[A convolutional neural network based model for assisting pathological diagnoses on thyroid liquid-based cytology].

Objective: To develop a convolutional neural network based model for assisting pathological diagnoses on thyroid liquid-based cytology specimens. Methods: Seven-hundred thyroid TCT slides were collected, scanned for whole slide imaging (WSI), and divided into training and test sets after labeling the correct diagnosis (benign versus malignant). The extracted regions of interest after noise filtering were cropped into pieces of 512 × 512 patch on 10 × and 40 × magnifications, respectively. A classification model was constructed using deeply learning algorithms, and applied to the training set, then automatically tuned in the test set. After data enhancement and parameters optimization, accuracy, sensitivity, specificity, positive predictive value and negative predictive value of the model were calculated. Results The training set with 560 WSI contained 4 926 cell clusters (11 164 patches), while the test set with 140 WSI contained 977 cell clusters (1 402 patches). YOLO network was selected to establish a detection model, and ResNet50 was used as a classification model. With 40 epochs training, results from 10× magnifications showed an accuracy of 90.01%, sensitivity of 89.31%, specificity of 92.51%, positive predictive value of 97.70% and negative predictive value of 70.82%. The area under curve was 0.97. The average diagnostic time was less than 1 second. Although the model for data of 40× magnifications was very sensitive (98.72%), but its specificity was poor, suggesting that the model was more reliable at 10× magnification. Conclusions: The performance of a deep-learning based model is equivalent to pathologists' diagnostic performance, but its efficiency is far beyond. The model can greatly improve consistency and efficiency, and reduce the missed diagnosis rate. In the future, larger studies should have more morphology diversity, improve model's accuracy and eventually develop a model for direct clinical use.

Polymorphism of the melatonin receptor genes and its relationship with seasonal reproduction in the Gulin Ma goat breed.

Melatonin is thought to be the main molecule that transmits the signal of seasonal change to the neuroendocrine system in seasonal breeding species. Melatonin exerts its effects through specific melatonin receptors, MTNR1A and MTNR1B. In the present study, six native goat breeds in China and one introduced goat breed were analysed to investigate the relationship between the genetic polymorphism of receptor genes and seasonal reproduction. Sequencing results showed that there were five polymorphic mutations in the MTNR1A gene and two in the MTNR1B gene. In the MTNR1A gene, genotypes AA, AB and BB for 424C>T and genotypes CC, CD and DD for 589C>A were observed in these goat breeds. In all six native goat breeds, only genotype AA was detected. In the MTNR1B gene, genotypes EE, EF and FF for 1179G>A and genotypes GG, GH and HH for 1529A>G were detected. However, in Gulin Ma goats, the genotypes EE and HH were not found. Moreover, the base of G at position 1179 and A at position 1529 were linked (By Arlequin ver 3.1, Zoological Institute, Berne, Switzerland, http://cmpg.unibe.ch/software/arlequin3,D' = 0.7496, r(2)  = 0.4421, χ(2)  = 489.8679, p = 0.000). Among these mutations, no amino acid change was found in MTNR1A, while both of the mutations in MTNR1B gene caused amino acid changes of R222H and S339G, respectively. The structural analysis showed that the R222H mutation occurred in the first amino acid residue of the third cytoplasmic loop, and the S339G mutation was located in the carboxyl terminus of the protein. In terms of seasonal breeding, all the genotypes we detected showed a similar kidding frequency distribution trend with a higher frequency in May-August than in January-April and in September-December. This suggests that the relationship between the polymorphisms in the MTNR1A and MTNR1B genes and seasonal breeding could not be established.

Building a disease risk model of osteoporosis based on traditional Chinese medicine symptoms and western medicine risk factors.

In the Traditional Chinese Medicine (TCM) cross-sectional survey conducted by our team, we were interested in determining the risk factors of osteoporosis. To analyze this TCM study, we had to deal with three statistical problems: (1) a very large number of potential risk factors, (2) interactions among potential risk factors, and (3) nonlinear effects of some continuous-scale risk factors. To address these analytic issues, we used two data mining methods, support vector machine recursive feature elimination and random forest; to deal with the curse of high-dimensional risk factors, we applied another data mining technique of association rule learning to discover the potential associations among risk factors. Finally, we employed the generalized partial linear model (GPLM) to determine nonlinear effects of an important continuous-scale risk factor. The final GPLM model shows that TCM symptoms play an important role in assessing the risk of osteoporosis. The GPLM also reveals a nonlinear effect of the important risk factor, menopause years, which might be missed by the generalized linear model.

Characteristics of Crimean-Congo hemorrhagic fever virus (Xinjiang strain) in China.

Virus strains isolated from blood of patients during a hemorrhagic fever outbreak in 1968 in southern Xinjiang, China, from Hyalomma asiaticum and from sheep, were found to be identical or closely related to Crimean-Congo hemorrhagic fever (C-CHF) virus by complement fixation and indirect immunofluorescence tests with convalescent sera of patients and with C-CHF reference antibody. The virus was inactivated by ether and acid. Viral synthesis was not suppressed by 5-iododeoxyuridine suggesting an RNA-containing genome. The buoyant density in sucrose was 1.16-1.18 g/cm3. The particle weight was estimated at 3.26 +/- 0.46 X 10(8). The diameter of the virus particles was 85-105 nm.

[An evaluation of urinary Tamm-Horsfall protein coating cell in the diagnosis of urologic neoplasms].

By utilizing enzymic immunoassay, in a total of 161 cases of urologic tumors, Tamm-Horsfall Protein Coating Cells (THPC) could be found uniquely in the urines of renal tumors, whereas was absent in those of normal control as well as in the urines of transitional cell carcinomas of pelvis, ureter and bladder. Renal cell carcinoma predominated over in the THPC positive cases, among which granular cell carcinoma demonstrated a higher positive rate than that of clear cell type, 40% versus 15%. THP antigen could be found in the 66.6% frozen sections of renal cell carcinoma and the THP concentration in urine were also higher. Therefore, the detection of THP coating cells in urine of urologic neoplasm may be useful in evaluating the renal origin of the exfoliated tumor cells and elevating the screening accuracy of renal cell carcinoma, with no harms to the patients and no special equipment.

Cellular origin of renal cell carcinoma--an immunohistological study on monoclonal antibodies.

It has been recognized in recent years that renal cell carcinoma possesses a proximal convoluted tubules origin. In an attempt of further intensive evaluation we assayed 51 renal cell carcinomas and 38 additional normal kidney specimens with 4 kidney segment-specific antibodies (Uro-2/S4, Uro-5/T16, Uro-10/T43, Anti Tamm-Horsfall Protein) directed against the nephrotic cells by indirect immunohistological techniques. Consistently stable staining was developed in the tests. The results showed that the majority of renal clear cell carcinomas expressed nephrogenous properties of proximal convoluted tubules but in part also of distal tubules and collecting ducts. Granular cell carcinoma mostly demonstrated histogenic appearance of distal convoluted tubules.