Clinical utility of artificial intelligence assistance in histopathologic review of lymph node metastasis for gastric adenocarcinoma.

BACKGROUND: Advances in whole-slide image capture and computer image analyses using deep learning technologies have enabled the development of computer-assisted diagnostics in pathology. Herein, we built a deep learning algorithm to detect lymph node (LN) metastasis on whole-slide images of LNs retrieved from patients with gastric adenocarcinoma and evaluated its performance in clinical settings. METHODS: We randomly selected 18 patients with gastric adenocarcinoma who underwent surgery with curative intent and were positive for LN metastasis at Chiba University Hospital. A ResNet-152-based assistance system was established to detect LN metastases and to outline regions that are highly probable for metastasis in LN images. Reference standards comprising 70 LN images from two different institutions were reviewed by six pathologists with or without algorithm assistance, and their diagnostic performances were compared between the two settings. RESULTS: No statistically significant differences were observed between these two settings regarding sensitivity, review time, or confidence levels in classifying macrometastases, isolated tumor cells, and metastasis-negative. Meanwhile, the sensitivity for detecting micrometastases significantly improved with algorithm assistance, although the review time was significantly longer than that without assistance. Analysis of the algorithm's sensitivity in detecting metastasis in the reference standard indicated an area under the curve of 0.869, whereas that for the detection of micrometastases was 0.785. CONCLUSIONS: A wide variety of histological types in gastric adenocarcinoma could account for these relatively low performances; however, this level of algorithm performance could suffice to help pathologists improve diagnostic accuracy.

Association between intraoperative pulmonary artery pressure and cardiovascular complications after off-pump coronary artery bypass surgery: a single-center observational study.

BACKGROUND: The impact of intraoperative pulmonary hemodynamics on prognosis after off-pump coronary artery bypass (OPCAB) surgery remains unknown. In this study, we examined the association between intraoperative vital signs and the development of major adverse cardiovascular events (MACE) during hospitalization or within 30 days postoperatively. METHODS: This retrospective study analyzed data from a university hospital. The study cohort comprised consecutive patients who underwent isolated OPCAB surgery between November 2013 and July 2021. We calculated the mean and coefficient of variation of vital signs obtained from the intra-arterial catheter, pulmonary artery catheter, and pulse oximeter. The optimal cut-off was defined as the receiver operating characteristic curve (ROC) with the largest Youden index (Youden index = sensitivity + specificity - 1). Multivariate logistic regression analysis ROC curves were used to adjust all baseline characteristics that yielded P values of < 0.05. RESULTS: In total, 508 patients who underwent OPCAB surgery were analyzed. The mean patient age was 70.0 ± 9.7 years, and 399 (79%) were male. There were no patients with confirmed or suspected preoperative pulmonary hypertension. Postoperative MACE occurred in 32 patients (heart failure in 16, ischemic stroke in 16). The mean pulmonary artery pressure (PAP) was significantly higher in patients with than without MACE (19.3 ± 3.0 vs. 16.7 ± 3.4 mmHg, respectively; absolute difference, 2.6 mmHg; 95% confidence interval, 1.5 to 3.8). The area under the ROC curve of PAP for the prediction of MACE was 0.726 (95% confidence interval, 0.645 to 0.808). The optimal mean PAP cut-off was 18.8 mmHg, with a specificity of 75.8% and sensitivity of 62.5% for predicting MACE. After multivariate adjustments, high PAP remained an independent risk factor for MACE. CONCLUSIONS: Our findings provide the first evidence that intraoperative borderline pulmonary hypertension may affect the prognosis of patients undergoing OPCAB surgery. Future large-scale prospective studies are needed to verify the present findings.

Reliability of non-contact tongue diagnosis for Sjögren's syndrome using machine learning method.

Sjögren's syndrome (SS) is an autoimmune disease characterized by dry mouth. The cause of SS is unknown, and its diverse symptoms make diagnosis difficult. The Saxon test, an intraoral examination, is used as the primary diagnostic method for SS, however, the risk of salivary infection is problematic. Therefore, we investigate the possibility of diagnosing SS by non-contact and imaging observation of the tongue surface. In this study, we obtained tongue photographs of 60 patients at the Tsurumi University School of Dentistry outpatient clinic to clarify the relationship between the features of the tongue and SS. We divided the tongue into four regions, and the color of each region was transformed into CIE1976L*a*b* space and statistically analyzed. To clarify experimentally the possibility of SS diagnosis using tongue color, we employed three machine-learning models: logistic regression, support vector machine, and random forest. In addition, we constructed diagnostic prediction models based on the Bagging and Stacking methods combined with three machine-learning models for comparative evaluation. This analysis used dimensionality compression by principal component analysis to eliminate redundancy in tongue color information. We found a significant difference between the a* value of the rear part of the tongue and the b* value of the middle part of the tongue in SS and non-SS patients. In addition to the principal component scores of tongue color, the support vector machine was trained using age, and achieved high accuracy (71.3%) and specificity (78.1%). The results indicate that the prediction of SS diagnosis by tongue color reaches a level comparable to machine learning models trained using the Saxon test. This is the first study using machine learning to predict SS diagnosis by non-contact tongue observation. Our proposed method can potentially support early SS detection simply and conveniently, eliminating the risk of infection at diagnosis, and it should be validated and optimized in clinical practice.

The Use of Deep Learning-Based Computer Diagnostic Algorithm for Detection of Lymph Node Metastases of Gastric Adenocarcinoma.

Objectives. The diversifying modalities of treatment for gastric cancer raise urgent demands for the rapid and precise diagnosis of metastases in regional lymph nodes, thereby significantly impact the workload of pathologists. Meanwhile, the recent advent of whole-slide scanners and deep-learning techniques have enabled the computer-assisted analysis of histopathological images, which could help to alleviate this impact. Thus, we developed a deep learning-based diagnostic algorithm to detect lymph node metastases of gastric adenocarcinoma and evaluated its performance. Methods. We randomly selected 20 patients with gastric adenocarcinoma who underwent surgery as definitive treatment and were found to be node metastasis-positive. HEMATOXYLIN-eosin (HE) stained glass slides, including a total of 51 metastasis-positive nodes, were retrieved from the specimens of these cases. Other slides with 776 metastasis-negative nodes were also retrieved from other twenty cases with the same disease that were diagnosed as metastasis-negative by the final pathological examinations. All glass slides were digitized using a whole-slide scanner. A deep-learning algorithm to detect metastases was developed using the data in which metastasis-positive parts of the images were annotated by a well-trained pathologist, and its performance in detecting metastases was evaluated. Results. Cross-validation analysis indicated an area of 0.9994 under the receiver operating characteristic curve. Free-response receiver operating characteristic curve (FROC) analysis indicated a sensitivity of 1.00 with three false positives. Further evaluation using an independent dataset also showed similar level of accuracies. Conclusion. This deep learning-based diagnosis-aid system is a promising tool that can assist pathologists involved in gastric cancer care and reduce their workload.

U-Net-Based Segmentation of Microscopic Images of Colorants and Simplification of Labeling in the Learning Process.

Colored product textures correspond to particle size distributions. The microscopic images of colorants must be divided into regions to determine the particle size distribution. The conventional method used for this process involves manually dividing images into areas, which may be inefficient. In this paper, we have overcome this issue by developing two different modified architectures of U-Net convolution neural networks to automatically determine the particle sizes. To develop these modified architectures, a significant amount of ground truth data must be prepared to train the U-Net, which is difficult for big data as the labeling is performed manually. Therefore, we also aim to reduce this process by using incomplete labeling data. The first objective of this study is to determine the accuracy of our modified U-Net architectures for this type of image. The second objective is to reduce the difficulty of preparing the ground truth data by testing the accuracy of training on incomplete labeling data. The results indicate that efficient segmentation can be realized using our modified U-Net architectures, and the generation of ground truth data can be simplified. This paper presents a preliminary study to improve the efficiency of determining particle size distributions with incomplete labeling data.

Superiority of Supervised Machine Learning on Reading Chest X-Rays in Intensive Care Units.

Purpose: Portable chest radiographs are diagnostically indispensable in intensive care units (ICU). This study aimed to determine if the proposed machine learning technique increased in accuracy as the number of radiograph readings increased and if it was accurate in a clinical setting. Methods: Two independent data sets of portable chest radiographs (n = 380, a single Japanese hospital; n = 1,720, The National Institution of Health [NIH] ChestX-ray8 dataset) were analyzed. Each data set was divided training data and study data. Images were classified as atelectasis, pleural effusion, pneumonia, or no emergency. DenseNet-121, as a pre-trained deep convolutional neural network was used and ensemble learning was performed on the best-performing algorithms. Diagnostic accuracy and processing time were compared to those of ICU physicians. Results: In the single Japanese hospital data, the area under the curve (AUC) of diagnostic accuracy was 0.768. The area under the curve (AUC) of diagnostic accuracy significantly improved as the number of radiograph readings increased from 25 to 100% in the NIH data set. The AUC was higher than 0.9 for all categories toward the end of training with a large sample size. The time to complete 53 radiographs by machine learning was 70 times faster than the time taken by ICU physicians (9.66 s vs. 12 min). The diagnostic accuracy was higher by machine learning than by ICU physicians in most categories (atelectasis, AUC 0.744 vs. 0.555, P < 0.05; pleural effusion, 0.856 vs. 0.706, P < 0.01; pneumonia, 0.720 vs. 0.744, P = 0.88; no emergency, 0.751 vs. 0.698, P = 0.47). Conclusions: We developed an automatic detection system for portable chest radiographs in ICU setting; its performance was superior and quite faster than ICU physicians.

Real-Time Expanded Field-of-View for Minimally Invasive Surgery Using Multi-Camera Visual Simultaneous Localization and Mapping.

Minimally invasive surgery is widely used because of its tremendous benefits to the patient. However, there are some challenges that surgeons face in this type of surgery, the most important of which is the narrow field of view. Therefore, we propose an approach to expand the field of view for minimally invasive surgery to enhance surgeons' experience. It combines multiple views in real-time to produce a dynamic expanded view. The proposed approach extends the monocular Oriented features from an accelerated segment test and Rotated Binary robust independent elementary features-Simultaneous Localization And Mapping (ORB-SLAM) to work with a multi-camera setup. The ORB-SLAM's three parallel threads, namely tracking, mapping and loop closing, are performed for each camera and new threads are added to calculate the relative cameras' pose and to construct the expanded view. A new algorithm for estimating the optimal inter-camera correspondence matrix from a set of corresponding 3D map points is presented. This optimal transformation is then used to produce the final view. The proposed approach was evaluated using both human models and in vivo data. The evaluation results of the proposed correspondence matrix estimation algorithm prove its ability to reduce the error and to produce an accurate transformation. The results also show that when other approaches fail, the proposed approach can produce an expanded view. In this work, a real-time dynamic field-of-view expansion approach that can work in all situations regardless of images' overlap is proposed. It outperforms the previous approaches and can also work at 21 fps.