Deep convolutional neural network model ResNeSt for discrimination of papillary thyroid carcinomas and benign nodules in thyroid nodules diagnosed as atypia of undetermined significance.

Background: A deep convolutional neural network (DCNN) model was employed for the differentiation of thyroid nodules diagnosed as atypia of undetermined significance (AUS) according to the 2023 Bethesda System for Reporting Thyroid Cytopathology (TBSRTC). The aim of this study was to investigate the efficiency of ResNeSt in improving the diagnostic accuracy of fine-needle aspiration (FNA) biopsy. Methods: Fragmented images were used to train and test DCNN models. A training dataset was built from 1,330 samples diagnosed as papillary thyroid carcinoma (PTC) or benign nodules, and a test dataset was built from 173 samples diagnosed as AUS. ResNeSt was trained and tested to provide a differentiation. With regard to AUS samples, the characteristics of the cell nuclei were compared using the Wilcoxon test. Results: The ResNeSt model achieved an accuracy of 92.49% (160/173) on fragmented images and 84.78% (39/46) from a patient wise viewpoint in discrimination of PTC and benign nodules in AUS nodules. The sensitivity and specificity of ResNeSt model were 95.79% and 88.46%. The κ value between ResNeSt and the pathological results was 0.847 (P<0.001). With regard to the cell nuclei of AUS nodules, both area and perimeter of malignant nodules were larger than those of benign ones, which were 2,340.00 (1,769.00, 2,807.00) vs. 1,941.00 (1,567.50, 2,455.75), P<0.001 and 190.46 (167.64, 208.46) vs. 171.71 (154.95, 193.65), P<0.001, respectively. The grayscale (0 for black, 255 for white) of malignant lesions was lower than that of benign ones, which was 37.52 (31.41, 46.67) vs. 45.84 (31.88, 57.36), P <0.001, indicating nuclear staining of malignant lesions were deeper than benign ones. Conclusions: In summary, the DCNN model ResNeSt showed great potential in discriminating thyroid nodules diagnosed as AUS. Among those nodules, malignant nodules showed larger and more deeply stained nuclei than benign nodules.

Prediction of tumor origin in cancers of unknown primary origin with cytology-based deep learning.

Cancer of unknown primary (CUP) site poses diagnostic challenges due to its elusive nature. Many cases of CUP manifest as pleural and peritoneal serous effusions. Leveraging cytological images from 57,220 cases at four tertiary hospitals, we developed a deep-learning method for tumor origin differentiation using cytological histology (TORCH) that can identify malignancy and predict tumor origin in both hydrothorax and ascites. We examined its performance on three internal (n = 12,799) and two external (n = 14,538) testing sets. In both internal and external testing sets, TORCH achieved area under the receiver operating curve values ranging from 0.953 to 0.991 for cancer diagnosis and 0.953 to 0.979 for tumor origin localization. TORCH accurately predicted primary tumor origins, with a top-1 accuracy of 82.6% and top-3 accuracy of 98.9%. Compared with results derived from pathologists, TORCH showed better prediction efficacy (1.677 versus 1.265, P < 0.001), enhancing junior pathologists' diagnostic scores significantly (1.326 versus 1.101, P < 0.001). Patients with CUP whose initial treatment protocol was concordant with TORCH-predicted origins had better overall survival than those who were administrated discordant treatment (27 versus 17 months, P = 0.006). Our study underscores the potential of TORCH as a valuable ancillary tool in clinical practice, although further validation in randomized trials is warranted.

A multicenter prospective study of lateral neck lymph node mapping in papillary thyroid cancer.

Background: Despite the high incidence of lateral neck lymph node (LN) metastasis in papillary thyroid cancer (PTC), the management of the lateral neck remains controversial. We aimed to map the draining LNs in the lateral neck using carbon nanoparticles and explore its potential in neck evaluation. Methods: We conducted a multicenter, prospective study in PTC patients who had non-palpable yet suspicious metastatic lateral LNs on ultrasound and/or computed tomography (CT) but could not be confirmed by fine needle aspiration. Carbon nanoparticle suspension was injected peritumorally into the thyroid and modified lateral neck dissection was subsequently performed. Results: A total of 154 patients were enrolled for analysis. And 5,070 lateral LNs were removed, of which 1,079 (21.3%) were dyed. The median of dyed LNs was 6 per case (range, 1-33). The distribution of dyed LNs in neck compartments was IV > III > IIA > IIB/V, independent of tumor size, location, multifocality or microscopic extra-thyroidal extension (ETE). Compared with undyed LNs, the probabilities of metastasis in dyed LNs were significantly increased in compartment III, IV, V, and II-V (III: 29.3% vs. 15.4%, P<0.001; IV: 26.3% vs. 14.5%, P<0.001; V: 16.7% vs. 3.3%, P=0.005; II-V: 26.3% vs. 10.0%, P<0.001). The relative risks of metastasis in dyed LNs compared with undyed LNs were 1.90, 1.82, 5.04 and 2.62 in compartment III, IV, V, and II-V, respectively. Conclusions: It was the first prospective multicenter study to map the lateral neck LNs with carbon nanoparticles, which could help surgeons visualize the suspicious LNs during surgery. Instead of unguided LN biopsy, this method has a potential role in lateral neck assessment for indeterminate lateral LNs in PTC.

HIF-1α drives resistance to ferroptosis in solid tumors by promoting lactate production and activating SLC1A1.

Solid tumors have developed robust ferroptosis resistance. The mechanism underlying ferroptosis resistance regulation in solid tumors, however, remains elusive. Here, we report that the hypoxic tumor microenvironment potently promotes ferroptosis resistance in solid tumors in a hypoxia-inducible factor 1α (HIF-1α)-dependent manner. In combination with HIF-2α, which promotes tumor ferroptosis under hypoxia, HIF-1α is the main driver of hypoxia-induced ferroptosis resistance. Mechanistically, HIF-1α-induced lactate contributes to ferroptosis resistance in a pH-dependent manner that is parallel to the classical SLC7A11 and FSP1 systems. In addition, HIF-1α also enhances transcription of SLC1A1, an important glutamate transporter, and promotes cystine uptake to promote ferroptosis resistance. In support of the role of hypoxia in ferroptosis resistance, silencing HIF-1α sensitizes mouse solid tumors to ferroptosis inducers. In conclusion, our results reveal a mechanism by which hypoxia drives ferroptosis resistance and identify the combination of hypoxia alleviation and ferroptosis induction as a promising therapeutic strategy for solid tumors.

Assessment of functional prognosis of anterior cruciate ligament reconstruction in athletes based on a body shape index.

BACKGROUND: A healthy body shape is essential to maintain athletes' sports level. At present, little is known about the effect of athletes' body shape on anterior cruciate ligament reconstruction (ACLR). Moreover, the relationship between body shape and variables such as knee joint function after operation and return to the field has not been well studied. AIM: To verify the relationship between a body shape index (ABSI) and the functional prognosis of the knee after ACLR in athletes with ACL injuries. METHODS: We reviewed 76 athletes with unilateral ACL ruptures who underwent ACLR surgery in the First Hospital of Shanxi Medical University between 2017 and 2020, with a follow-up period of more than 24 mo. First, all populations were divided into a High-ABSI group (ABSI > 0.835, n = 38) and a Low-ABSI group (ABSI < 0.835, n = 38) based on the arithmetic median (0.835) of ABSI values. The primary exposure factor was ABSI, and the outcome indicators were knee function scores as well as postoperative complications. The correlation between ABSI and postoperative knee function scores and postoperative complications after ACLR were analyzed using multifactorial logistic regression. RESULTS: The preoperative knee function scores of the two groups were similar. The surgery and postoperative rehabilitation exercises, range of motion (ROM) compliance rate, Lysholm score, and Knee Injury and Osteoarthritis Outcome Score of the two groups gradually increased, whereas the quadriceps atrophy index gradually decreased. The knee function scores were higher in the Low-ABSI group than in the High-ABSI group at the 24-mo postoperative follow-up (P < 0.05). In multifactorial logistic regression, ABSI was a risk factor of low knee joint function score after surgery, specifically low ROM scores (odds ratio [OR] = 1.31, 95% confidence interval [CI] [1.10-1.44]; P < 0.001), low quadriceps atrophy index (OR = 1.11, 95%CI [0.97-1.29]; P < 0.05), low Lysholm scores (OR = 2.34, 95%CI [1.78-2.94]; P < 0.001), low symptoms (OR = 1.14, 95%CI [1.02-1.34]; P < 0.05), low activity of daily living (OR = 1.34, 95%CI [1.18-1.65]; P < 0.05), low sports (OR = 2.47, 95%CI [1.78-2.84]; P < 0.001), and low quality of life (OR = 3.34, 95%CI [2.88-3.94]; P < 0.001). ABSI was also a risk factor for deep vein thrombosis of the lower limb (OR = 2.14, 95%CI [1.88-2.36], P < 0.05] and ACL recurrent rupture (OR = 1.24, 95%CI [0.98-1.44], P < 0.05) after ACLR. CONCLUSION: ABSI is a risk factor for the poor prognosis of knee function in ACL athletes after ACLR, and the risk of poor knee function after ACLR, deep vein thrombosis of lower limb, and ACL recurrent rupture gradually increases with the rise of ABSI.

ELISA-like QDB method to meet the emerging need of Her2 assessment for breast cancer patients.

Inherent issues of subjectivity and inconsistency have long plagued immunohistochemistry (IHC)-based Her2 assessment, leading to the repeated issuance of guidelines by the American Society of Clinical Oncology/College of American Pathologists (ASCO/CAP) for its standardization for breast cancer patients. Yet, all these efforts may prove insufficient with the advent of Trastuzumab deruxtecan (T-Dxd), a drug with the promise to expand to tumors traditionally defined as Her2 negative (Her2-). In this study, we attempted to address these issues by exploring an ELISA-like quantitative dot blot (QDB) method as an alternative to IHC. The QDB method has been used to measure multiple protein biomarkers including ER, PR, Ki67, and cyclin D1 in breast cancer specimens. Using an independent cohort (cohort 2) of breast cancer formalin-fixed paraffin-embedded (FFPE) specimens, we validated cutoffs developed in cohort 1 (Yu et al., Scientific Reports 2020 10:10502) with overall 100% specificity (95% CI: 100-100) and 97.56% sensitivity (95% CI: 92.68-100) in cohort 2 against standard practice with the dichotomized absolutely quantitated values. Using the limit of detection (LOD) of the QDB method as the putative cutoff point, tumors with no Her2 expression were identified with the number comparable to those of IHC 0. Our results support further evaluation of the QDB method as an alternative to IHC to meet the emerging need of identifying tumors with low Her2 expression (Her2-low) in daily clinical practice.

SIRT6 drives sensitivity to ferroptosis in anaplastic thyroid cancer through NCOA4-dependent autophagy.

The sirtuin family has been reported to participate in the regulation of oxidative stress, cancer metabolism, aging, and so on. However, few studies have demonstrated its role in ferroptosis. Our previous studies confirmed that SIRT6 is upregulated in thyroid cancer and associated with cancer development by regulating glycolysis and autophagy. In this research, we aimed to elucidate the association between SIRT6 and ferroptosis. RSL3, erastin, ML210, and ML162 were applied to induce ferroptosis. Cell death and lipid peroxidation were measured by flow cytometry. We found that overexpression of SIRT6 significantly increased the sensitivity of cells to ferroptosis, whereas knockout of SIRT6 promoted resistance to ferroptosis. Furthermore, we demonstrated that SIRT6 induced NCOA4-dependent autophagic degradation of ferritin, thus driving sensitivity to ferroptosis. The clinically used ferroptosis inducer sulfasalazine showed promising therapeutic effects on SIRT6-upregulated thyroid cancer cells in vivo. In conclusion, our research demonstrated SIRT6-driven sensitivity to ferroptosis via NCOA4-dependent autophagy and proposed ferroptosis inducers as promising therapeutic agents for anaplastic thyroid cancer patients.

Quaterization Derivatization with Bis(Pyridine) Iodine Tetrafluoroboride: High-Sensitivity Mass Spectrometric Analysis of Unsaturated Fatty Acids in Human Thyroid Tissues.

Accurate quantification of disease-related unsaturated fatty acids (UFAs) in biomedical samples plays an important role in clinical diagnosis. Here, we reported a quaterization derivatization-stable isotope labeling strategy for accurate quantitative analysis of UFAs by high-performance liquid chromatography-mass spectrometry. [d0]/[d10]-Bis(pyridine) iodine tetrafluoroboride ([d0]/[d10]-IPy2BF4) was employed as the carbon-carbon double bond derivatization reagent with high efficiency and high specificity, to introduce a charge tag on UFAs and avoid the interference of saturated fatty acids. After labeling, the detection sensitivity was significantly enhanced by up to three orders of magnitude compared to intact UFAs. The standard curves showed good linearity (R2 > 0.999) over a wide concentration range. This strategy was successfully applied to determine the content of 12 UFAs in human thyroid carcinoma and para-carcinoma tissues. A significant difference was found in the content of several UFAs between these two kinds of tissues (p < 0.05). These results indicated that the proposed strategy may be valuable for the discovery of abnormal UFA content in early clinical diagnosis.

Chloramine-T-Enabled Mass Spectrometric Analysis of C═C Isomers of Unsaturated Fatty Acids and Phosphatidylcholines in Human Thyroids.

Specific locations of carbon-carbon double bonds (C═C) in lipids often play an essential role in biological processes, and there has been a booming development in C═C composition analysis by mass spectrometry. However, a universal derivatization and fragmentation pattern for the annotation of C═C positions in lipids is still challenging and attractive. To expand this field in lipidomics, a flexible and convenient N-tosylaziridination method was developed, with high derivatization efficiency, sensitivity, and specificity. The derivatization was very fast (15 s), and C═C numbers as well as locations could be pinpointed specifically in tandem mass spectra. By qualitative and quantitative studies of paratumor and tumor thyroid tissues of human beings, the total content of unsaturated fatty acids was suggested to be increased in tumor tissues, and good correlations in and between lysophosphatidylcholines and phosphatidylcholines were revealed by Spearman analysis. Further studies of C═C isomers showed that n-6/n-3 ratios were closely associated with human thyroid tumorigenesis, and high ratios of n-6/n-3 isomers seemed to suffer a high risk of carcinogenesis. Other isomers were not very representative; however, C═C in n-9/n-7 could also be significant for oncology research. Generally, it is supposed that both total amounts and C═C isomer ratios were related to cancer, and N-tosylaziridine derivatization could provide an alternative strategy for the C═C isomer study of disease models.

Identification of differentially expressed genes and signaling pathways in papillary thyroid cancer: a study based on integrated microarray and bioinformatics analysis.

BACKGROUND: The techniques of DNA microarray and bioinformatic analysis have exhibited efficiency in identifying dysregulated gene expression in human cancers. In this study, we used integrated bioinformatics analysis to improve our understanding of the pathogenesis of papillary thyroid cancer (PTC). METHODS: In this study, we integrated four Gene Expression Omnibus (GEO) datasets, GSE33630, GSE35570, GSE60542 and GSE29265, including 136 normal samples and 157 PTC specimens. The contents of the four datasets are based on GPL570, an Affymetrix Human Genome U133 Plus 2.0 array. Gene ontology (GO) analysis was used to identify characteristic the biological attributes of differentially expressed genes (DEGs) between PTC and normal samples. GO annotation was performed on the DEGs obtained, and the process relied on the DAVID online tool. Kyoto Encyclopedia of Genes and Genomes (KEGG) approach enrichment analyses were adopted to obtain the basic functions of the DEGs. The KOBAS online analysis database was used to complete DEG KEGG pathway comparison and analysis. The search tool (STRING) database was mainly used to search for interacting genes and complete the construction of protein-protein interaction (PPI) networks. RESULTS: Five hundred-ninety DEGs were consistently expressed in the four datasets; 327 of them were upregulated, while 263 were downregulated. Ten DEGs, including five upregulated (ENTPD1, THRSP, KLK10, ADAMTS9, MIR31HG) and five downregulated (SCARA5, EPHB1, CHRDL1, LOC440934, FOXP2) genes, were randomly selected for q-PCR in our own tissue samples to validate the integrated data. The most highly enriched GO terms were extracellular exosome (GO:0070062), cell adhesion (GO:0070062), positive regulation of gene expression (GO:0010628), and extracellular matrix (ECM) organization (GO:0030198). KEGG pathway analysis was performed, and it was found that abnormally expressed genes effectively participated in pathways such as tyrosine metabolism, complement and coagulation cascades, cell adhesion molecules (CAMs), transcriptional misregulation and ECM-receptor interaction pathways. CONCLUSIONS: Five hundred-ninety DEGs were identified in PTC by integrated microarray analysis. The GO and KEGG analyses presented here suggest that the DEGs were enriched in extracellular exosome, tyrosine metabolism, CAMs, complement and coagulation cascades, transcriptional misregulation and ECM-receptor interaction pathways. Functional studies of PTC should focus on these pathways.

A Comparison of Two Operation Methods Revealed the Risk Factors and the Necessity of LN-prRLN Dissection in Papillary Thyroid Carcinoma: A Retrospective Cohort Study in FUSCC.

BACKGROUND: Although the American Thyroid Association (ATA) guidelines indicate that central lymph nodes posterior to the right recurrent laryngeal nerve (LN-prRLN) should be routinely dissected, pr-RLN dissection is often neglected due to the high risk of injury to the recurrent laryngeal nerve (RLN). The purpose of this study was to investigate the risk factors associated with LN-prRLN metastasis in patients with papillary thyroid carcinoma (PTC) by preoperative examination and the indications for LN-prRLN dissection. METHODS: A total of 1487 consecutive patients with PTC who underwent total thyroidectomy or right lobectomy plus isthmic resection with central LN dissection (CLND) were divided into two groups: patients with LN-prRLN dissection (group A) and patients without LN-prRLN dissection (group B). Clinicopathologic data were reviewed of the patients who were operated on by the same thyroid surgery team in the Department of Head Neck Surgery, Fudan University Shanghai Cancer Center (FUSCC) between August 2011 and May 2019. The relationships of LN-prRLN metastasis with clinicopathologic characteristics were analyzed by univariate and multivariate logistic regression. RESULTS: The incidence of LN-prRLN metastasis was 34.1% (129/378). Univariate analysis showed that sex (P ≤ 0.001), tumor size (P ≤ 0.001), extrathyroidal extension (P=0.002), concurrent Hashimoto's thyroiditis (P=0.009), cLNMa (central lymph nodes anterior to the right recurrent laryngeal nerve) (P ≤ 0.001), cLNMa number (P ≤ 0.001), and lateral LN metastasis (LLNM) (P ≤ 0.001) were significantly associated with LN-prRLN metastasis in PTC. Multivariate logistic regression analysis revealed that tumor size (P=0.039), cLNMa (P=0.001), and LLNM (P=0.025) were independent risk factors for LN-prRLN metastasis in patients with PTC. Although there was no significant difference between the two groups in recurrence, we found that 4 cases relapsed in the LN-prRLN compartment in group B, while none relapsed in group A. CONCLUSION: LN-prRLN metastasis is often identified in patients with PTC. Patients with large tumor sizes, cLNMa and LLNM are at a high risk of LN-prRLN metastasis and should be recommended for careful LN-prRLN dissection.

Follicular lymphoma with paraneoplastic pemphigus as the first symptom: a case report and review of the literature.

Paraneoplastic pemphigus (PNP) is an autoimmune bullous dermatosis associated with tumors, first described by Anhalt et al. in 1990. Reports of paraneoplastic pemphigus complicated by follicular lymphoma (FL) are rare in the medical literature. Here, we retrospectively analyze a case of PNP accompanied by FL. The patient was a 54-year-old woman who suffered from PNP associated with FL at the beginning. She had received a pathological diagnosis and was treated with R-CHOP and other drugs. Her mucosal lesions and cutaneous lesions improved, and the FL was in remission. Eleven months later, she died of BO after receiving the diagnosis of PNP. We also review most of the studies and reports about PNP accompanied by FL. We list the clinicopathologic features, therapeutic schedule, and prognosis in order to improve hematologists' understanding and treatment of the diseases.

Simultaneous Analysis of Fatty Alcohols, Fatty Aldehydes, and Sterols in Thyroid Tissues by Electrospray Ionization-Ion Mobility-Mass Spectrometry Based on Charge Derivatization.

In this work, we developed a rapid and high-sensitivity method for simultaneous analyses of fatty alcohols, fatty aldehydes, and sterols by combining the optimized derivatization reaction with electrospray ionization-ion mobility-mass spectrometry (ESI-IM-MS). Pyridine and thionyl chloride were used as derivatization reagents as they were easily removed after the derivatization reaction and could generate permanently charged tags on different functional groups including hydroxyls and aldehydes. Through this one-step derivatization reaction, the sensitivity of detection for fatty alcohols, fatty aldehydes, and sterols was significantly increased. Moreover, the introduction of ion mobility spectrometry (IMS), offering an additional resolution power, ensured more sensitive and accurate detection of derivative products without increasing analytical time. Being connected with high-performance liquid chromatography, more than 15 kinds of compounds were analyzed within 4 min. Relative quantification using peak intensity ratios between d0-/d5-labeled ions were subsequently applied for analyzing these 15 kinds of compounds in human thyroid carcinoma and para-carcinoma tissues. The results showed significant differences in content of some analytes between these two kinds of tissues (p < 0.05). The correlations between most of the analytes in thyroid carcinoma tissues are better than the correlations in para-carcinoma tissues.

Discordant lymphoma consisting of mantle cell lymphoma and angioimmunoblastic T cell lymphoma: homology or heterogeneity?

OBJECTS: To investigate the pathologic characteristic of discordant lymphoma with mantle cell lymphoma and angioimmunoblastic T cell lymphoma. METHODS: The clinicopathologic data of cases of discordant lymphoma were organized and clinicopathologic features were analyzed by literature review. RESULTS: A 49-year-old male was taken to the hospital due to the lymphandenopathy in January 2007 and mantle cell lymphoma was diagnosed in the pathology report. EBV-EBER staining was negative. Active chemotherapy was received and the patient achieved complete response. Seven years later since diagnosis, in 2014 scattered rashes were found. A skin biopsy was taken and the result was not mantle cell lymphoma but angioimmunoblastic T cell lymphoma. EBV-EBER positivity was detected. Clonal T cell receptor gamma locus gene rearrangements were detected while no clonal immunoglobin heavy locus gene rearrangement was detected in the skin sample. CONCLUSIONS: This is the first report on discordant lymphoma consisting of mantle cell lymphoma and angioimmunoblastic T cell lymphoma. There seems to be no relation these two different kinds of lymphoma, and EBV infection might prompt the development of angioimmunoblastic T cell lymphoma after transplantation. Rash is a common clinical manifestation when T cell lymphoma develops after treatment for MCL.

Tripartite motif containing 14: An oncogene in papillary thyroid carcinoma.

Tripartite motif (TRIM)-containing protein 14 (TRIM14) is implicated in many malignancies. Presently, we studied whether TRIM14 played a role in human papillary thyroid carcinoma (PTC). Herein, TRIM14 was up-regulated in tumor tissues when compared with normal thyroid samples. Among PTC patients, enhanced TRIM14 predicted short recurrence free survival (RFS) time. Then, we found that knockdown of TRIM14 in human PTC K1 cells inhibited cell proliferation, induced cell apoptosis and prevented the tumorigenicity in nude mice. On the contrary, TRIM14 overexpression in human PTC TPC1 cells promoted cell proliferation while inhibited cell apoptosis. TRIM14 exerted its oncogenic activities via promoting the activation of signal transducer and activator of transcription 3 (STAT3). Further, TRIM14 interacted with the suppressor of cytokine-signaling-1 (SOCS1), a negative regulator of STAT3 activation, and knockdown of TRIM14 inhibited SOCS1 ubiquitination. In conclusion, TRIM14 may be a prognostic factor and oncogene in PTC, and may be a potential target for PTC intervention.

Using deep convolutional neural networks for multi-classification of thyroid tumor by histopathology: a large-scale pilot study.

BACKGROUND: To explore whether deep convolutional neural networks (DCNNs) have the potential to improve diagnostic efficiency and increase the level of interobserver agreement in the classification of thyroid nodules in histopathological slides. METHODS: A total of 11,715 fragmented images from 806 patients' original histological images were divided into a training dataset and a test dataset. Inception-ResNet-v2 and VGG-19 were trained using the training dataset and tested using the test dataset to determine the diagnostic efficiencies of different histologic types of thyroid nodules, including normal tissue, adenoma, nodular goiter, papillary thyroid carcinoma (PTC), follicular thyroid carcinoma (FTC), medullary thyroid carcinoma (MTC) and anaplastic thyroid carcinoma (ATC). Misdiagnoses were further analyzed. RESULTS: The total 11,715 fragmented images were divided into a training dataset and a test dataset for each pathology type at a ratio of 5:1. Using the test set, VGG-19 yielded a better average diagnostic accuracy than did Inception-ResNet-v2 (97.34% vs. 94.42%, respectively). The VGG-19 model applied to 7 pathology types showed a fragmentation accuracy of 88.33% for normal tissue, 98.57% for ATC, 98.89% for FTC, 100% for MTC, 97.77% for PTC, 100% for nodular goiter and 92.44% for adenoma. It achieved excellent diagnostic efficiencies for all the malignant types. Normal tissue and adenoma were the most challenging histological types to classify. CONCLUSIONS: The DCNN models, especially VGG-19, achieved satisfactory accuracies on the task of differentiating thyroid tumors by histopathology. Analysis of the misdiagnosed cases revealed that normal tissue and adenoma were the most challenging histological types for the DCNN to differentiate, while all the malignant classifications achieved excellent diagnostic efficiencies. The results indicate that DCNN models may have potential for facilitating histopathologic thyroid disease diagnosis.

Deep convolutional neural network VGG-16 model for differential diagnosing of papillary thyroid carcinomas in cytological images: a pilot study.

Objective: In this study, we exploited a VGG-16 deep convolutional neural network (DCNN) model to differentiate papillary thyroid carcinoma (PTC) from benign thyroid nodules using cytological images. Methods: A pathology-proven dataset was built from 279 cytological images of thyroid nodules. The images were cropped into fragmented images and divided into a training dataset and a test dataset. VGG-16 and Inception-v3 DCNNs were trained and tested to make differential diagnoses. The characteristics of tumor cell nucleus were quantified as contours, perimeter, area and mean of pixel intensity and compared using independent Student's t-tests. Results: In the test group, the accuracy rates of the VGG-16 model and Inception-v3 on fragmented images were 97.66% and 92.75%, respectively, and the accuracy rates of VGG-16 and Inception-v3 in patients were 95% and 87.5%, respectively. The contours, perimeter, area and mean of pixel intensity of PTC in fragmented images were more than the benign nodules, which were 61.01±17.10 vs 47.00±24.08, p=0.000, 134.99±21.42 vs 62.40±29.15, p=0.000, 1770.89±627.22 vs 1157.27±722.23, p=0.013, 165.84±26.33 vs 132.94±28.73, p=0.000), respectively. Conclusion: In summary, after training with a large dataset, the DCNN VGG-16 model showed great potential in facilitating PTC diagnosis from cytological images. The contours, perimeter, area and mean of pixel intensity of PTC in fragmented images were more than the benign nodules.

A Novel N Staging System for Predicting Survival in Patients with Medullary Thyroid Cancer.

INTRODUCTION: Despite the crucially prognostic value of lymph node metastasis (LNM) in patients with medullary thyroid cancer (MTC), only the LNM compartment alone was reflected in the 8th edition of the American Joint Committee on Cancer (AJCC) system. OBJECTIVE: This study aimed to incorporate the metastatic lymph node number and metastatic lymph node ratio to generate a more accurate and appropriate N staging system for patients with MTC based on recursive partitioning analysis. DESIGN, SETTING, AND PATIENTS: Two cohorts were included in the analysis, including 1374 MTC patients from the Surveillance, Epidemiology, and End Results database as the derivation cohort, and 164 patients from Fudan University Shanghai Cancer Center as the validation cohort. The predictive performance of the alternative proposed N staging system was compared with that of the 8th AJCC system by using the Harrell concordance index (C-index) and the area under the receiver operating characteristic curve (AUC). RESULTS: In the derivation cohort, the C-index and the AUC at 10 years were 0.778 and 0.789, respectively, for the novel N staging system, and 0.749 and 0.741, respectively, for the 8th AJCC N staging system. Similar trends were also observed in the validation cohort. The proposed N staging system had a better prognostic performance. CONCLUSION: With some improvements, the novel N staging system for MTC suggested from this research may be assessed for potential adoption in the next edition of the AJCC N staging system.

Deep convolutional neural network Inception-v3 model for differential diagnosing of lymph node in cytological images: a pilot study.

BACKGROUND: In this study, we exploited the Inception-v3 deep convolutional neural network (DCNN) model to differentiate cervical lymphadenopathy using cytological images. METHODS: A dataset of 80 cases was collected through the fine-needle aspiration (FNA) of enlarged cervical lymph nodes, which consisted of 20 cases of reactive lymphoid hyperplasia, 24 cases of non-Hodgkin's lymphoma (NHL), 16 cases of squamous cell carcinoma (SCC), and 20 cases of adenocarcinoma. The images were cropped into fragmented images and divided into a training dataset and a test dataset. Inception-v3 was trained to make differential diagnoses and then tested. The features of misdiagnosed images were further analysed to discover the features that may influence the diagnostic efficiency of such a DCNN. RESULTS: A total of 742 original images were derived from the cases, from which a total of 7,934 fragmented images were cropped. The classification accuracies for the original images of reactive lymphoid hyperplasia, NHL, SCC and adenocarcinoma were 88.46%, 80.77%, 89.29% and 100%, respectively. The total accuracy on the test dataset was 89.62%. Three fragmented images of reactive lymphoid hyperplasia and three fragmented images of SCC were misclassified as NHL. Three fragmented images of NHL were misclassified as reactive lymphoid hyperplasia, one was misclassified as SCC, and one was misclassified as adenocarcinoma. CONCLUSIONS: In summary, after training with a large dataset, the Inception-v3 DCNN model showed great potential in facilitating the diagnosis of cervical lymphadenopathy using cytological images. Analysis of the misdiagnosed cases revealed that NHL was the most challenging cytology type for DCNN to differentiate.

Improved Doping and Emission Efficiencies of Mn-Doped CsPbCl3 Perovskite Nanocrystals via Nickel Chloride.

It is challenging to improve the emission efficiency of Mn-doped CsPbCl3 (Mn:CsPbCl3) nanocrystals (NCs) because the excellent optical performances are dependent on high doping efficiency and few defects and traps. Steady-state and time-resolved photoluminescence (PL) spectroscopies were used to investigate the luminescence properties of Mn:CsPbCl3 NCs with different Mn doping levels synthesized in the presence of nickel chloride. The doping efficiency of Mn ions in Mn:CsPbCl3 NCs was greatly enhanced in the presence of NiCl2, and the PL wavelength of Mn2+ ions was tuned from 594 to 638 nm by varying the concentration of dopant Mn from 0.11% to 15.25%. The high emission quantum yields of Mn:CsPbCl3 NCs with orange and red emissions peaked at 600 and 620 nm in hexane were 70% and 39%, respectively. The improvement in doping and emission efficiencies of Mn2+ was attributed to the enhanced formation energies of the Mn doping under the Mn and Ni codoped configuration and the resulting reduction of defects and traps in Mn:CsPbCl3 NCs with incorporation of Ni2+ ions.