Histopathological Classification Research Articles

Enhancing Histopathological Image Classification Performance through Synthetic Data Generation with Generative Adversarial Networks.

Breast cancer is the second most common cancer worldwide, primarily affecting women, while histopathological image analysis is one of the possibile methods used to determine tumor malignancy. Regarding image analysis, the application of deep learning has become increasingly prevalent in recent years. However, a significant issue is the unbalanced nature of available datasets, with some classes having more images than others, which may impact the performance of the models due to poorer generalizability. A possible strategy to avoid this problem is downsampling the class with the most images to create a balanced dataset. Nevertheless, this approach is not recommended for small datasets as it can lead to poor model performance. Instead, techniques such as data augmentation are traditionally used to address this issue. These techniques apply simple transformations such as translation or rotation to the images to increase variability in the dataset. Another possibility is using generative adversarial networks (GANs), which can generate images from a relatively small training set. This work aims to enhance model performance in classifying histopathological images by applying data augmentation using GANs instead of traditional techniques.

Prognostic Impact of Proteinuria at Manifestation in Adult Nephrotic Syndrome Patients: Insights from a Prospective Cohort Study.

Background and objective Nephrotic syndrome is a significant worldwide health concern impacting millions of people and is marked by heavy proteinuria, edema, and decreased serum levels of albumin. Albuminuria arises from abnormal glomerular permeability and impaired tubular reabsorption, contributing to declining kidney function and a heightened risk of cardiovascular complications. The objective of this study was to investigate the prognostic role of proteinuria on the persistent decline in estimated glomerular filtration rate (eGFR) (<30 ml/minute/1.73m2) during follow-up and the dynamics of remission and relapse in various subtypes of nephrotic syndrome. Methods A total of 134 adult patients, diagnosed with various histopathological categories of nephrotic syndrome, were prospectively studied. Urine protein levels were assessed using the pyrogallol red-molybdate (PRM) method. The Kaplan-Meier analysis and log-rank test were utilized to assess the prognostic role of proteinuria at manifestation on persistent decline in estimated glomerular filtration rate(eGFR) (<30 ml/minute/1.73m2) and to evaluate remission and relapse based on proteinuria levels over an 18-month follow-up period. Results Patients with sub-nephrotic levels of proteinuria at manifestation did not progress to end-stage renal disease on follow-up. Patients with sub-nephrotic levels of albuminuria at manifestation were significantly associated with remission on follow-up. The Kaplan-Meier analysis indicated a significant probability of persistent eGFR decline (p < 0.001) in adult nephrotics with higher levels of albuminuria. Furthermore, patients with sub-nephrotic range proteinuria had earlier remission (p < 0.001) compared to those with relapse (p = 0.001) during the follow-up, as demonstrated by log-rank tests. Conclusion This study highlights that sub-nephrotic albuminuria at manifestation is linked to a reduced risk of renal progression and persistent eGFR decline compared to adult nephrotics with higher levels of albuminuria. Early detection and effective management of proteinuria, are crucial for preventing renal function decline and improving patient outcomes.

Feasibility study of ResNet-50 in the distinction of intraoral neural tumors using histopathological images.

Neural tumors are difficult to distinguish based solely on cellularity and often require immunohistochemical staining to aid in identifying the cell lineage. This article investigates the potential of a Convolutional Neural Network for the histopathological classification of the three most prevalent benign neural tumor types: neurofibroma, perineurioma, and schwannoma. A model was developed, trained, and evaluated for classification using the ResNet-50 architecture, with a database of 30 whole-slide images stained in hematoxylin and eosin (106, 782 patches were generated from and divided among the training, validation, and testing subsets, with strategies to avoid data leakage). The model achieved an accuracy of 70% (64% normalized), and showed satisfactory results for differentiating two of the three classes, reaching approximately 97% and 77% as true positives for neurofibroma and schwannoma classes, respectively, and only 7% for perineurioma class. The AUROC curves for neurofibroma and schwannoma classes was 0.83%, and 0.74% for perineurioma. However, the specificity rate for the perineurioma class was greater (83%) than in the other two classes (neurofibroma with 61%, and schwannoma with 60%). This investigation demonstrated significant potential for proficient performance with a limitation regarding the perineurioma class (the limited feature variability observed contributed to a lower performance).

Evaluation of the extended Japan NBI expert team classification of subtype 2B in laterally spreading colorectal tumors based on blue laser imaging.

The Japan NBI Expert Team (JNET) classification has good diagnostic potential for colorectal diseases. We aimed to explore the diagnostic value of the JNET classification type 2B (JNET2B) criteria for colorectal laterally spreading tumors (LSTs) based on magnifying endoscopy with blue laser imaging (ME-BLI) examination. Between January 2017 and June 2023, 218 patients who were diagnosed as having JNET2B-type LSTs using ME-BLI were included retrospectively. Endoscopic images were reinterpreted to categorize the LSTs as JNET2B-low (n = 178) and JNET2B-high (n = 53) LSTs. The JNET2B-low and JNET2B-high LSTs were compared based on their histopathological and morphological classifications. Among the 178 JNET2B-low LSTs, 86 (48.3%) were histopathologically classified as low-grade intraepithelial neoplasia, 54 (30.3%) as high-grade intraepithelial neoplasia (HGIN), 37 (20.8%) as intramucosal carcinoma (IMC), and one (0.6%) as superficial invasive submucosal carcinoma (SMC1). Among the 53 JNET2B-high LSTs, five (9.4%) were classified as HGIN, 28 (52.9%) as IMC, 15 (28.3%) as SMC1, and 5 (9.4%) as deep invasive submucosal carcinoma. There were significant differences in this histopathological classification between the two groups (P < 0.001). However, there was no significant difference between JNET2B-low and JNET2B-high LSTs based on their morphological classification (granularvs nongranular) or size (<20 mm vs ≥20 mm). Besides, the κ value for JNET2B subtyping was 0.698 (95% confidence interval 0.592-0.804) between the two endoscopists who reassessed the endoscopic images. The JNET2B subtyping of LSTs has a diagnostic potential in the preoperative setting, and may be valuable for treatment decision-making.

Explainable machine learning predicts survival of retroperitoneal liposarcoma: A study based on the SEER database and external validation in China

AbstractObjectiveWe have developed explainable machine learning models to predict the overall survival (OS) of retroperitoneal liposarcoma (RLPS) patients. This approach aims to enhance the explainability and transparency of our modeling results.MethodsWe collected clinicopathological information of RLPS patients from The Surveillance, Epidemiology, and End Results (SEER) database and allocated them into training and validation sets with a 7:3 ratio. Simultaneously, we obtained an external validation cohort from The First Affiliated Hospital of Naval Medical University (Shanghai, China). We performed LASSO regression and multivariate Cox proportional hazards analysis to identify relevant risk factors, which were then combined to develop six machine learning (ML) models: Cox proportional hazards model (Coxph), random survival forest (RSF), ranger, gradient boosting with component‐wise linear models (GBM), decision trees, and boosting trees. The predictive performance of these ML models was evaluated using the concordance index (C‐index), the integrated cumulative/dynamic area under the curve (AUC), and the integrated Brier score, as well as the Cox–Snell residual plot. We also used time‐dependent variable importance, analysis of partial dependence survival plots, and the generation of aggregated survival SHapley Additive exPlanations (SurvSHAP) plots to provide a global explanation of the optimal model. Additionally, SurvSHAP (t) and survival local interpretable model‐agnostic explanations (SurvLIME) plots were used to provide a local explanation of the optimal model.ResultsThe final ML models are consisted of six factors: patient's age, gender, marital status, surgical history, as well as tumor's histopathological classification, histological grade, and SEER stage. Our prognostic model exhibits significant discriminative ability, particularly with the ranger model performing optimally. In the training set, validation set, and external validation set, the AUC for 1, 3, and 5 year OS are all above 0.83, and the integrated Brier scores are consistently below 0.15. The explainability analysis of the ranger model also indicates that histological grade, histopathological classification, and age are the most influential factors in predicting OS.ConclusionsThe ranger ML prognostic model exhibits optimal performance and can be utilized to predict the OS of RLPS patients, offering valuable and crucial references for clinical physicians to make informed decisions in advance.

Podocyte injury as a potential therapeutic target in IgA nephropathy: should pathology guide us?

The Oxford histopathologic classification (MEST-C: scores for lesions indicating active glomerular inflammation, mesangial [M] and endocapillary [E] hypercellularity as well as cellular or fibrocellular crescents [C], and for segmental glomerulosclerosis [S] and interstitial fibrosis and/or tubular atrophy [T]) is useful in helping assess prognosis in patients with IgA nephropathy. Elements of this classification indicative of active glomerular inflammation, endocapillary hypercellularity and crescents, also have been found to be responsive to immunosuppressive therapy, potentially including newer agents specifically targeting mediators of such inflammation. In this issue of Kidney International, Bellur and coworkers identify histopathologic subtypes of segmental glomerulosclerosis in IgA nephropathy showing podocyte injury that also behave like active lesions, including showing improved outcomes with immunosuppression. This podocyte injury, identifiable only by kidney biopsy, may represent a potential therapeutic target in some patients with IgA nephropathy.

MPSA: Multi-Position Supervised Soft Attention-based convolutional neural network for histopathological image classification

In recent years, significant achievements have been made in the field of histopathological image analysis using convolutional neural networks (CNNs). However, existing CNNs fail to fully capture the important local structures and regional information in histopathological images due to the complex tissue structures and variable pathological features present in these images. They often treat all regions equally, which further exacerbates the challenge of accurately analyzing such images. Current network model can’t extract deep layer features efficiently without guiding. To alleviate this problem, we propose a novel network model called Multi-Position Supervised Soft Attention (MPSA). MPSA adds regions of interest (RoI) labels at multiple feature layers for deep supervision, and then uses the supervised layers as soft attention to guide the learning of the classification network, enabling the network to accurately extract features of the lesion target. Additionally, we design a Multi-level Attention Feature Enhancement Module (MAFEM), which combines multiple levels of attention mechanisms to enhance the performance of the convolutional neural network in histopathological image classification. MAFEM includes spatial attention, soft attention of the main branch, and our proposed soft attention for multi-branch feature fusion. The proposed soft attention for multi-branch feature fusion aims to enhance the predictive performance of the classification model by activating relevant neurons in the diagnostic area in a highly activated state, while effectively avoiding noise activation. This innovative approach ensures that the model can focus on the most pertinent information, leading to improved classification accuracy. We conducted classification experiments on the liver cancer histopathological images dataset and the results showed that our method achieved a classification accuracy of 95.79%, indicating that it is very effective in the analysis of liver histopathological images. Our proposed network architecture has also demonstrated good generalization ability in other medical datasets, achieving a classification accuracy of 84.41% on the ultrasound carotid plaque dataset.

Evaluating Postoperative Outcomes and Investigating the Usefulness of EU-TIRADS Scoring in Managing Pediatric Thyroid Nodules Bethesda 3 and 4

The aim was to assess postoperative outcomes in pediatric thyroid nodules with atypia of undetermined significance (AUS/FLUS) or suspicious for a follicular neoplasm (SFN) and their respective the European-Thyroid Imaging Reporting and Data System (EU-TIRADS) scores. Forty-four pediatric patients at a single center with thyroid nodules classified as AUS/FLUS or SFN from August 2019 to December 2022 were retrospectively reviewed. Data on demographics, thyroid function, nodule size, and ultrasonographic features were collected. Postoperative pathologies were categorized into benign, low-risk, and malignant neoplasms according to the World Health Organization 2022 criteria, and EU-TIRADS was used for retrospective radiological scoring. Among 21 (47.7%) of patients who had surgical intervention, 72% had Bethesda 3 and 28% had Bethesda 4 thyroid nodules. Post-surgical histopathological classifications were 43% benign, 19% low-risk, and 38% malignant. Of note, EU-TIRADS 3 and 5 scores were present in 44% and 56% of the benign cases, respectively. Malignant cases tended to produce higher EU-TIRADS scores, with 64% rated as EU-TIRADS 5. Bethesda category 4 nodules had a 66% malignancy rate, significantly higher than the 27% in category 3. A substantial proportion of histologically benign cases were classified as EU-TIRADS 5, suggesting that EU-TIRADS may lead to unnecessary biopsies in benign cases. Malignant cases were more likely to have a higher EU-TIRADS score, indicating a positive correlation with malignancy risk, particularly in Bethesda 4 cases. However, the EU-TIRADS system’s predictive value for malignancy in Bethesda 3 cases was poorer.

#1833 Analysis of clinical and laboratory manifestations in different classes of lupus nephritis—a Serbian single center experience

Abstract Background and Aims Lupus nephritis represents one of the most serious manifestations of systemic lupus erythematosus. The gold standard for determining the class of lupus nephritis is percutaneous kidney biopsy and the histopathological analysis of the biopsy specimen. The aim of this study was to analyze the clinical and laboratory characteristics of different histopathological classes of lupus nephritis (LN) at the time of the systemic lupus diagnosis and onset of symptoms. Method This retrospective cross-sectional study included a total of 54 subjects who were hospitalized at the Clinic for Nephrology and Clinical Immunology of the University Clinical Center of Vojvodina from 2017 to 2022 and underwent percutaneous kidney biopsy. Available clinical and laboratory data, as well as the histopathological findings of kidney biopsies, were analyzed. Based on the histopathological classification, patients were divided into four groups (Class II, III, IV, and V), and the clinical and laboratory characteristics were analyzed for all subjects. Due to a low number of patients, no analysis of Class I and Class VI was made. Results The most prevalent class among the subjects was Class IV, in which the lowest values of the C3 complement component were recorded. Subjects in this LN class had higher blood pressure values, a higher frequency of clinical manifestations such as arthralgia, and joint swelling. Class IV LN was also associated with elevated anti-dsDNA antibody values with a significant statistical significance (p = 0.004), lower red blood cell parameters, and higher levels of nitrogenous substances compared to other classes. The lowest values of total serum proteins and albumins, as well as the highest values of total cholesterol, LDL cholesterol, and triglycerides, were observed in the group of subjects with Class V LN; C-reactive protein had the highest average values in Class III and IV, which showed a statistical significance (p = 0.007). The lowest values of total serum calcium and iron were in Class III, and 24-hour proteinuria was most pronounced in Class V (p = 0.001), whereas microscopic hematuria was most present in Class II LN (p = 0.03). Conclusion Patients with lupus nephritis exhibit different clinical and laboratory manifestations based on histopathological classification and the time of diagnosis. The results of this study suggest their potential use in diagnosis, assessment of histopathological classification, and the development of a targeted and personalized therapeutic approach. These parameters will be used to develop machine learning models which might help in further evaluation and treatment.

#675 Application of ANCA-associated vasculitis Histopathological Classification and Renal Risk Score for prediction of end-stage kidney disease

Abstract Background and Aims Antineutrophil cytoplasmic antibody (ANCA)-associated vasculitis (AAV) is a systemic autoimmune disease predominantly characterized by inflammation of small-sized blood vessels along with the presence of circulating ANCA. The kidney is frequently involved and, even though immunosuppressive therapy has dramatically improved the outcome of AAV patients, a significant proportion of patients progress to end-stage kidney disease (ESKD). Therefore, establishing and validating tools to predict renal outcomes in patients with AAV is crucial. Here we analyzed the application of two tools in predicting kidney survival in a Portuguese single-center cohort with AAV. Method We conducted a retrospective cohort study at a tertiary-care hospital in Portugal (Hospital de São João, Porto), from December 2010 to June 2023. Patients with a kidney biopsy-proven AAV were included. Serological ANCA-negative patients and patients with overlap syndromes, such as AAV in combination with anti-glomerular basement membrane disease, were excluded. Based on electronic records, information on demographics and clinical variables were collected. Kidney biopsies were reviewed to calculate: Berden histopathological classification and the Brix Renal Risk Score (RRS). Kidney survival was defined as the time from diagnosis to the need for kidney replacement therapy. A Kaplan-Meier analysis was performed to compare kidney survival based on the mentioned classifications. Results Fifty-one patients were enrolled, 78% (n = 40) of whom were positive for myeloperoxidase (MPO). The majority were men (61%, n = 31) with a mean age of 65 (± 12) years and a mean serum creatinine of 4.59 mg/dL (± 2.74) at diagnosis. During a median follow-up of 39 (IQR, 7-84) months, 20 patients (39%) progressed to end-stage kidney disease (ESKD), 10 patients (20%) died with preserved renal function and 21 patients (41%) did not need hemodialysis. Concerning Berden classification, 13 patients with a crescentic class, 4 patients with a mixed class and 3 patients with a sclerotic class developed ESKD. After 48 months of follow-up, renal survival was 100%, 90%, 61% and 26.7% in the focal, mixed, crescentic and sclerotic class groups, respectively. Berden classification did not correlate with kidney survival (p = 0.211). Concerning RRS, 10 patients in the high-risk group and 5 patients in the medium-risk group developed ESKD. After 48 months of follow-up, renal survival was 100% in the low-risk group, 85.7% in the medium-risk group, and 41.6% in the high-risk group. A correlation was established between kidney survival and RRS score (p = 0.003). Conclusion In our cohort RRS was predictive of ESKD, contrary to Berden histopathological classification. The information derived from these risk models strongly add to standard biochemical and clinical data. Patients with higher risk of ESKD need close follow-up and a tailored treatment plan regarding preparation for renal replacement therapy.

#2257 Kidney outcomes in ANCA-associated vasculitis—application of Berden's Classification and ANCA Renal Risk Score in a single centre cohort

Abstract Background and Aims Anti-neutrophil cytoplasmic antibody (ANCA) -associated vasculitis is the most common cause of rapidly progressive glomerulonephritis and is a life-threatening condition. The course of the disease is variable and disease prognostication remains challenging. Histological and clinical scores have been developed to predict kidney outcomes, to tailor treatment. The aim of this study is to apply Berden´s histopathological classification (BHC) and ANCA renal risk score (ARRS), proposed by Brix and co-workers, in a cohort of patients with ANCA-associated vasculitis and to analyse their performance in renal outcomes prediction. Method In this retrospective observational study, we included patients with biopsy proven ANCA-associated vasculitis, diagnosed at our centre between March 2011 and March 2023. We consulted the clinical records and the kidney biopsy reports. Biopsies were categorized according to both scores. Follow up time was up to 48 months since the date of diagnosis. In BHC patients were divided into 4 groups (focal, crescentic, sclerotic and mixed) according to normal glomeruli, cellular crescents and globally sclerosed glomeruli proportion. When applying the ARRS, patients were divided into 3 groups: low, medium or high risk. This is a histopathological and biochemical score that combines normal glomeruli and tubular atrophy and interstitial fibrosis (IFTA) percentages with estimated glomerular filtration rate (eGFR) at time of diagnosis. Factors related to renal survival were evaluated by Spearman Rank Correlation, Binary Logistic Regression analysis and Kaplan-Meier curves were used for assessing time between diagnosis and end stage kidney disease (ESKD). Results Our study included 31 patients, 54.8% were male, median age at diagnosis was 62 (IQR, 53-69) years, with a median eGFR of 8.3 (IQR, 6.2-13.3) mL/min/1.73 m2 and a median urine albumin to creatinine ratio (uACR) of 1.5 (IQR, 0.88-2.2) g/g. 19 patients (61.3%) were positive for myeloperoxidase and 5 (16.1%) were proteinase 3 positive. Median follow up time was 36 (IQR, 9-48) months. 10 patients (32%) developed ESKD and 9 patients (29%) died. Regarding BHC, 4 patients (13%) presented a focal class, 18 (58%) crescentic, 7 (23%) mixed and 2 (6%) sclerotic. Progression to ESKD occurred in 7 patients with crescentic class (38.9%), 2 patients with mixed (28.6%) and 1 patient with sclerotic (50%). None of the patients with focal class developed ESKD. There were no significant differences in kidney survival curves by BHC class (Log Rank p = 0.47) (Fig. 1a). In ARRS, 1 patient (3%) was in the low-risk group, 19 patients (61%) in the medium risk and 11 patients (35%) in the high risk. Characteristics of the population by ARRS score group and detailed score discrimination are specified in Table 1. Progression to ESKD, occurred in 5 patients of medium risk group (26.3%) and also in 5 patients in the high-risk group which represented 45.5% of this group. Kidney survival curves by ARRS score group had no statistic difference (Log Rank p = 0.46) (Fig. 1b). ESKD was positively associated with serum creatinine (sCr) at admission (p = 0.02), highest sCr (p = 0.001) and leukocyturia (p = 0.05). By multivariate analysis adjusted to gender, age, sCr at admission, leukocyturia and BHC or ARRS, the only factor associated with ESKD was sCr at admission (OR 1.45, CI 1.03-2.06, p = 0.05). Conclusion In our cohort of patients, sCr at patient hospital admission was the best predictor for progression to ESKD. By ARRS, just one patient had low risk, and as expected, kidney survival was higher in the medium risk group (68.5%) than in the high-risk group (53%), although not reaching statistical significance. BHC also had no association with kidney outcomes, although kidney survival was 100% in focal class and sclerotic class had the lowest survival (50%), yet crescentic class revealed a worse kidney survival (57.1%) than mixed class (71.4%). The small number of patients, overall and in each group, and the short follow up period can be pointed as limitations of this study.

#2168 Lupus nephritis histopathological classification system: a questionnaire-based survey of the Renal Pathology Society (RPS)

Abstract Background and Aims Kidney biopsy is the gold standard for lupus nephritis diagnosis, defining histologic class, disease activity and chronicity in detail and providing guidance for treatment. The updated 2018 ISN/RPS histopathological classification in lupus nephritis is widely used for prognosis and treatment decisions. Given the evolving phenotypes of lupus nephritis and the success of new targeted agents in clinical trials, there is ongoing discussion about the need for updating existing classification systems. A survey to assess the current use of the lupus nephritis histopathological classification systems in everyday practice was conducted on behalf of the RPS. Method An online survey was sent between September 27th and October 24th, 2023 to active members of the RPS. The survey used an internet platform and contained both multiple-choice and open-ended questions; results were analyzed anonymously. Results 185 of 562 RPS members replied to the questionnaire. 179 (97%) were pathologists and the level of expertise in lupus nephritis was good: 120 (65%) participants indicated they encounter more than 20 biopsies with lupus nephritis per year, while only 13% encounter fewer than 10. Almost 90% of responders discuss kidney biopsy results in a multidisciplinary meeting involving both clinicians and pathologists, on a weekly or monthly basis. The 2018 ISN/RPS lupus nephritis classification is used in most cases (92%), while 7% of participants prefer the earlier 2004 version. The average grade on the utility of the 2018 ISN/RPS classification was 8 (IQR 7-9) on a scale from 0 (not useful) to 10 (extremely useful). Almost 90% of the pathologist responders indicated they include the NIH activity/chronicity indices in biopsy reports. Reasons for not adding the activity/chronicity indices were given as responses to open-ended questions: (1) clinicians usually do not ask for activity/chronicity indices—(2) the extra workload outweighs the perceived clinical benefit—(3) a quantitative description of the degree of activity/chronicity is less time-consuming and more understandable than the indices—(4) reproducibility is not guaranteed. The average grade on the utility of the activity/chronicity indices was 7 (IQR 7-9) on a scale from 0 (not useful) to 10 (extremely useful). Pathologists rated treating clinicians' understanding of kidney biopsy reports and lupus classification at light microscopy with an average score of 8 (IQR 7-9) on a scale from 0 (no understanding) to 10 (complete understanding). Suggested improvements of the 2018 ISN/RPS lupus nephritis classification were further standardization and simplicity. Clearer definitions were requested for: class III vs class IV; segmental sclerosis vs fibrotic crescents; the role of globally sclerosed glomeruli; lupus-like entities; and lupus podocytopathy. Furthermore, there was a request on guidelines for new parameters: e.g., focus on extraglomerular involvement, laboratory and clinical features and biomarkers; CD68+ staining to detect endocapillary hypercellularity; and EXT1/2 immunostaining. Conclusion Our survey shows the 2018 ISN/RPS lupus nephritis classification is widely used in everyday practice by pathologists, and well known by clinicians. The results obtained by the RPS survey serve as a starting point for the new Working Group of lupus nephritis classification. We will consider the ideas and suggestions of the international community to evaluate biopsies from extensively phenotyped cohorts, to modify the classification and to facilitate improved targeted treatment decisions—a current unmet need.

#2423 Molecular classification of lupus nephritis based on immune pathway profiling—correlations with histological class and clinical manifestations

Abstract Background and Aims The ISN/RPS 2003 classification is the current gold standard histopathological classification for lupus nephritis (LN), and has important roles in guiding treatment and predicting prognosis. The classification system, largely based on light microscopy features, does not inform the underlying immunological aberrations. A classification based on molecular typing of immune-related pathways may better stratify LN patients and help select treatments that target specific immune cascades. Method The normalized expression matrix and clinical data were extracted from Gene Expression Omnibus database, Human Cell Atlas (Hacohen-Human-CELseq2) and Nephroseq. Four chip datasets (GSE127797, 32591, 69438 and 37463) and one single cell dataset were analyzed. Sample gene set enrichment analysis (ssGSEA) method was used to derive immune phenotype with gene expression data. The hierarchical clustering with ssGSEA scores of selected gene sets was used to classify LN tissues. Functional characterization and protein-protein interactions of the differential expression genes (DEGs) of three groups were compared with healthy donors using Metascape (absolute log FC &amp;gt; 1 and adjusted P-value &amp;lt; .05). Kruskal-Walli test and chi-square test were performed to compare the clinical parameters of the different groups. Results We analyzed 96 immune-related gene sets in the glomerular and tubulointerstitial regions of LN kidney biopsy samples. The immune-related genes were categorized into three clusters (Immune-Mild, Immune-Moderate and Immune-Severe) based on the degree of immune pathway activation. Interferon-stimulated genes are a hallmark of LN, with high signals in all three clusters. When compared to other groups, the Immune-Mild cluster from the glomerular region showed no T-cell activation or T cell receptor pathway. Also, the Immune-Mild (tubulointerstitial) group also showed minimal lymphocyte activation or migration. Protein-protein interactions networks with MOCDE components further suggested that STAT1 and ISG15 the hub genes in different clusters. The gene expressions of hub genes were then compared among epithelial cells and 21 immune cells from LN and heath controls. Immune clusters are correlated with clinical phenotypes, in which Immune-Mild Group (Glomerular) showed high prevalence of Class II and V LN, and Immune-Mild (Tubulointerstitial) showed more Class V LN. Immune-Severe group also showed higher degree of proteinuria and blood pressure but lower eGFR compared to Immune-Moderate and Immune-Mild groups. Conclusion A new molecular classification of LN based on the degree of immune pathway activation shows correlation with clinical manifestations and may potentially guide personalized therapy in LN patients.

#464 Survey results on the utility of the Renal Risk Score and Histopathological Classification of ANCA-glomerulonephritis

Abstract Background and Aims The importance of kidney biopsy in ANCA-associated glomerulonephritis (AAV) is widely recognized: both the Berden classification and the ANCA renal risk score (ARRS) are widely used and reliable predictors of prognosis. AAV phenotypes, outcomes and treatment have changed over the last decades: whether and how the existing classification and ARRS need to be updated is a matter of discussion. A survey to assess the current use of the classification and ARRS in clinical practice was conducted on behalf of the newly formed “Working Group on classifications and risk scores”. Method An online survey was sent to the active members of the European Vasculitis Society (EUVAS) between September 13th and September 27th, 2023. The survey used an internet platform, contained 12 multiple-choice and 2 open questions and results were analyzed anonymously. Results 76 of 295 EUVAS members replied to the questionnaire: 40 (52.6%) were nephrologists, 25 (32.9%) rheumatologists, 4 (5.3%) pathologists and 7 (9.3%) “others”. Their level of expertise was high, with 61 (80.3%) physicians encountering more than 10 new-diagnosed AAV patients per year. A total of 64 (84.2%) responders usually performed a kidney biopsy in suspected AAV patients, considering the pathological examination a useful tool for prognosis (89.5%), diagnosis (85.5%) and treatment (81.6%). More than 90% of responders indicated they discuss kidney biopsy results in a multidisciplinary meeting involving both clinicians and pathologists. Both the Berden classification and the ARRS are known by around 70% of experts in the field; the first being commonly used by clinicians in everyday practice (70.1% vs 52.5%, p = 0.08). Both systems were considered useful on a scale from 0 to 10, with a median grade of 7 [(IQR 5-8) vs (IQR 6-8), p = 0.21]. Suggested improvements were further standardizations and simplicity. New parameters (e.g., extraglomerular involvement, laboratory and clinical features) and biomarkers should be considered for future inclusion. Conclusion A survey assessing the current use of the Berden classification and ARRS in clinical practice showed that both systems are widely used among vasculitis clinicians and that the ideal system should be able to guide treatment decisions. The validation of pathological systems for targeted treatment decisions is an unmet need.

Automated histopathological detection and classification of lung cancer with an image pre-processing pipeline and spatial attention with deep neural networks

Automated histopathological detection and classification of lung cancer with an image pre-processing pipeline and spatial attention with deep neural networks

Lung and Colon Cancer Histopathological Image Classification Using 1D Convolutional Channel-based Attention Networks

Lung and Colon cancer are the leading diseases of death and disability in humans caused by a combination of genetic diseases and biochemical abnormalities. If these are diagnosed in their early stages, they can not be spread in organs and negatively impact human life. Many deep-learning networks have recently been proposed to detect and classify these malignancies. However, incorrect detection or misclassification of these fatal diseases can significantly affect an individual's health and well-being. This paper introduces a novel, cost-effective, and mobile-embedded architecture to diagnose and classify Lung squamous cell carcinomas and adenocarcinomas of the lung and colon from digital pathology images. Extensive experiment shows that our proposed modifications achieve 100% testing results for lung, colon, and lung-and-colon cancer detection. Our novel architecture takes around 0.65 million trainable parameters and around 6.4 million flops to achieve the best lung and colon cancer detection performance. Compared with the other results, our proposed architecture shows state-of-the-art performance.

Testicular histopathology and follicular stimulating hormone to predict fertility in nonobstructive azoospermia.

To examine the ability of testicular histopathology in Non-obstructive azoospermia (NOA) in predicting sperm retrieval rate (SR), sperm quality and assisted reproductive technology success. A retrospective study recruited clinically diagnosed NOA patients between 2007 and 2015. Testicular biopsy and conventional sperm extraction (TESE) were done concomitantly. Correlation between pathological categories, SR rate, sperm quality and success of intracytoplasmic sperm injection (ICSI) was studied. FSH was measured as a predictor of fertility. One hundred eighteen patients were recruited. Histopathological classification was hypospermatogenesis (HS) 45 (38%), maturation arrest (MA) 22(19%), Sertoli cell only syndrome (SCOS) 34 (29%) and normal spermatogenesis (NS) 17 (14%). FSH value was above normal level in 34 (76%) of HS, 19 (86%) of MA, 32 (94%) of SCOS and 5 (29%) of NS. Positive SR was obtained in 108 (92%) patients. The highest SR rate was seen in NS group 100% and the lowest was in SCOS 26 (77%). The worst sperm quality was found in SCOS as type C represents 46%, followed by MA 40% and HS 24%. Patients had ICSI following TESE had variable success rate as success of ICSI was seen (9/15) for HS, (0/7) for MA, (5/15) for SCOS and (8/9) for NS. FSH is strongly correlated to SR, quality of sperm and success of ICSI as positive SR in normal FSH patients was obtained in 28 (100%) of normal FSH, 70 (97%) of high FSH and 10 (56%) of double high FSH (p value < 001). The success of ICSI significantly correlates with FSH value as normal FSH has 77% success ICSI rate, high FSH (52%) and double high FSH (0%) (p value < 0.001). Testicular biopsy and histopathology findings in NOA are strongly correlated SR rate, quality of sperms, and success of ICSI. FSH is a strong noninvasive predictor of fertility in NOA patients.

Inception-v3 with reduce learning rate for optimization of lung cancer histopathology classification

Inception-v3 with reduce learning rate for optimization of lung cancer histopathology classification

Enhancing Histopathology Breast Cancer Detection and Classification with the Deep Ensemble Graph Network

Enhancing Histopathology Breast Cancer Detection and Classification with the Deep Ensemble Graph Network

Cancer population norms using a new value set for the SF-6Dv2 based on the preferences of patients with breast or colorectal cancer in Quebec.

Because health resources are limited, health programs should be compared to allow the most efficient ones to emerge. To that aim, health utility instruments have been developed to allow the calculation of quality-adjusted life-year (QALY). However, generic instruments, which can be used by any individual regardless of their health profile, typically consider the preferences of the general population when developing their value set. Consequently, they are often criticized for lacking sensitivity in certain domains, such as cancer. In response, the latest version of the Short Form 6-Dimension (SF-6Dv2) has been adapted to suit the preferences of patients with breast or colorectal cancer in the Canadian province of Quebec. By extension, our study's aim was to determine cancer population norms of utility among patients with breast or colorectal cancer in Quebec using the SF-6Dv2. To determine the cancer population norms, we exploited the data that were used in the development of a new value set for the SF-6Dv2. This value set was developed considering the preferences of patients with breast or colorectal cancer. Stratification by time of data collection (i.e., T1 and T2), sociodemographic variables (i.e., age, sex, body mass index, and self-reported health problems affecting quality of life), and clinical aspects (i.e., cancer site, histopathological classification, cancer stage at diagnosis, modality, and treatment characteristics) was performed. In 353 observations, patients were more likely to have negative utility scores at T1 than at T2. Males had higher mean utility scores than females considering type of cancer and comorbidities. Considering the SF-6Dv2's dimensions, more females than males reported having health issues, most which concerned physical functioning. Significant differences by sex surfaced for all dimensions except "Role Limitation" and "Mental health." Patients with multifocal cancer had the highest mean and median utility values in all cancer sites considered. Cancer population norms can serve as a baseline for interpreting the scores obtained by a given population in comparison to the situation of another group. In this way, our results can assist in comparing utility scores among cancer patients with different sociodemographic groups to other patients/populations groups. To our knowledge, our identified utility norms are the first for patients with breast or colorectal cancer from Quebec.