Sample Inadequacy Research Articles

Physical prior-guided deep learning for SIM reconstruction: modeling object-to-image degradation.

Structured illumination microscopy (SIM) provides an enhanced spatial resolution of up to twice the conventional capacity. Recently, many approaches have attempted to combine deep learning frameworks with SIM reconstruction for better capability. Nonetheless, the inadequacy of training samples highlights the challenge of limited credibility and low generalization ability of deep learning, thus significantly constraining the application in biology. To tackle this issue, we propose an object-to-image plane degradation network (OIDN) guided by the physical process of optical imaging. Specifically, the proposed OIDN embeds the object-to-image plane degradation process into the reconstruction network to provide explicit guidance. With a set of learnable point spread function (PSF) parameters constrained by physical prior, OIDN successfully converts the conventional image-to-image data pattern mapping into the object-to-image plane degradation mapping that highly aligns with the optical processes of SIM imaging. Comprehensive experiments demonstrate that the proposed method reliably yields high-quality images across signal-to-noise ratio conditions, exhibiting superior generalization ability across diverse datasets and sample types.

Efficacy of Follicular Cell Pattern Analysis in Thyroid Fine-needle Aspiration Cytology Evaluation.

Fine-needle aspiration cytology (FNAC) is widely utilized for thyroid lesion diagnosis but faces challenges such as sample inadequacy and overlapping cytological features. This study examines how accurately these patterns correlate with histopathological diagnoses, shedding light on FNAC's limitations and diagnostic potential. To study the application of the architectural pattern of follicular cells in the interpretation of thyroid lesions and to demonstrate the diagnostic accuracy (DA) of FNAC. Cross-sectional study carried over 1 year. A total of 110 cases were reviewed by the cytopathologists. The prominent follicular cell architecture, namely macrofollicular, microfollicular, papillary, trabecular, three-dimensional clusters, and dispersed cells, was described in each case. In addition to these patterns, cellular morphology and background features were also noted, and a final cytological diagnosis was established. The cytology diagnosis was correlated with the histopathological diagnosis. Sensitivity, specificity, positive predictive value, negative predictive value, DA of FNAC in diagnosing nonneoplastic and neoplastic lesions. Macrofollicular pattern was seen in 80.26% of colloid goiter cases. Microfollicular pattern was observed in 72.2% of follicular neoplasm. About 62.5% of papillary thyroid carcinomas showed a papillary pattern. The trabecular pattern was seen in 42.86% of chronic lymphocytic thyroiditis and 16.67% of follicular neoplasms. The sensitivity and specificity of FNAC in diagnosing neoplastic lesions was 92.59% and 97.59%, respectively. FNAC is a simple, rapid, definite, and cost-effective primary diagnostic tool for thyroid evaluation. Cell architecture pattern is a simple and appropriate approach that complements cell morphology and background details in arriving at the final cytological diagnosis of thyroid lesions.

Molecular classification of endometrial carcinoma on endometrial biopsy: an early prognostic value to guide personalized treatment

ObjectiveMolecular features are essential for estimating the risk of recurrence and impacting overall survival in patients with endometrial cancer. Additionally, the surgical procedure itself could be personalized based on the...

Application of liquid biopsy in lung cancer management.

Current advances in the understanding of the lung cancer landscape have drastically changed the approach to treating a patient with lung carcinoma. The field has progressed from analyzing single gene to using advanced techniques like next-generation sequencing and microarray technology. While a tumor tissue sample is considered the gold standard, it has several limitations. The limitations of invasive procedures, long processing periods, inaccessibility, and sample inadequacy are being addressed by sampling biofluids, termed 'liquid biopsy,' which offers a less invasive and more accessible way to obtain tumor-related information. Liquid biopsy has transformed the care of lung cancer patients by directly targeting somatic alterations from tumors. This article provides insights into the biology, technical aspects, limitations, and practical applications of 'liquid biopsy,' focusing on cell-free DNA and circulating tumor DNA in the context of lung cancer.

An Improved SAR Ship Classification Method Using Text-to-Image Generation-Based Data Augmentation and Squeeze and Excitation

Synthetic aperture radar (SAR) plays a crucial role in maritime surveillance due to its capability for all-weather, all-day operation. However, SAR ship recognition faces challenges, primarily due to the imbalance and inadequacy of ship samples in publicly available datasets, along with the presence of numerous outliers. To address these issues, this paper proposes a SAR ship classification method based on text-generated images to tackle dataset imbalance. Firstly, an image generation module is introduced to augment SAR ship data. This method generates images from textual descriptions to overcome the problem of insufficient samples and the imbalance between ship categories. Secondly, given the limited information content in the black background of SAR ship images, the Tokens-to-Token Vision Transformer (T2T-ViT) is employed as the backbone network. This approach effectively combines local information on the basis of global modeling, facilitating the extraction of features from SAR images. Finally, a Squeeze-and-Excitation (SE) model is incorporated into the backbone network to enhance the network’s focus on essential features, thereby improving the model’s generalization ability. To assess the model’s effectiveness, extensive experiments were conducted on the OpenSARShip2.0 and FUSAR-Ship datasets. The performance evaluation results indicate that the proposed method achieves higher classification accuracy in the context of imbalanced datasets compared to eight existing methods.

An integrated and homogenized global surface solar radiation dataset and its reconstruction based on a convolutional neural network approach

Abstract. Surface solar radiation (SSR) is an essential factor in the flow of surface energy, enabling accurate capturing of long-term climate change and understanding of the energy balance of Earth's atmosphere system. However, the long-term trend estimation of SSR is subject to significant uncertainties due to the temporal inhomogeneity and the uneven spatial distribution of in situ observations. This paper develops an observational integrated and homogenized global terrestrial (except for Antarctica) station SSR dataset (SSRIHstation) by integrating all available SSR observations, including the existing homogenized SSR results. The series is then interpolated in order to obtain a 5∘ × 5∘ resolution gridded dataset (SSRIHgrid). On this basis, we further reconstruct a long-term (1955–2018) global land (except for Antarctica) SSR anomaly dataset with a 5∘ × 2.5∘ resolution (SSRIH20CR) by training improved partial convolutional neural network deep-learning methods based on 20th Century Reanalysis version 3 (20CRv3). Based on this, we analysed the global land- (except for Antarctica) and regional-scale SSR trends and spatiotemporal variations. The reconstruction results reflect the distribution of SSR anomalies and have high reliability in filling and reconstructing the missing values. At the global land (except for Antarctica) scale, the decreasing trend of the SSRIH20CR (−1.276 ± 0.205 W m−2 per decade) is smaller than the trend of the SSRIHgrid (−1.776 ± 0.230 W m−2 per decade) from 1955 to 1991. The trend of the SSRIH20CR (0.697 ± 0.359 W m−2 per decade) from 1991 to 2018 is also marginally lower than that of the SSRIHgrid (0.851 ± 0.410 W m−2 per decade). At the regional scale, the difference between the SSRIH20CR and SSRIHgrid is more significant in years and areas with insufficient coverage. Asia, Africa, Europe and North America cause the global dimming of the SSRIH20CR, while Europe and North America drive the global brightening of the SSRIH20CR. Spatial sampling inadequacies have largely contributed to a bias in the long-term variation of global and regional SSR. This paper's homogenized gridded dataset and the Artificial Intelligence reconstruction gridded dataset (Jiao and Li, 2023) are both available at https://doi.org/10.6084/m9.figshare.21625079.v1.

Influence of biliary stents on the diagnostic outcome of endoscopic ultrasound-guided tissue acquisition from solid pancreatic lesions: a systematic review and meta-analysis.

This meta-analysis analyzed the effect of an indwelling biliary stent on endoscopic ultrasound (EUS)-guided tissue acquisition from pancreatic lesions. A literature search was performed to identify studies published between 2000 and July 2022 comparing the diagnostic outcomes of EUS-TA in patients with or without biliary stents. For non-strict criteria, samples reported as malignant or suspicious for malignancy were included, whereas for strict criteria, only samples reported as malignant were included in the analysis. Nine studies were included in this analysis. The odds of an accurate diagnosis were significantly lower in patients with indwelling stents using both non-strict (odds ratio [OR], 0.68; 95% confidence interval [CI], 0.52-0.90) and strict criteria (OR, 0.58; 95% CI, 0.46-0.74). The pooled sensitivity with and without stents were similar (87% vs. 91%) using non-strict criteria. However, patients with stents had a lower pooled sensitivity (79% vs. 88%) when using strict criteria. The sample inadequacy rate was comparable between groups (OR, 1.12; 95% CI, 0.76-1.65). The diagnostic accuracy and sample inadequacy were comparable between plastic and metal biliary stents. The presence of a biliary stent may negatively affect the diagnostic outcome of EUS-TA for pancreatic lesions.

Conventional versus Liquid-based Cytology: "Man versus Machine".

Liquid-based cytology (LBC) can improve adequacy, monolayer quality with a clean background compared to conventional smears (CS). The objective was to compare the quality and diagnostic yield of CS and LBC in routine cytological investigations. This retrospective study consisted of 306 samples (255 gynecological, 39 nongynecological, and 12 fine needle aspiration cytology [FNAC]) during a 2-year period (2019-2020). From each patient, two samples were collected in the same manner in the same sitting and processed by CS and LBC (ThinPrep® 2000, Hologic Inc.). Both CS and LBC were compared for adequacy, quality, representativeness, inflammation, hemorrhage, necrosis, preservation, reactive changes, organisms, atypia/dysplasia/malignancy, and preparation/screening time. Statistical analysis was performed. No statistically significant difference was noted for adequacy, representativeness, reactive changes, preservation, and atypia/dysplasia/malignancy. CS was better in cellularity and diagnosis of inflammation and organisms, whereas LBC had a clean background and the difference was statistically significant (P = 0.0005). CS was equivalent to LBC in adequacy, representativeness, reactive changes, and atypia/dysplasia/malignancy. Adequacy comparable to LBC can be achieved in CS by careful sample collection, processing, and screening by trained cytotechnologists. CS was better in detecting organisms and inflammation than LBC. The advantages of LBC were monolayer smear, clean background, and lesser screening time, but the demerit was higher cost and longer processing time. Therefore, LBC is best suited to those laboratories having high sample inadequacy rates, lack of competent cytotechnologists, and no financial constraints. Either man or machine, appropriate and adequate sample collection by trained personnel forms the cornerstone for ensuring adequacy in both CS and LBC.

New perspectives in liquid biopsy for glioma patients.

Gliomas are the most common primary tumors of the central nervous system. They are characterized by a disappointing prognosis and ineffective therapy that has shown no substantial improvements in the past 20 years. The lack of progress in treating gliomas is linked with the inadequacy of suitable tumor samples to plan translational studies and support laboratory developments. To overcome the use of tumor tissue, this commentary review aims to highlight the potential for the clinical application of liquid biopsy (intended as the study of circulating biomarkers in the blood), focusing on circulating tumor cells, circulating DNA and circulating noncoding RNA. Thanks to the increasing sensitivity of sequencing techniques, it is now possible to analyze circulating nucleic acids and tumor cells (liquid biopsy). Although studies on the use of liquid biopsy are still at an early stage, the potential clinical applications of liquid biopsy in the study of primary brain cancer are many and have the potential to revolutionize the approach to neuro-oncology, and importantly, they offer the possibility of gathering information on the disease at any time during its history.

45P Environmental scan of molecular profiling in Canada

Colorectal Cancer Canada (CCC) developed a pan-tumour, pan-Canadian survey that was sent to healthcare professionals that aimed to conduct a high-level environmental scan of current molecular profiling practices at cancer care institutions across Canada. The survey was disseminated in August 2020 using the free web-based software Google Forms. Questions were developed by members of CCC in collaboration with a scientific advisory committee. Questions were written to include all cancer types and were optional to answer. Data was collected in Excel format and analyzed using descriptive statistics in November 2020. Data was reported in aggregate only and pooled by province. This research was funded by sponsorships granted to CCC’s Get Personal program. The survey sample consisted of 42 healthcare professionals within 26 different institutions, with at least one representative from each province. According to the survey results, turnaround times (TAT) for predictive biomarkers vary depending on the type of biomarker, testing modality, and the availability of testing in-house, with an average TAT of 3.26 weeks for NGS. 69% of respondents reported that molecular profiling is used to make patient care decisions mostly for patients with metastatic disease, and 59.5% of respondents reported that be done only if there is a Health Canada approved therapy associated with that drug. 86% of respondents reported that medical oncologists are ordering molecular profiling for their patients; however, 80% of respondents suggested pathologists should be the ones ordering molecular profiling for patients. As reported by respondents, the main challenges to molecular profiling included: inadequacy and insufficiency of samples (78%), inadequate reimbursement for the test (66%), unavailability of a certain test (53.7%), and long TATs (48.8%). The lack of standardized guidelines and coordination between centres prevent the administration of precision oncology care in Canada; therefore, in addition to generating knowledge of current profiling practices, this research could potentially influence the health of Canadians on a broader scale, by informing decision-making, influencing standardization of cancer care, and enabling capacity-building in patient-driven research.

Overcoming Challenges in Assessing Mathematical Reasoning

Despite mathematical reasoning being necessary for in-depth understanding of mathematical concepts, many teacher experience difficulty in assessing it. Data were collected from 34 primary teachers at 4 Victorian government schools at two post- lesson reflective sessions following lessons with a focus on reasoning. These sessions facilitated teachers’ collaborative efforts to assess their students’ reasoning from students’ work samples. The data included transcripts of all the reflective sessions; written work samples; and associated completed rubrics. Analysis of these data enabled identification of seven challenges teachers experienced in assessing reasoning: Limited guidance provided by curriculum documents; Teachers’ knowledge of reasoning; Teacher noticing and interpretation of student reasoning; Students’ difficulties in articulating their reasoning; Assessing progress in reasoning; Inadequacy of work samples; and Challenges in tracking and reporting student progress in reasoning. The discussion presents strategies to overcome these challenges.

GENERATING SYNTHETIC 3D POINT SEGMENTS FOR IMPROVED CLASSIFICATION OF MOBILE LIDAR POINT CLOUDS

Abstract. Mobile lidar point clouds are commonly used for 3d mapping of road environments as they provide a rich, highly detailed geometric representation of objects on and around the road. However, raw lidar point clouds lack semantic information about the type of objects, which is necessary for various applications. Existing methods for the classification of objects in mobile lidar data, including state of the art deep learning methods, achieve relatively low accuracies, and a primary reason for this under-performance is the inadequacy of available 3d training samples to sufficiently train deep networks. In this paper, we propose a generative model for creating synthetic 3d point segments that can aid in improving the classification performance of mobile lidar point clouds. We train a 3d Adversarial Autoencoder (3dAAE) to generate synthetic point segments that exhibit a high resemblance to and share similar geometric features with real point segments. We evaluate the performance of a PointNet-like classifier trained with and without the synthetic point segments. The evaluation results support our hypothesis that training a classifier with training data augmented with synthetic samples leads to significant improvement in the classification performance. Specifically, our model achieves an F1 score of 0.94 for vehicles and pedestrians and 1.00 for traffic signs.

Diagnostic yield, accuracy, and complication rate of CT-guided biopsy for spinal lesions: a systematic review and meta-analysis

BackgroundCT-guided biopsy is a commonly used diagnostic procedure for spinal lesions. This meta-analysis aims to investigate its diagnostic performance and complications, as well as factors influencing outcomes.MethodsA systematic review of...

Percutaneous wide bore needle biopsy for spinal lesions using the kyphoplasty cannula—a technical note

Management of vertebral and discovertebral lesions depends on highly specific modalities like biopsy and histopathology. The transpedicular and transforaminal routes have gained popularity for obtaining samples for diagnosing these lesions. Sample inadequacy plays an important part in lowering the diagnostic accuracy. Present study introduces a technical modification by using a kyphoplasty cannula and pituitary forceps through it thus improving sample adequacy and diagnostic accuracy. Seventy-one patients with radiological evidence of spinal lesions from T1 to S1 were included. After obtaining samples via the transpedicular route they were sent for aerobic, anaerobic, acid fast bacilli & fungal cultures, Mycobacterium Tuberculosis/Radio Immuno Frequency (MTB/RIF) Assay and histopathological examination. In lesions with minimal vertebral erosions the transforaminal route was used to obtain samples from the disc space. Sixty-eight patients (95.8%) had a definite diagnosis of which 37 patients (54.4%) tested positive for vertebral osteomyelitis/discitis. All the samples were found to be adequate on histopathological examination. The remaining 3 patients showed chronic inflammation and responded to oral and intravenous antibiotics. None of the patients underwent repeat biopsy. There were no perioperative complications. Percutaneous transpedicular biopsy has evolved as the intervention of choice in diagnosing radiologically proven vertebral body lesions. Our technique of using a pituitary forceps through a cannula is highly effective in getting an adequate representative sample with excellent accuracy in diagnosis. This procedure is beneficial for soft tissue lesions and for infective pathology especially discitis.

A48 USE OF ENDOSCOPIC ULTRASOUND FINE NEEDLE ASPIRATE AND ENDOSCOPIC ULTRASOUND FINE NEEDLE BIOPSY FOR DETECTION OF GATA6 EXPRESSON IN PANCREATIC DUCTAL ADENOCARINCOMA

Abstract Background GATA6 is a transcription factor that can be used to distinguish between the basal-like and classical subtypes of pancreatic ductal adenocarcinoma (PDAC). The basal-like subtype has been demonstrated to be less responsive to modified FOLFIRINOX chemotherapy and thus can be used to predict response to specific chemotherapies. To date, GATA6 expression has only been evaluated in surgically resected PDAC specimens. Less than 15% of patients with PDAC are eligible for surgery. Endoscopic ultrasound guided fine-needle aspirate (EUS-FNA) and biopsy (EUS-FNB) can potentially help assess GATA6 expression in PDAC and in turn, help guide personalized treatment selection in all cases of PDAC. Aims The primary objective of this study was to explore the yield of EUS-FNA and EUS-FNB for the detection of GATA6 among patients with PDAC. The study also aimed to explore the impact of lesion location on sample adequacy and type of fixative on validity of GATA6 staining. Methods This study was conducted from November 2017 to October 2019. Consecutive patients with a diagnosis of PDAC confirmed by biopsy were included. Patients underwent either EUS-FNB or EUS-FNA to obtain tissues samples. Samples were fixed in either neutral buffered formalin (NBF) or a methanol based buffered solution (Cytolyt) and evaluated by a specialized cytopathology team. Fisher’s exact test was used and a p-value ≤0.05 was considered to indicate statistical significance. Results Forty-four patients were included in the study. Twenty-three (52%) patients were male and the median age of patients was 67.5 years. Twenty-five lesions were located in the head and neck of the pancreas, 14 were located in the body, and 4 were located in the tail. One patient was found to have a local recurrence of PDAC at the surgical bed of a previous Whipple procedure. Eighteen lesions were sampled by EUS-FNA and 26 were sampled using EUS-FNB. Thirty-eight (86%) samples were adequate for assessment of GATA6. Sampling technique (p=0.68) and fixative type (p=1.00) did not appear to affect sample adequacy. Compared to pancreatic body or tail specimens, samples obtained from the head or neck of the pancreas were more likely to be inadequate for analysis (p=0.03). Conclusions EUS-FNA and EUS-FNB samples are efficacious methods of assessing GATA6 expression in PDAC. This is the first predictor of treatment response that has been demonstrated to be obtained without surgical resection. Neither EUS needle type or alcohol fixation before cell block preparation appear to impact GATA6 detection. Lesions in the pancreatic head or neck appear to be associated with higher rates of sample inadequacy. Larger, prospective studies are required to confirm our findings. Funding Agencies None

FP07.14 Implementation of a Lung Biopsy Histology Protocol Supports Successful Cancer Biomarker Testing

Non-small cell lung cancer (NSCLC) management increasingly relies on biomarker testing, which presents a logistical challenge for utilization of lung biopsy specimens beyond routine diagnosis. Appropriate allocation of tissue from small biopsies is critical in order to permit for accurate histologic diagnosis while preserving tissue for ancillary studies such as PD-L1 testing and next generation sequencing (NGS). International surveys show that satisfaction with molecular testing is less than ideal, driven in part by inadequacy of formalin-fixed paraffin embedded (FFPE) tumor tissue for biomarker testing in over a quarter of clinical cases by some reports. We developed a histology protocol (LUNGCOR) that generates multiple unstained FFPE slides up front in order to increase tissue yields for pathologic diagnosis and ancillary testing. We examined the efficacy of this protocol in regards to molecular testing success rates. Biopsy samples (e.g. transbronchial or fine needle biopsies) were flagged for the LUNGCOR protocol based on a clinical indication of "lung mass" or need for genomics. Proceduralists were requested to prospectively specify when samples were obtained for the purposes of biomarker testing. Per protocol, after embedding, 2 H&E slides with 18 intervening 4-micron unstained FFPE slides were prepared and reserved for immunohistochemistry or molecular studies. We performed a retrospective review of consecutive LUNGCOR specimens within the pathology information system and electronic medical record to determine the usage patterns for the unstained slides and success rates of molecular testing. This analysis is restricted to the first 2 years (2015-2016) following protocol initiation. 300 specimens were designated for the LUNGCOR protocol in this time period. Of these specimens, 84% resulted in a malignant diagnosis and 175 (58%) were non-squamous NSCLC. Of these, 87 (50%) of NSCLC underwent molecular profiling of some kind, either in house NGS, single gene PCR assays, or send-out NGS. For cases undergoing in house next generation sequencing (447-gene panel), 61 of 69 (88%) cases generated complete genotyping for all required and emerging biomarkers for targeted therapy. Sequencing failures were most commonly due to inadequate tumor content in the biopsy specimen, rather than inadequate size of the overall sample. Generation of unstained slides during upfront histologic processing of known or suspected NSCLC increases availability of tissue for molecular testing and reduces risk of sample inadequacy due to specimen loss during routine handling steps, such as "refacing" the block. This straightforward process improvement can improve the yield of invasive biopsies for essential biomarker testing and contribute to improved patient outcomes.

Comparative study of conventional Ziehl−Neelsen and bleach concentration method in detecting acid-fast bacilli in fine-needle aspiration material of lymph nodes

Background: Tuberculosis (TB), specific infectious disease, affects most of the organs of the human body. Extrapulmonary TB manifests commonly as lymphadenopathy. Emergence of acquired immunodeficiency syndrome warrants the rapid diagnosis to limit its spread. Aim and Objectives: The application of bleach concentration method in detecting tubercle bacilli in the material obtained by fine-needle aspiration (FNA) of lymph nodes, evaluating sensitivity of bleach concentration method over conventional direct smear method in diagnosing tubercular lymphadenitis, detecting bacilli in the background of granuloma formation. Materials and Methods: The prospective, hospital-based, cross-sectional study was carried out in a study population of 78 patients where 15 cases got eliminated following the cytological diagnosis of malignancy and inadequacy of sample. Patients with lymphadenopathy underwent FNA after clinical examination, history taking, and informed consent. The aspirated material was stained with Leishman, H and E, conventional Ziehl−Neelsen (ZN), and bleach concentration method. Data were collected and statistical analysis done. Results: The predominant age group affected was (16–30) years with male preponderance. 44.44% cases were cytomorphologically diagnosed with tuberculous lymphadenitis. Acid-fast Bacilli (AFB) was found in 23 cases by conventional ZN method and 39 cases by bleach concentration method. Sensitivity of bleach method was 92.35% compared to 54.76% in conventional ZN method. Bacilli positivity was more in cases presenting with necrosis than with granuloma alone. Conclusion: Bleach method for detecting AFB is more sensitive than conventional ZN method. It is safe, inexpensive, easy to perform, requires no additional equipment, and safe to handle in laboratory.

Molecular Therapeutics of Non-Small Cell Lung Cancer (NSCLC) and Challenges in Repeat Tissue Biopsy

With advances in targeted and personalized treatment for lung cancer, molecular analysis of tumors is routinely performed for sequencing of treatment options in patients with advanced non-small-cell lung cancer (NSCLC). Oncogene addiction due to driver mutations includes EGFR exon 20 insertion mutations, MET amplification, EML4-AL, KRAS G12C point mutations, RET rearrangements, HER2 amplification and mutations, and FGFR amplification and translocations. A re-biopsy at the time of tumor recurrence or progression after first-line treatment failure is important for further molecular assessment and personalized therapy. However, repeat tumor biopsies are fraught with challenges including access to the tumor, sample inadequacy, patient consent, patient performance status, safety, or physician’s choice or assessment. Cytological specimens are gaining importance but are limited due to validation difficulties. Liquid biopsies, which are minimally invasive have shown promise to assess dynamic biomarkers using ctDNA analysis and are thus frequently considered in routine clinical practice in advanced NSCLC patients to guide further targeted treatment. Here we present a comprehensive review that emphasizes the significance of performing tumor re-biopsy in advanced stage NSCLC patients following resistance to first-line treatment and simultaneously highlights the current challenges in performing the same and the current status and future perspectives of liquid biopsy in NSCLC.

Model-Based Transfer Learning and Sparse Coding for Partial Face Recognition

With the growing needs of practical applications such as security monitoring, partial face recognition is a challenging but important issue, because the captured faces in real-world surveillance videos may be occluded or with variations. Though current face recognition methods perform well in relatively constrained scenes, they may suffer from degradation for partial faces. In this paper, we propose a framework of model-based transfer learning and sparse coding (MTLSC) for partial face recognition. First, due to less information in partial face image, we exploit the mirrored image of an original probe sample as sample augment to provide further information. Considering the inadequacy of training face samples, we obtain face features based on model-based transfer learning VGGNet that is pre-trained on VGGFace dataset. Then we reconstruct face features by sliding window in view of different sizes of partial face hard to extract the same feature dimension. Finally we carry out sparse coding with rectification and calculate the minimum score of the probe and mirrored samples among all classes to get the results. Thus, by model-based transfer learning, sliding window for feature reconstruction and sparse coding with rectification, the proposed framework improves partial face recognition performance. Experimental results on three face databases (LFW, AR and NIR), and two person re-identification databases (iLIDS-VID and PKU-Reid) demonstrate our method is effective for partial face recognition.

Endometrial pathology in postmenopausal women: Analysis of 112 cases in tertiary care centre

Background: Postmenopausal bleeding is a frequent finding accounting for 5-10% of women in gynaecology clinic. About 10% of these patients have primary or secondary malignancy. Aim: The histopathological patterns of endometrium, in postmenopausal women presenting with bleeding were studied along with the frequency of endometrial cancer. Materials and Methods: It was both a prospective and retrospective, observational study carried out over a period of one and half year in tertiary care teaching hospital on 112 postmenopausal women, above 40 year of age with history of one year of amenorrhoea without hormone replacement therapy. The samples were obtained by dilatation and curettage and endometrial pipelle procedure. Results and Discussion: Maximum patients belonged to age group of 46-50 year. The most common histopathological pattern observed was atrophic endometrium in 42 (37.5%) patients followed by simple hyperplasia in 13(11.6%) and endometrial polyp in 7(6.25%) cases. Irregular shedding of endometrium was diagnosed in 8 case (7.14%). The maximum patients of endometrial carcinoma (5.35 %) were noted in 61-65 age groups. Inflammatory pathology was found in 4(3.56%) and atypical hyperplasia was seen a single case (0.89%). In 8 cases, the opinion could not be offered due to inadequacy of sample. Conclusion: As the incidence of malignancy in postmenopausal period remains sufficiently high, it requires immediate investigation in the form of endometrial sampling for early diagnosis, prompt treatment and vigilant follow up. Keywords: Postmenopausal bleeding, Histopathology, Endometrial cancer.