Automated Detection of Endometrial Polyps from Hysteroscopic Videos Using Deep Learning.
Ontology highlight
ABSTRACT:
SUBMITTER: Zhao A
PROVIDER: S-EPMC10137953 | biostudies-literature | 2023 Apr
REPOSITORIES: biostudies-literature
Publications
Automated Detection of Endometrial Polyps from Hysteroscopic Videos Using Deep Learning.
Zhao Aihua A Du Xin X Yuan Suzhen S Shen Wenfeng W Zhu Xin X Wang Wenwen W
Diagnostics (Basel, Switzerland) 20230413 8
Endometrial polyps are common gynecological lesions. The standard treatment for this condition is hysteroscopic polypectomy. However, this procedure may be accompanied by misdetection of endometrial polyps. To improve the diagnostic accuracy and reduce the risk of misdetection, a deep learning model based on YOLOX is proposed to detect endometrial polyps in real time. Group normalization is employed to improve its performance with large hysteroscopic images. In addition, we propose a video adjac ...[more]
Similar Datasets
Project description:ObjectiveTo assess the performance of convolutional neural networks (CNNs) for automated detection of keratoconus (KC) in standalone Scheimpflug-based dynamic corneal deformation videos.DesignRetrospective cohort study.ParticipantsWe retrospectively analyzed datasets with records of 734 nonconsecutive, refractive surgery candidates, and patients with unilateral or bilateral KC.MethodsWe first developed a video preprocessing pipeline to translate dynamic corneal deformation videos into 3-dimensional pseudoimage representations and then trained a CNN to directly identify KC from pseudoimages. We calculated the model's KC probability score cut-off and evaluated the performance by subjective and objective accuracy metrics using 2 independent datasets.Main outcome measuresArea under the receiver operating characteristics curve (AUC), accuracy, specificity, sensitivity, and KC probability score.ResultsThe model accuracy on the test subset was 0.89 with AUC of 0.94. Based on the external validation dataset, the AUC and accuracy of the CNN model for detecting KC were 0.93 and 0.88, respectively.ConclusionsOur deep learning-based approach was highly sensitive and specific in separating normal from keratoconic eyes using dynamic corneal deformation videos at levels that may prove useful in clinical practice.Financial disclosuresProprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.
| S-EPMC10587634 | biostudies-literature
Project description:BackgroundAn endometrial polyp is a frequently encountered gynecologic disease with abnormal uterine bleeding and infertility being the two common presenting problems, and hysteroscopic polypectomy is an effective method to remove them. The postoperative polyp recurrence might result in reappearance of abnormal uterine bleeding or infertility, whereas factors influencing the postoperative recurrence potential have limited data.MethodsThis case-series report included 168 premenopausal women who suffered from endometrial polyps and underwent hysteroscopic polypectomy. All of them were awaiting a future pregnancy. Office hysteroscopy was done before and after hysteroscopic polypectomy, in which preoperative hysteroscopy examined the number, type, and location of endometrial polyps, and postoperative hysteroscopy checked the polyp recurrence. Surgical indications, either infertility or the presentation of abnormal uterine bleeding, and follow-up duration were recorded.ResultsSeventy-three out of 168 (43%) women had polyp recurrence after hysteroscopic polypectomy. Multivariate logistic regression analysis revealed that more endometrial polyps (P = 0.015) and longer duration of follow-up (P = 0.004) were significantly associated with an increased risk of postoperative polyp recurrence. The type of endometrial polyps was not correlated with polyp recurrence potential, whereas pedunculated type endometrial polyps were closely related to the presentation of abnormal uterine bleeding (P = 0.001).ConclusionsA higher number of endometrial polyps and longer follow-up duration are associated with a greater potential of polyp recurrence after hysteroscopic polypectomy.
| S-EPMC4676604 | biostudies-literature
Project description:ImportanceCompetence in cataract surgery is a public health necessity, and videos of cataract surgery are routinely available to educators and trainees but currently are of limited use in training. Machine learning and deep learning techniques can yield tools that efficiently segment videos of cataract surgery into constituent phases for subsequent automated skill assessment and feedback.ObjectiveTo evaluate machine learning and deep learning algorithms for automated phase classification of manually presegmented phases in videos of cataract surgery.Design, setting, and participantsThis was a cross-sectional study using a data set of videos from a convenience sample of 100 cataract procedures performed by faculty and trainee surgeons in an ophthalmology residency program from July 2011 to December 2017. Demographic characteristics for surgeons and patients were not captured. Ten standard labels in the procedure and 14 instruments used during surgery were manually annotated, which served as the ground truth.ExposuresFive algorithms with different input data: (1) a support vector machine input with cross-sectional instrument label data; (2) a recurrent neural network (RNN) input with a time series of instrument labels; (3) a convolutional neural network (CNN) input with cross-sectional image data; (4) a CNN-RNN input with a time series of images; and (5) a CNN-RNN input with time series of images and instrument labels. Each algorithm was evaluated with 5-fold cross-validation.Main outcomes and measuresAccuracy, area under the receiver operating characteristic curve, sensitivity, specificity, and precision.ResultsUnweighted accuracy for the 5 algorithms ranged between 0.915 and 0.959. Area under the receiver operating characteristic curve for the 5 algorithms ranged between 0.712 and 0.773, with small differences among them. The area under the receiver operating characteristic curve for the image-only CNN-RNN (0.752) was significantly greater than that of the CNN with cross-sectional image data (0.712) (difference, -0.040; 95% CI, -0.049 to -0.033) and the CNN-RNN with images and instrument labels (0.737) (difference, 0.016; 95% CI, 0.014 to 0.018). While specificity was uniformly high for all phases with all 5 algorithms (range, 0.877 to 0.999), sensitivity ranged between 0.005 (95% CI, 0.000 to 0.015) for the support vector machine for wound closure (corneal hydration) and 0.974 (95% CI, 0.957 to 0.991) for the RNN for main incision. Precision ranged between 0.283 and 0.963.Conclusions and relevanceTime series modeling of instrument labels and video images using deep learning techniques may yield potentially useful tools for the automated detection of phases in cataract surgery procedures.
| S-EPMC6450320 | biostudies-literature
Project description:The evaluation of scratching behavior is important in experimental animals because there is significant interest in elucidating mechanisms and developing medications for itching. The scratching behavior is classically quantified by human observation, but it is labor-intensive and has low throughput. We previously established an automated scratching detection method using a convolutional recurrent neural network (CRNN). The established CRNN model was trained by white mice (BALB/c), and it could predict their scratching bouts and duration. However, its performance in black mice (C57BL/6) is insufficient. Here, we established a model for black mice to increase prediction accuracy. Scratching behavior in black mice was elicited by serotonin administration, and their behavior was recorded using a video camera. The videos were carefully observed, and each frame was manually labeled as scratching or other behavior. The CRNN model was trained using the labels and predicted the first-look videos. In addition, posterior filters were set to remove unlikely short predictions. The newly trained CRNN could sufficiently detect scratching behavior in black mice (sensitivity, 98.1%; positive predictive rate, 94.0%). Thus, our established CRNN and posterior filter successfully predicted the scratching behavior in black mice, highlighting that our workflow can be useful, regardless of the mouse strain.
| S-EPMC9352956 | biostudies-literature
Project description:ObjectiveAsherman's syndrome (AS) is a significant cause of subfertility in women from developing countries. Over 80% of AS cases in these regions are linked to dilation and curettage (D&C) procedures following pregnancy. The incidence of AS in patients with infertility and recurrent miscarriage can be as high as 10%, while the pregnancy rate in cases of moderate to severe adhesions can be as low as 34%. We aimed to establish a hysteroscopic artificial intelligence system using image-deep-learning algorithms for fertility assessment.MethodsThis diagnostic study included 555 cases with 4922 hysteroscopic images from a Chinese intrauterine adhesions cohort clinical database (NCT05381376). The study evaluated two image-deep-learning algorithms' effectiveness in predicting pregnancy within one year, using AUCs and decision curve analysis. The models' performance was evaluated for two-year prediction via concordance index and cumulative time-dependent ROC. A quantifiable visualization panel of the system was established.ResultsThe proportional hazard CNN system accurately predicted conception, with AUCs of 0.982, 0.992, and 0.990 in three randomly assigned datasets, superior to the InceptionV3 framework, and achieved a net benefit of 69.4% for subfertility assessment. The system fitted well with c-indexes of 0.920-0.940 and was time-stable. The quantifiable visualization panel displayed four intrauterine pathologies intuitively. The performance was comparable to senior hysteroscopists, with a kappa value of 0.84-0.89.ConclusionsThe CNN based on the proportional hazard approach accurately assesses fertility postoperatively. The quantifiable visualization panel could assist in intrauterine pathologies assessment, optimize treatment strategies, and achieve individualized and cost-efficient practices.
| S-EPMC11921166 | biostudies-literature
Project description:The color of particular parts of a flower is often employed as one of the features to differentiate between flower types. Thus, color is also used in flower-image classification. Color labels, such as 'green', 'red', and 'yellow', are used by taxonomists and lay people alike to describe the color of plants. Flower image datasets usually only consist of images and do not contain flower descriptions. In this research, we have built a flower-image dataset, especially regarding orchid species, which consists of human-friendly textual descriptions of features of specific flowers, on the one hand, and digital photographs indicating how a flower looks like, on the other hand. Using this dataset, a new automated color detection model was developed. It is the first research of its kind using color labels and deep learning for color detection in flower recognition. As deep learning often excels in pattern recognition in digital images, we applied transfer learning with various amounts of unfreezing of layers with five different neural network architectures (VGG16, Inception, Resnet50, Xception, Nasnet) to determine which architecture and which scheme of transfer learning performs best. In addition, various color scheme scenarios were tested, including the use of primary and secondary color together, and, in addition, the effectiveness of dealing with multi-class classification using multi-class, combined binary, and, finally, ensemble classifiers were studied. The best overall performance was achieved by the ensemble classifier. The results show that the proposed method can detect the color of flower and labellum very well without having to perform image segmentation. The result of this study can act as a foundation for the development of an image-based plant recognition system that is able to offer an explanation of a provided classification.
| S-EPMC8550396 | biostudies-literature
Project description:Fabry disease is a congenital lysosomal storage disease, and most of these cases develop organ damage in middle age. There are some promising therapeutic options for this disorder, which can stabilize the progression of the disease. However, a long delay in diagnosis prevents early intervention, resulting in treatment failure. Because Fabry disease is a rare disease, it is not well recognized and disease specific screening tests are rarely performed. Hence, a novel approach to for detecting patients with a widely practiced clinical test is crucial for the early detection of the disease. Recently, decision support systems based on artificial intelligence (AI) have been developed in many clinical fields. However, the construction of these models requires datasets from a large number of samples; this aspect is one of the main obstacles in AI-based approaches for rare diseases. In this study, with a novel image amplification method to construct the dataset for AI-model training, we built the deep neural-network model to detect Fabry cases from their urine samples. Sensitivity, specificity, and the AUC of the models on validation dataset were 0.902 (95% CI, 0.900-0.903), 0.977 (0.950-0.980), and 0.968 (0.964-0.972), respectively. This model could also extract disease-specific findings that are interpretable with human recognition. These results indicate that we can apply novel AI models for rare diseases based on this image amplification method we developed. We expect this approach could contribute to the diagnosis of Fabry disease.SynopsisThis is the first reported AI-based decision support system to detect undiagnosed Fabry cases, and our new image amplification method will contribute to the AI models for other rare disorders.
| S-EPMC9523392 | biostudies-literature
Project description:BackgroundThe aim of this study was to develop and evaluate a deep neural network model in the automated detection of pulmonary embolism (PE) from computed tomography pulmonary angiograms (CTPAs) using only weakly labelled training data.MethodsWe developed a deep neural network model consisting of two parts: a convolutional neural network architecture called InceptionResNet V2 and a long-short term memory network to process whole CTPA stacks as sequences of slices. Two versions of the model were created using either chest X-rays (Model A) or natural images (Model B) as pre-training data. We retrospectively collected 600 CTPAs to use in training and validation and 200 CTPAs to use in testing. CTPAs were annotated only with binary labels on both stack- and slice-based levels. Performance of the models was evaluated with ROC and precision-recall curves, specificity, sensitivity, accuracy, as well as positive and negative predictive values.ResultsBoth models performed well on both stack- and slice-based levels. On the stack-based level, Model A reached specificity and sensitivity of 93.5% and 86.6%, respectively, outperforming Model B slightly (specificity 90.7% and sensitivity 83.5%). However, the difference between their ROC AUC scores was not statistically significant (0.94 vs 0.91, p = 0.07).ConclusionsWe show that a deep learning model trained with a relatively small, weakly annotated dataset can achieve excellent performance results in detecting PE from CTPAs.
| S-EPMC8919639 | biostudies-literature
Project description:The COVID-19 pandemic has caused a worldwide catastrophe and widespread devastation that reeled almost all countries. The pandemic has mounted pressure on the existing healthcare system and caused panic and desperation. The gold testing standard for COVID-19 detection, reverse transcription-polymerase chain reaction (RT-PCR), has shown its limitations with 70% accuracy, contributing to the incorrect diagnosis that exaggerated the complexities and increased the fatalities. The new variations further pose unseen challenges in terms of their diagnosis and subsequent treatment. The COVID-19 virus heavily impacts the lungs and fills the air sacs with fluid causing pneumonia. Thus, chest X-ray inspection is a viable option if the inspection detects COVID-19-induced pneumonia, hence confirming the exposure of COVID-19. Artificial intelligence and machine learning techniques are capable of examining chest X-rays in order to detect patterns that can confirm the presence of COVID-19-induced pneumonia. This research used CNN and deep learning techniques to detect COVID-19-induced pneumonia from chest X-rays. Transfer learning with fine-tuning ensures that the proposed work successfully classifies COVID-19-induced pneumonia, regular pneumonia, and normal conditions. Xception, Visual Geometry Group 16, and Visual Geometry Group 19 are used to realize transfer learning. The experimental results were promising in terms of precision, recall, F1 score, specificity, false omission rate, false negative rate, false positive rate, and false discovery rate with a COVID-19-induced pneumonia detection accuracy of 98%. Experimental results also revealed that the proposed work has not only correctly identified COVID-19 exposure but also made a distinction between COVID-19-induced pneumonia and regular pneumonia, as the latter is a very common disease, while COVID-19 is more lethal. These results mitigated the concern and overlap in the diagnosis of COVID-19-induced pneumonia and regular pneumonia. With further integrations, it can be employed as a potential standard model in differentiating the various lung-related infections, including COVID-19.
| S-EPMC9351538 | biostudies-literature
Project description:Background and aimsIdentification and photo-documentation of the ileocecal valve (ICV) and appendiceal orifice (AO) confirm completeness of colonoscopy examinations. We aimed to develop and test a deep convolutional neural network (DCNN) model that can automatically identify ICV and AO, and differentiate these landmarks from normal mucosa and colorectal polyps.MethodsWe prospectively collected annotated full-length colonoscopy videos of 318 patients undergoing outpatient colonoscopies. We created three nonoverlapping training, validation, and test data sets with 25,444 unaltered frames extracted from the colonoscopy videos showing four landmarks/image classes (AO, ICV, normal mucosa, and polyps). A DCNN classification model was developed, validated, and tested in separate data sets of images containing the four different landmarks.ResultsAfter training and validation, the DCNN model could identify both AO and ICV in 18 out of 21 patients (85.7%). The accuracy of the model for differentiating AO from normal mucosa, and ICV from normal mucosa were 86.4% (95% CI 84.1% to 88.5%), and 86.4% (95% CI 84.1% to 88.6%), respectively. Furthermore, the accuracy of the model for differentiating polyps from normal mucosa was 88.6% (95% CI 86.6% to 90.3%).ConclusionThis model offers a novel tool to assist endoscopists with automated identification of AO and ICV during colonoscopy. The model can reliably distinguish these anatomical landmarks from normal mucosa and colorectal polyps. It can be implemented into automated colonoscopy report generation, photo-documentation, and quality auditing solutions to improve colonoscopy reporting quality.
| S-EPMC10395661 | biostudies-literature