Diagnosis of endometrium hyperplasia and screening of endometrial intraepithelial neoplasia in histopathological images using a global-to-local multi-scale convolutional neural network

Deep Learning for Grading Endometrial Cancer

Endometrial cancer is the fourth most common cancer in women in the United States; the lifetime risk for developing this disease is approximately 2.8%. Precise histologic evaluation and molecular classification of endometrial cancer are important for effective patient management and determining the best treatment modalities. This study introduces EndoNet, which uses convolutional neural networks for extracting histologic features and a vision transformer for aggregating these features and classifying slides based on their visual characteristics into high- and low-grade cases. The model was trained on 929 digitized hematoxylin and eosin-stained whole-slide images of endometrial cancer from hysterectomy cases at Dartmouth-Health. It classifies these slides into low-grade (endometrioid grades 1 and 2) and high-grade (endometrioid carcinoma International Federation of Gynecology and Obstetrics grade 3, uterine serous carcinoma, or carcinosarcoma) categories. EndoNet was evaluated on an internal test set of 110 patients and an external test set of 100 patients from The Cancer Genome Atlas public database. The model achieved a weighted average F1 score of 0.91 (95% CI, 0.86 to 0.95) and an area under the curve of 0.95 (95% CI, 0.89 to 0.99) on the internal test, and 0.86 (95% CI, 0.80 to 0.94) for F1 score and 0.86 (95% CI, 0.75 to 0.93) for area under the curve on the external test. Pending further validation, EndoNet has the potential to support pathologists without the need of manual annotations in classifying the grades of gynecologic pathology tumors.

Factors associated with interobserver variation amongst pathologists in the diagnosis of endometrial hyperplasia: A systematic review

OBJECTIVE

Reproducible diagnoses of endometrial hyperplasia (EH) remains challenging and has potential implications for patient management. This systematic review aimed to identify pathologist-specific factors associated with interobserver variation in the diagnosis and reporting of EH.

METHODS

Three electronic databases, namely MEDLINE, Embase and Web of Science, were searched from 1st January 2000 to 25th March 2023, using relevant key words and subject headings. Eligible studies reported on pathologist-specific factors or working practices influencing interobserver variation in the diagnosis of EH, using either the World Health Organisation (WHO) 2014 or 2020 classification or the endometrioid intraepithelial neoplasia (EIN) classification system. Quality assessment was undertaken using the QUADAS-2 tool, and findings were narratively synthesised.

RESULTS

Eight studies were identified. Interobserver variation was shown to be significant even amongst specialist gynaecological pathologists in most studies. Few studies investigated pathologist-specific characteristics, but pathologists were shown to have different diagnostic styles, with some more likely to under-diagnose and others likely to over-diagnose EH. Some novel working practices were identified, such as grading the "degree" of nuclear atypia and the incorporation of objective methods of diagnosis such as semi-automated quantitative image analysis/deep learning models.

CONCLUSIONS

This review highlighted the impact of pathologist-specific factors and working practices in the accurate diagnosis of EH, although few studies have been conducted. Further research is warranted in the development of more objective criteria that could improve reproducibility in EH diagnostic reporting, as well as determining the applicability of novel methods such as grading the degree of nuclear atypia in clinical settings.

Radiomics model based on multi-sequence MRI for preoperative prediction of ki-67 expression levels in early endometrial cancer

To validate a radiomics model based on multi-sequence magnetic resonance imaging (MRI) in predicting the ki-67 expression levels in early-stage endometrial cancer, 131 patients with early endometrial cancer who had undergone pathological examination and preoperative MRI scan were retrospectively enrolled and divided into two groups based on the ki-67 expression levels. The radiomics features were extracted from the T2 weighted imaging (T2WI), dynamic contrast enhanced T1 weighted imaging (DCE-T1WI), and apparent diffusion coefficient (ADC) map and screened using the Pearson correlation coefficients (PCC). A multi-layer perceptual machine and fivefold cross-validation were used to construct the radiomics model. The receiver operating characteristic (ROC) curves analysis, calibration curves, and decision curve analysis (DCA) were used to assess the models. The combined multi-sequence radiomics model of T2WI, DCE-T1WI, and ADC map showed better discriminatory powers than those using only one sequence. The combined radiomics models with multi-sequence fusions achieved the highest area under the ROC curve (AUC). The AUC value of the validation set was 0.852, with an accuracy of 0.827, sensitivity of 0.844, specificity of 0.773, and precision of 0.799. In conclusion, the combined multi-sequence MRI based radiomics model enables preoperative noninvasive prediction of the ki-67 expression levels in early endometrial cancer. This provides an objective imaging basis for clinical diagnosis and treatment.

A self-supervised classification model for endometrial diseases

PURPOSE

Ultrasound imaging is the preferred method for the early diagnosis of endometrial diseases because of its non-invasive nature, low cost, and real-time imaging features. However, the accurate evaluation of ultrasound images relies heavily on the experience of radiologist. Therefore, a stable and objective computer-aided diagnostic model is crucial to assist radiologists in diagnosing endometrial lesions.

METHODS

Transvaginal ultrasound images were collected from multiple hospitals in Quzhou city, Zhejiang province. The dataset comprised 1875 images from 734 patients, including cases of endometrial polyps, hyperplasia, and cancer. Here, we proposed a based self-supervised endometrial disease classification model (BSEM) that learns a joint unified task (raw and self-supervised tasks) and applies self-distillation techniques and ensemble strategies to aid doctors in diagnosing endometrial diseases.

RESULTS

The performance of BSEM was evaluated using fivefold cross-validation. The experimental results indicated that the BSEM model achieved satisfactory performance across indicators, with scores of 75.1%, 87.3%, 76.5%, 73.4%, and 74.1% for accuracy, area under the curve, precision, recall, and F1 score, respectively. Furthermore, compared to the baseline models ResNet, DenseNet, VGGNet, ConvNeXt, VIT, and CMT, the BSEM model enhanced accuracy, area under the curve, precision, recall, and F1 score in 3.3-7.9%, 3.2-7.3%, 3.9-8.5%, 3.1-8.5%, and 3.3-9.0%, respectively.

CONCLUSION

The BSEM model is an auxiliary diagnostic tool for the early detection of endometrial diseases revealed by ultrasound and helps radiologists to be accurate and efficient while screening for precancerous endometrial lesions.

A Radiomic-Based Machine Learning Model Predicts Endometrial Cancer Recurrence Using Preoperative CT Radiomic Features: A Pilot Study

BACKGROUND

Current prognostic models lack the use of pre-operative CT images to predict recurrence in endometrial cancer (EC) patients. Our study aimed to investigate the potential of radiomic features extracted from pre-surgical CT scans to accurately predict disease-free survival (DFS) among EC patients.

METHODS

Contrast-Enhanced CT (CE-CT) scans from 81 EC cases were used to extract the radiomic features from semi-automatically contoured volumes of interest. We employed a 10-fold cross-validation approach with a 6:4 training to test set and utilized data augmentation and balancing techniques. Univariate analysis was applied for feature reduction leading to the development of three distinct machine learning (ML) models for the prediction of DFS: LASSO-Cox, CoxBoost and Random Forest (RFsrc).

RESULTS

In the training set, the ML models demonstrated AUCs ranging from 0.92 to 0.93, sensitivities from 0.96 to 1.00 and specificities from 0.77 to 0.89. In the test set, AUCs ranged from 0.86 to 0.90, sensitivities from 0.89 to 1.00 and specificities from 0.73 to 0.90. Patients classified as having a high recurrence risk prediction by ML models exhibited significantly worse DSF (p-value < 0.001) across all models.

CONCLUSIONS

Our findings demonstrate the potential of radiomics in predicting EC recurrence. While further validation studies are needed, our results underscore the promising role of radiomics in forecasting EC outcomes.

Automated Prognostic Assessment of Endometrial Hyperplasia for Progression Risk Evaluation Using Artificial Intelligence

Endometrial hyperplasia is a precursor to endometrial cancer, characterized by excessive proliferation of glands that is distinguishable from normal endometrium. Current classifications define 2 types of EH, each with a different risk of progression to endometrial cancer. However, these schemes are based on visual assessments and, therefore, subjective, possibly leading to overtreatment or undertreatment. In this study, we developed an automated artificial intelligence tool (ENDOAPP) for the measurement of morphologic and cytologic features of endometrial tissue using the software Visiopharm. The ENDOAPP was used to extract features from whole-slide images of PAN-CK⁺-stained formalin-fixed paraffin-embedded tissue sections from 388 patients diagnosed with endometrial hyperplasia between 1980 and 2007. Follow-up data were available for all patients (mean = 140 months). The most prognostic features were identified by a logistic regression model and used to assign a low-risk or high-risk progression score. Performance of the ENDOAPP was assessed for the following variables: images from 2 different scanners (Hamamatsu XR and S60) and automated placement of a region of interest versus manual placement by an operator. Then, the performance of the application was compared with that of current classification schemes: WHO94, WHO20, and EIN, and the computerized-morphometric risk classification method: D-score. The most significant prognosticators were percentage stroma and the standard deviation of the lesser diameter of epithelial nuclei. The ENDOAPP had an acceptable discriminative power with an area under the curve of 0.765. Furthermore, strong to moderate agreement was observed between manual operators (intraclass correlation coefficient: 0.828) and scanners (intraclass correlation coefficient: 0.791). Comparison of the prognostic capability of each classification scheme revealed that the ENDOAPP had the highest accuracy of 88%-91% alongside the D-score method (91%). The other classification schemes had an accuracy between 83% and 87%. This study demonstrated the use of computer-aided prognosis to classify progression risk in EH for improved patient treatment.

Detection of malignancy in whole slide images of endometrial cancer biopsies using artificial intelligence

In this study we use artificial intelligence (AI) to categorise endometrial biopsy whole slide images (WSI) from digital pathology as either "malignant", "other or benign" or "insufficient". An endometrial biopsy is a key step in diagnosis of endometrial cancer, biopsies are viewed and diagnosed by pathologists. Pathology is increasingly digitised, with slides viewed as images on screens rather than through the lens of a microscope. The availability of these images is driving automation via the application of AI. A model that classifies slides in the manner proposed would allow prioritisation of these slides for pathologist review and hence reduce time to diagnosis for patients with cancer. Previous studies using AI on endometrial biopsies have examined slightly different tasks, for example using images alongside genomic data to differentiate between cancer subtypes. We took 2909 slides with "malignant" and "other or benign" areas annotated by pathologists. A fully supervised convolutional neural network (CNN) model was trained to calculate the probability of a patch from the slide being "malignant" or "other or benign". Heatmaps of all the patches on each slide were then produced to show malignant areas. These heatmaps were used to train a slide classification model to give the final slide categorisation as either "malignant", "other or benign" or "insufficient". The final model was able to accurately classify 90% of all slides correctly and 97% of slides in the malignant class; this accuracy is good enough to allow prioritisation of pathologists' workload.

Clinical-grade endometrial cancer detection system via whole-slide images using deep learning

The accurate pathological diagnosis of endometrial cancer (EC) improves the curative effect and reduces the mortality rate. Deep learning has demonstrated expert-level performance in pathological diagnosis of a variety of organ systems using whole-slide images (WSIs). It is urgent to build the deep learning system for endometrial cancer detection using WSIs. The deep learning model was trained and validated using a dataset of 601 WSIs from PUPH. The model performance was tested on three independent datasets containing a total of 1,190 WSIs. For the retrospective test, we evaluated the model performance on 581 WSIs from PUPH. In the prospective study, 317 consecutive WSIs from PUPH were collected from April 2022 to May 2022. To further evaluate the generalizability of the model, 292 WSIs were gathered from PLAHG as part of the external test set. The predictions were thoroughly analyzed by expert pathologists. The model achieved an area under the receiver operating characteristic curve (AUC), sensitivity, and specificity of 0.928, 0.924, and 0.801, respectively, on 1,190 WSIs in classifying EC and non-EC. On the retrospective dataset from PUPH/PLAGH, the model achieved an AUC, sensitivity, and specificity of 0.948/0.971, 0.928/0.947, and 0.80/0.938, respectively. On the prospective dataset, the AUC, sensitivity, and specificity were, in order, 0.933, 0.934, and 0.837. Falsely predicted results were analyzed to further improve the pathologists’ confidence in the model. The deep learning model achieved a high degree of accuracy in identifying EC using WSIs. By pre-screening the suspicious EC regions, it would serve as an assisted diagnostic tool to improve working efficiency for pathologists.