on traditional ROIs in the images. Another set of BA data with 1 229 subjects also in the hospital was adopted for test. Mean average precision(mAP) and mean absolute error(MAE) were used to assess model accuracy of detection and BA prediction, respectively. Results The mAP of ROIs detection of the model was 0.91,and MAE of all male and female subjects was 0.461 and 0.431 years respectively in validation and test sets. The difference less than 1 year in test accounted for 90.07% between BA assessment of the model and of the peadiatric radiologists, with an accuracy rate of 96.67%.The difference over 1 year was 9.03% (with underestimation of 6.43% and overestimation of 2.60%), in which corresponding age data was of being less in training set or sesamoid nearby adductor pollicis or fusion of epiphysis appeared in test set. Conclusion An AI model based on deep learning of traditional ROIs′features in hand DR images is initially achieved to automatically predict BA rapidly and effectively, yet it still needs further optimization.

Correlation of quantitative parameters of dynamic enhanced magnetic resonance imaging with Dukes stage, lymph node metastasis and tumor differentiation degree of rectal cancer / 中国医师杂志

Objective To explore the correlation between quantitative parameters of dynamic enhanced magnetic resonance imaging (DCE-MRI) and Dukes stage,lymph node metastasis,tumor differentiation degree and molecular biological indicators [Ki67 and human epidermal growth factor receptor 2 (CerbB-2)] of rectal cancer.Methods This study was a prospective study.From October 2014 to October 2017,168 cases of rectal cancer patients were selected as the research objects.DCE-MRI was performed preoperatively to obtain the quantitative parameter values of region of interest (ROI) [apparent diffusion coefficient (ADC) Ktrans,Ve,and Kep] in tumor site.The expression of Ki67 and CerbB-2 were detected by immunohistochemical.Correlations of DCE-MRI quantitative parameter values and rectal cancer Dukes staging,lymph node metastasis,tumor differentiation degree to Ki67 and CerbB-2 expression level were analysised.Results With the increase of tumor differentiation,ADCmean was increasing,while Ktrans and Ve showed a downward trend,with significant difference (P ＜ 0.05).Among them,Ktrans and Ve in the high-differentiation group and the medium-differentiation group were significantly lower than those in the low-differentiation group (P ＜ 0.05),while there was no statistical significance between the high-differentiation group and the medium-differentiation group (P ＞ 0.05).The ADC was statistically significant between different groups (P ＜ 0.05),while Kep was not statistically significant (P ＞ 0.05).The ADCmean of lymph node metastasis group was significantly lower than that of no lymph node metastasis group (P ＜0.001),while Ktrans,Ve and Kep were significantly higher than that of no lymph node metastasis group (P ＜0.001).With the increase of Dukes staging in rectal cancer,ADCmean was decreasing and Ktrans was increasing,with statistically significant difference (P ＜0.001);among which the ADCmean of Dukes A and B was significantly higher than that of Dukes C and D (P ＜ 0.05),and Ktrans was significantly lower than that of Dukes C and D (P ＜ 0.05).Ktrans and Kep were increased with the increased expression levels of Ki67 and CerbB-2,and the difference between Ktrans and Kep was statistically significant (P ＜ 0.05).The ADCmean decreased with the increase of Ki67 and CerbB-2 expression level,and the difference was statistically significant (P ＜ 0.05).Conclusions DCE-MRI quantitatively participates can reflect the biological behavior of rectal cancer,and can be used to evaluate the prognosis of tumors and guide the clinical selection of more appropriate treatment options.

Application and development of artificial intelligence in dermatology / 中华皮肤科杂志

Artificial intelligence,known as the science and engineering of manufacturing intelligent machines,is one of the frontier science and technology that affects the development of modern society.With the progress of computer science and internet technology,artificial intelligence has gradually formed 3 main branches:cognitive computing,machine learning and deep learning.In recent years,the cooperation between artificial intelligence and medicine has been applied in many aspects,such as image recognition,assistant diagnosis,medical robots,drug research and development.Furthermore,characteristics of dermatology and advantages of artificial intelligence in image recognition make the application of artificial intelligence in dermatology become a focus nowadays.The authors summarize the application scope of artificial intelligence in dermatology,such as skin imaging,skin pathology and medical robots,analyze the current situation of artificial intelligence application in dermatology,and make prospects for the trend of development in the future.

Overview of Deep Learning in Gastrointestinal Endoscopy

Artificial intelligence is likely to perform several roles currently performed by humans, and the adoption of artificial intelligence-based medicine in gastroenterology practice is expected in the near future. Medical image-based diagnoses, such as pathology, radiology, and endoscopy, are expected to be the first in the medical field to be affected by artificial intelligence. A convolutional neural network, a kind of deep-learning method with multilayer perceptrons designed to use minimal preprocessing, was recently reported as being highly beneficial in the field of endoscopy, including esophagogastroduodenoscopy, colonoscopy, and capsule endoscopy. A convolutional neural network-based diagnostic program was challenged to recognize anatomical locations in esophagogastroduodenoscopy images, Helicobacter pylori infection, and gastric cancer for esophagogastroduodenoscopy; to detect and classify colorectal polyps; to recognize celiac disease and hookworm; and to perform small intestine motility characterization of capsule endoscopy images. Artificial intelligence is expected to help endoscopists provide a more accurate diagnosis by automatically detecting and classifying lesions; therefore, it is essential that endoscopists focus on this novel technology. In this review, we describe the effects of artificial intelligence on gastroenterology with a special focus on automatic diagnosis, based on endoscopic findings.

Preliminary clinical application of fluorescence microscopic imaging and computer-aided diagnosis system in the diagnosis of superficial cutaneous fungal infections / 中华皮肤科杂志

Objective To evaluate the accuracy of automated fluorescence microscopic imaging and computer-aided diagnosis system (AFMICADS) in the auxiliary diagnosis of superficial cutaneous fungal infections.Methods Totally,106 outpatients and inpatients with suspected superficial fungal infections were enrolled from clinical departments of Union Hospital,Tongji Medical College,Huazhong University of Science and Technology between July 2018 and September 2018.A total of 126 specimens were collected,including 83 skin scales and 43 nail parings.Each specimen was divided into 3 groups to be examined by conventional fungal microscopy,culture with modified Sabouraud dextrose agar and fluorescence microscopy (artificial fluorescence microscopy and AFMICADS-based fluorescence microscopy) respectively.A positive result was defined as that conventional fungal microscopy and/or fungal culture was positive.Consistency rate,sensitivity and specificity of the 3 microscopic methods were calculated.Statistical analysis was carried out with SPSS 10.0 software by using McNemar test and Kappa test for analyzing difference in the positive rate,as well as consistency,between the 3 microscopic methods and the positive standard,and by using efficiency test for comparing the consistency rate among the 3 microscopic methods.Results Of 126 specimens,124 (98.4％) were positive for artificial fluorescence microscopy,and 123 (97.6％) for AFMICADS-based fluorescence microscopy.Both positive rates of the above 2 microscopic methods were significantly higher than the positive rate of the positive standard (77.8％,both P ＜ 0.001).The sensitivity,specificity and consistency rate of AFMICADS-based fluorescence microscopy were 100％,10.7％ and 80.2％ respectively,and those of artificial fluorescence microscopy were 100％,7.1％ and 79.4％ respectively.Additionally,no significant difference in the consistency was observed between the AFMICADS-based and artificial fluorescence microscopy (P ＞0.05).Conclusion The accuracy of AFMICADS-based fluorescence microscopy in the diagnosis of superficial cutaneous fungal infections is similar to that of artificial fluorescence microscopy.