Gleason grading of histological images is important in risk assessment and treatment planning for prostate cancer patients. Much research has been done in classifying small homogeneous cancer regions within histological images. However, semi-supervised methods published to date depend on pre-selected regions and cannot be easily extended to an image of heterogeneous tissue composition. In this paper, we propose a multi-scale U-Net model to classify images at the pixel-level using 224 histological image tiles from radical prostatectomies of 20 patients. Our model was evaluated by a patient-based 10-fold cross validation, and achieved a mean Jaccard index of 65.8% across 4 classes (stroma, Gleason 3, Gleason 4 and benign glands), and 75.5% for 3 classes (stroma, benign glands, prostate cancer), outperforming other methods.

Practitioner guidelines simultaneously provide broad overviews and in-depth details of disease. Written for experts, they are difficult for patients to understand, yet patients often use these guidelines as a source of information to help them to learn about their health. Using practitioner guidelines along with patient information needs and preferences, we created a method to design an information model for providing patients access to their personal health information, linked to individualized, relevant supporting information from guidelines within a patient portal. This model consists of twelve classes of concepts. We manually reviewed and annotated medical records to demonstrate the validity of our model. Each class of the model was found within at least one patient's record, and seven classes of concepts appeared in over half of the patients' records annotated. These annotations show that the model produced by the method can be used to determine what guideline information is relevant to an individual patient, based on concepts in their health information.

Patient portals have the potential to provide content that is specifically tailored to a patient's information needs based on diagnoses and other factors. In this work, we conducted a survey of 41 lung cancer patients at an outpatient lung cancer clinic at the medical center of the University of California Los Angeles, to gain insight into these perceived information needs and opinions on the design of a portal to fulfill them. We found that patients requested access to information related to diagnosis and imaging, with more than half of the patients reporting that they did not anticipate an increase in anxiety due to access to medical record information via a portal. We also found that patient educational background did not lead to a significant difference in desires for explanations of reports and definitions of terms.

Predicting infarct volume from magnetic resonance perfusion-weighted imaging can provide helpful information to clinicians in deciding how aggressively to treat acute stroke patients. Models have been developed to predict tissue fate, yet these models are mostly built using hand-crafted features (e.g., time-to-maximum) derived from perfusion images, which are sensitive to deconvolution methods. We demonstrate the application of deep convolution neural networks (CNNs) on predicting final stroke infarct volume using only the source perfusion images. We propose a deep CNN architecture that improves feature learning and achieves an area under the curve of 0.871±0.024 , outperforming existing tissue fate models. We further validate the proposed deep CNN with existing 2-D and 3-D deep CNNs for images/video classification, showing the importance of the proposed architecture. Our work leverages deep learning techniques in stroke tissue outcome prediction, advancing magnetic resonance imaging perfusion analysis one step closer to an operational decision support tool for stroke treatment guidance.

Prostate cancer is the most common and second most deadly form of cancer in men in the United States. The classification of prostate cancers based on Gleason grading using histological images is important in risk assessment and treatment planning for patients. Here, we demonstrate a new region-based convolutional neural network framework for multi-task prediction using an epithelial network head and a grading network head. Compared with a single-task model, our multi-task model can provide complementary contextual information, which contributes to better performance. Our model is achieved a state-of-the-art performance in epithelial cells detection and Gleason grading tasks simultaneously. Using fivefold cross-validation, our model is achieved an epithelial cells detection accuracy of 99.07% with an average area under the curve of 0.998. As for Gleason grading, our model is obtained a mean intersection over union of 79.56% and an overall pixel accuracy of 89.40%.

RATIONALE AND OBJECTIVES: Early prostate cancer detection and staging from MRI is extremely challenging for both radiologists and deep learning algorithms, but the potential to learn from large and diverse datasets remains a promising avenue to increase their performance within and across institutions. To enable this for prototype-stage algorithms, where the majority of existing research remains, we introduce a flexible federated learning framework for cross-site training, validation, and evaluation of custom deep learning prostate cancer detection algorithms. MATERIALS AND METHODS: We introduce an abstraction of prostate cancer groundtruth that represents diverse annotation and histopathology data. We maximize use of this groundtruth if and when they are available using UCNet, a custom 3D UNet that enables simultaneous supervision of pixel-wise, region-wise, and gland-wise classification. We leverage these modules to perform cross-site federated training using 1400+ heterogeneous multi-parameteric prostate MRI exams from two University hospitals. RESULTS: We observe a positive result, with significant improvements in cross-site generalization performance with negligible intra-site performance degradation for both lesion segmentation and per-lesion binary classification of clinically-significant prostate cancer. Cross-site lesion segmentation performance intersection-over-union (IoU) improved by 100%, while cross-site lesion classification performance overall accuracy improved by 9.5-14.8%, depending on the optimal checkpoint selected by each site. CONCLUSION: Federated learning can improve the generalization performance of prostate cancer detection models across institutions while protecting patient health information and institution-specific code and data. However, even more data and participating institutions are likely required to improve the absolute performance of prostate cancer classification models. To enable adoption of federated learning with limited re-engineering of federated components, we open-source our FLtools system at https://federated.ucsf.edu, including examples that can be easily adapted to other medical imaging deep learning projects.

Automated Gleason grading is an important preliminary step for quantitative histopathological feature extraction. Different from the traditional task of classifying small pre-selected homogeneous regions, semantic segmentation provides pixel-wise Gleason predictions across an entire slide. Deep learning-based segmentation models can automatically learn visual semantics from data, which alleviates the need for feature engineering. However, performance of deep learning models is limited by the scarcity of large-scale fully annotated datasets, which can be both expensive and time-consuming to create. One way to address this problem is to leverage external weakly labeled datasets to augment models trained on the limited data. In this paper, we developed an expectation maximization-based approach constrained by an approximated prior distribution in order to extract useful representations from a large number of weakly labeled images generated from low-magnification annotations. This method was utilized to improve the performance of a model trained on a limited fully annotated dataset. Our semi-supervised approach trained with 135 fully annotated and 1800 weakly annotated tiles achieved a mean Jaccard Index of 49.5% on an independent test set, which was 14% higher than the initial model trained only on the fully annotated dataset.

Rationale and objectives

The purpose of this study was to quantify the degree of imaging-histologic discordance in a cohort of patients undergoing computed tomography (CT)-guided lung biopsy for focal lung disease.

Materials and methods

A retrospective review was performed of 186 patients who underwent percutaneous lung biopsy of a parenchymal lesion at our institution between January and December 2009. Diagnostic radiology reports of CT or positron emission tomography-CTs performed before biopsy were used to classify the lesion as malignant or benign by five readers. Pathology reports of the biopsied lesions were classified by three readers. Inter-reader agreement and imaging-histologic concordance were quantified using kappa statistics. Discordant benign cases were then revisited to determine downstream effects.

Results

Inter-reader agreement on report content was substantial or almost perfect with kappas >0.783. Kappas for concordance were as follows: malignant (0.448), primary lung cancer (0.517), metastatic disease to lung (0.449), benign (0.510), and overall agreement (0.381). Of the twelve discordant benign cases that were revisited, four were found to be false negatives, resulting in a delay in diagnosis.

Conclusions

Our study of imaging-histologic discordance in percutaneous biopsy of lung lesions supports the need for imaging report standardization and improved integration and communication between the fields of radiology and pathology.

Social media presents a rich opportunity to gather health information with limited intervention through the analysis of completely unstructured and unlabeled microposts. We sought to estimate the health-related quality of life (HRQOL) of Twitter users using automated semantic processing methods. We collected tweets from 878 Twitter users recruited through online solicitation and in-person contact with patients. All participants completed the four-item Centers for Disease Control Healthy Days Questionnaire at the time of enrollment and 30 days later to measure "ground truth" HRQOL. We used a combination of document frequency analysis, sentiment analysis, topic analysis, and concept mapping to extract features from tweets, which we then used to estimate dichotomized HRQOL ("high" vs. "low") using logistic regression. Binary HRQOL status was estimated with moderate performance (AUC = 0.64). This result indicates that free-range social media data only offers a window into HRQOL, but does not afford direct access to current health status.