Deepometry, a framework for applying supervised and weakly supervised deep learning to imaging cytometry.
Ontology highlight
ABSTRACT:
SUBMITTER: Doan M
PROVIDER: S-EPMC8506936 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:The application of machine learning approaches to imaging flow cytometry (IFC) data has the potential to transform the diagnosis of hematological diseases. However, the need for manually labeled single-cell images for machine learning model training has severely limited its clinical application. To address this, we present iCellCnn, a weakly supervised deep learning approach for label-free IFC-based blood diagnostics. We demonstrate the capability of iCellCnn to achieve diagnosis of Sézary syndrome (SS) from patient samples on the basis of bright-field IFC images of T cells obtained after fluorescence-activated cell sorting of human peripheral blood mononuclear cell specimens. With a sample size of four healthy donors and five SS patients, iCellCnn achieved a 100% classification accuracy. As iCellCnn is not restricted to the diagnosis of SS, we expect such weakly supervised approaches to tap the diagnostic potential of IFC by providing automatic data-driven diagnosis of diseases with so-far unknown morphological manifestations.
| S-EPMC9017143 | biostudies-literature
Project description:The yield of cereal crops such as sorghum (Sorghum bicolor L. Moench) depends on the distribution of crop-heads in varying branching arrangements. Therefore, counting the head number per unit area is critical for plant breeders to correlate with the genotypic variation in a specific breeding field. However, measuring such phenotypic traits manually is an extremely labor-intensive process and suffers from low efficiency and human errors. Moreover, the process is almost infeasible for large-scale breeding plantations or experiments. Machine learning-based approaches like deep convolutional neural network (CNN) based object detectors are promising tools for efficient object detection and counting. However, a significant limitation of such deep learning-based approaches is that they typically require a massive amount of hand-labeled images for training, which is still a tedious process. Here, we propose an active learning inspired weakly supervised deep learning framework for sorghum head detection and counting from UAV-based images. We demonstrate that it is possible to significantly reduce human labeling effort without compromising final model performance (R 2 between human count and machine count is 0.88) by using a semitrained CNN model (i.e., trained with limited labeled data) to perform synthetic annotation. In addition, we also visualize key features that the network learns. This improves trustworthiness by enabling users to better understand and trust the decisions that the trained deep learning model makes.
| S-EPMC7706102 | biostudies-literature
Project description:Identification of somatic mutations in tumor samples is commonly based on statistical methods in combination with heuristic filters. Here we develop VarNet, an end-to-end deep learning approach for identification of somatic variants from aligned tumor and matched normal DNA reads. VarNet is trained using image representations of 4.6 million high-confidence somatic variants annotated in 356 tumor whole genomes. We benchmark VarNet across a range of publicly available datasets, demonstrating performance often exceeding current state-of-the-art methods. Overall, our results demonstrate how a scalable deep learning approach could augment and potentially supplant human engineered features and heuristic filters in somatic variant calling.
| S-EPMC9307817 | biostudies-literature
Project description:Deep learning and computer vision have become emerging tools for diseased plant phenotyping. Most previous studies focused on image-level disease classification. In this paper, pixel-level phenotypic feature (the distribution of spot) was analyzed by deep learning. Primarily, a diseased leaf dataset was collected and the corresponding pixel-level annotation was contributed. A dataset of apple leaves samples was used for training and optimization. Another set of grape and strawberry leaf samples was used as an extra testing dataset. Then, supervised convolutional neural networks were adopted for semantic segmentation. Moreover, the possibility of weakly supervised models for disease spot segmentation was also explored. Grad-CAM combined with ResNet-50 (ResNet-CAM), and that combined with a few-shot pretrained U-Net classifier for weakly supervised leaf spot segmentation (WSLSS), was designed. They were trained using image-level annotations (healthy versus diseased) to reduce the cost of annotation work. Results showed that the supervised DeepLab achieved the best performance (IoU = 0.829) on the apple leaf dataset. The weakly supervised WSLSS achieved an IoU of 0.434. When processing the extra testing dataset, WSLSS realized the best IoU of 0.511, which was even higher than fully supervised DeepLab (IoU = 0.458). Although there was a certain gap in IoU between the supervised models and weakly supervised ones, WSLSS showed stronger generalization ability than supervised models when processing the disease types not involved in the training procedure. Furthermore, the contributed dataset in this paper could help researchers get a quick start on designing their new segmentation methods in future studies.
| S-EPMC10076051 | biostudies-literature
Project description:Conventional deep learning (DL) algorithm requires full supervision of annotating the region of interest (ROI) that is laborious and often biased. We aimed to develop a weakly-supervised DL algorithm that diagnosis breast cancer at ultrasound without image annotation. Weakly-supervised DL algorithms were implemented with three networks (VGG16, ResNet34, and GoogLeNet) and trained using 1000 unannotated US images (500 benign and 500 malignant masses). Two sets of 200 images (100 benign and 100 malignant masses) were used for internal and external validation sets. For comparison with fully-supervised algorithms, ROI annotation was performed manually and automatically. Diagnostic performances were calculated as the area under the receiver operating characteristic curve (AUC). Using the class activation map, we determined how accurately the weakly-supervised DL algorithms localized the breast masses. For internal validation sets, the weakly-supervised DL algorithms achieved excellent diagnostic performances, with AUC values of 0.92-0.96, which were not statistically different (all Ps > 0.05) from those of fully-supervised DL algorithms with either manual or automated ROI annotation (AUC, 0.92-0.96). For external validation sets, the weakly-supervised DL algorithms achieved AUC values of 0.86-0.90, which were not statistically different (Ps > 0.05) or higher (P = 0.04, VGG16 with automated ROI annotation) from those of fully-supervised DL algorithms (AUC, 0.84-0.92). In internal and external validation sets, weakly-supervised algorithms could localize 100% of malignant masses, except for ResNet34 (98%). The weakly-supervised DL algorithms developed in the present study were feasible for US diagnosis of breast cancer with well-performing localization and differential diagnosis.
| S-EPMC8692405 | biostudies-literature
Project description:The Gleason score contributes significantly in predicting prostate cancer outcomes and selecting the appropriate treatment option, which is affected by well-known inter-observer variations. We present a novel deep learning-based automated Gleason grading system that does not require extensive region-level manual annotations by experts and/or complex algorithms for the automatic generation of region-level annotations. A total of 6664 and 936 prostate needle biopsy single-core slides (689 and 99 cases) from two institutions were used for system discovery and validation, respectively. Pathological diagnoses were converted into grade groups and used as the reference standard. The grade group prediction accuracy of the system was 77.5% (95% confidence interval (CI): 72.3-82.7%), the Cohen's kappa score (κ) was 0.650 (95% CI: 0.570-0.730), and the quadratic-weighted kappa score (κquad) was 0.897 (95% CI: 0.815-0.979). When trained on 621 cases from one institution and validated on 167 cases from the other institution, the system's accuracy reached 67.4% (95% CI: 63.2-71.6%), κ 0.553 (95% CI: 0.495-0.610), and the κquad 0.880 (95% CI: 0.822-0.938). In order to evaluate the impact of the proposed method, performance comparison with several baseline methods was also performed. While limited by case volume and a few more factors, the results of this study can contribute to the potential development of an artificial intelligence system to diagnose other cancers without extensive region-level annotations.
| S-EPMC8203612 | biostudies-literature
Project description:PurposeThe purpose of this study was to design an automated algorithm that can detect fluorescence leakage accurately and quickly without the use of a large amount of labeled data.MethodsA weakly supervised learning-based method was proposed to detect fluorescein leakage without the need for manual annotation of leakage areas. To enhance the representation of the network, a residual attention module (RAM) was designed as the core component of the proposed generator. Moreover, class activation maps (CAMs) were used to define a novel anomaly mask loss to facilitate more accurate learning of leakage areas. In addition, sensitivity, specificity, accuracy, area under the curve (AUC), and dice coefficient (DC) were used to evaluate the performance of the methods.ResultsThe proposed method reached a sensitivity of 0.73 ± 0.04, a specificity of 0.97 ± 0.03, an accuracy of 0.95 ± 0.05, an AUC of 0.86 ± 0.04, and a DC of 0.87 ± 0.01 on the HRA data set; a sensitivity of 0.91 ± 0.02, a specificity of 0.97 ± 0.02, an accuracy of 0.96 ± 0.03, an AUC of 0.94 ± 0.02, and a DC of 0.85 ± 0.03 on Zhao's publicly available data set; and a sensitivity of 0.71 ± 0.04, a specificity of 0.99 ± 0.06, an accuracy of 0.87 ± 0.06, an AUC of 0.85 ± 0.02, and a DC of 0.78 ± 0.04 on Rabbani's publicly available data set.ConclusionsThe experimental results showed that the proposed method achieves better performance on fluorescence leakage detection and can detect one image within 1 second and thus has great potential value for clinical diagnosis and treatment of retina-related diseases, such as diabetic retinopathy and malarial retinopathy.Translational relevanceThe proposed weakly supervised learning-based method that automates the detection of fluorescence leakage can facilitate the assessment of retinal-related diseases.
| S-EPMC8934548 | biostudies-literature
Project description:The development of decision support systems for pathology and their deployment in clinical practice have been hindered by the need for large manually annotated datasets. To overcome this problem, we present a multiple instance learning-based deep learning system that uses only the reported diagnoses as labels for training, thereby avoiding expensive and time-consuming pixel-wise manual annotations. We evaluated this framework at scale on a dataset of 44,732 whole slide images from 15,187 patients without any form of data curation. Tests on prostate cancer, basal cell carcinoma and breast cancer metastases to axillary lymph nodes resulted in areas under the curve above 0.98 for all cancer types. Its clinical application would allow pathologists to exclude 65-75% of slides while retaining 100% sensitivity. Our results show that this system has the ability to train accurate classification models at unprecedented scale, laying the foundation for the deployment of computational decision support systems in clinical practice.
| S-EPMC7418463 | biostudies-literature
Project description:Deep eutectic solvents (DES) are recently synthesized to cover limitations of conventional solvents. These green solvents have wide ranges of potential usages in real-life applications. Precise measuring or accurate estimating thermophysical properties of DESs is a prerequisite for their successful applications. Density is likely the most crucial affecting characteristic on the solvation ability of DESs. This study utilizes seven machine learning techniques to estimate the density of 149 deep eutectic solvents. The density is anticipated as a function of temperature, critical pressure and temperature, and acentric factor. The LSSVR (least-squares support vector regression) presents the highest accuracy among 1530 constructed intelligent estimators. The LSSVR predicts 1239 densities with the mean absolute percentage error (MAPE) of 0.26% and R2 = 0.99798. Comparing the LSSVR and four empirical correlations revealed that the earlier possesses the highest accuracy level. The prediction accuracy of the LSSVR (i.e., MAPE = 0. 26%) is 74.5% better than the best-obtained results by the empirical correlations (i.e., MAPE = 1.02%).
| S-EPMC8943155 | biostudies-literature
Project description:As a hot topic in brain disease prognosis, predicting clinical measures of subjects based on brain magnetic resonance imaging (MRI) data helps to assess the stage of pathology and predict future development of the disease. Due to incomplete clinical labels/scores, previous learning-based studies often simply discard subjects without ground-truth scores. This would result in limited training data for learning reliable and robust models. Also, existing methods focus only on using hand-crafted features (e.g., image intensity or tissue volume) of MRI data, and these features may not be well coordinated with prediction models. In this paper, we propose a weakly supervised densely connected neural network (wiseDNN) for brain disease prognosis using baseline MRI data and incomplete clinical scores. Specifically, we first extract multiscale image patches (located by anatomical landmarks) from MRI to capture local-to-global structural information of images, and then develop a weakly supervised densely connected network for task-oriented extraction of imaging features and joint prediction of multiple clinical measures. A weighted loss function is further employed to make full use of all available subjects (even those without ground-truth scores at certain time-points) for network training. The experimental results on 1469 subjects from both ADNI-1 and ADNI-2 datasets demonstrate that our proposed method can efficiently predict future clinical measures of subjects.
| S-EPMC8034591 | biostudies-literature