Dual-View Single-Shot Multibox Detector at Urban Intersections: Settings and Performance Evaluation

The explosion of artificial intelligence methods has paved the way for more sophisticated smart mobility solutions. In this work, we present a multi-camera video content analysis (VCA) system that exploits a single-shot multibox detector (SSD) network to detect vehicles, riders, and pedestrians and triggers alerts to drivers of public transportation vehicles approaching the surveilled area. The evaluation of the VCA system will address both detection and alert generation performance by combining visual and quantitative approaches. Starting from a SSD model trained for a single camera, we added a second one, under a different field of view (FOV) to improve the accuracy and reliability of the system. Due to real-time constraints, the complexity of the VCA system must be limited, thus calling for a simple multi-view fusion method. According to the experimental test-bed, the use of two cameras achieves a better balance between precision (68%) and recall (84%) with respect to the use of a single camera (i.e., 62% precision and 86% recall). In addition, a system evaluation in temporal terms is provided, showing that missed alerts (false negatives) and wrong alerts (false positives) are typically transitory events. Therefore, adding spatial and temporal redundancy increases the overall reliability of the VCA system.

Interpretatively automated identification of circulating tumor cells from human peripheral blood with high performance

The detection and analysis of circulating tumor cells (CTCs) would be of aid in a precise cancer diagnosis and an efficient prognosis assessment. However, traditional methods that rely heavily on the isolation of CTCs based on their physical or biological features suffer from intensive labor, thus being unsuitable for rapid detection. Furthermore, currently available intelligent methods are short of interpretability, which creates a lot of uncertainty during diagnosis. Therefore, we propose here an automated method that takes advantage of bright-field microscopic images with high resolution, so as to take an insight into cell patterns. Specifically, the precise identification of CTCs was achieved by using an optimized single-shot multi-box detector (SSD)-based neural network with integrated attention mechanism and feature fusion modules. Compared to the conventional SSD system, our method exhibited a superior detection performance with the recall rate of 92.2%, and the maximum average precision (AP) value of 97.9%. To note, the optimal SSD-based neural network was combined with advanced visualization technology, i.e., the gradient-weighted class activation mapping (Grad-CAM) for model interpretation, and the t-distributed stochastic neighbor embedding (T-SNE) for data visualization. Our work demonstrates for the first time the outstanding performance of SSD-based neural network for CTCs identification in human peripheral blood environment, showing great potential for the early detection and continuous monitoring of cancer progression.

Sea Cucumber Detection Algorithm Based on Deep Learning

The traditional single-shot multiBox detector (SSD) for the recognition process in sea cucumbers has problems, such as an insufficient expression of features, heavy computation, and difficulty in application to embedded platforms. To solve these problems, we proposed an improved algorithm for sea cucumber detection based on the traditional SSD algorithm. MobileNetv1 is selected as the backbone of the SSD algorithm. We increase the feature receptive field by receptive field block (RFB) to increase feature details and location information of small targets. Combined with the attention mechanism, features at different depths are strengthened and irrelevant features are suppressed. The experimental results show that the improved algorithm has better performance than the traditional SSD algorithm. The average precision of the improved algorithm is increased by 5.1%. The improved algorithm is also more robust. Compared with YOLOv4 and the Faster R-CNN algorithm, the performance of this algorithm on the P-R curve is better, indicating that the performance of this algorithm is better. Thus, the improved algorithm can stably detect sea cucumbers in real time and provide reliable feedback information.

Hyperspectral Imaging Combined with Artificial Intelligence in the Early Detection of Esophageal Cancer

Simple Summary

Detection of early esophageal cancer is important to improve patient survival, however, early diagnosis of the cancer cells is difficult, even for experienced endoscopists. This article provides a new method by using hyperspectral imaging and a deep learning diagnosis model to classify and diagnose esophageal cancer using a single-shot multibox detector. The accuracy of the results when using an RGB image in WLI was 83% and while using the spectrum data the accuracy was increased to 88%. There was an increase of 5% in WLI. The accuracy of the results when using an RGB image in NBI was 86% and while using the spectrum data the accuracy was increased to 91%. There was an increase of 5% in NBI. This study proves that the accuracy of prediction when using the spectrum data has been significantly improved and the diagnosis of narrow-band endoscopy data is more sensitive than that of white-light endoscopy.

Abstract

This study uses hyperspectral imaging (HSI) and a deep learning diagnosis model that can identify the stage of esophageal cancer and mark the locations. This model simulates the spectrum data from the image using an algorithm developed in this study which is combined with deep learning for the classification and diagnosis of esophageal cancer using a single-shot multibox detector (SSD)-based identification system. Some 155 white-light endoscopic images and 153 narrow-band endoscopic images of esophageal cancer were used to evaluate the prediction model. The algorithm took 19 s to predict the results of 308 test images and the accuracy of the test results of the WLI and NBI esophageal cancer was 88 and 91%, respectively, when using the spectral data. Compared with RGB images, the accuracy of the WLI was 83% and the NBI was 86%. In this study, the accuracy of the WLI and NBI was increased by 5%, confirming that the prediction accuracy of the HSI detection method is significantly improved.

Endoscopic Images by a Single-Shot Multibox Detector for the Identification of Early Cancerous Lesions in the Esophagus: A Pilot Study

Diagnosis of early esophageal neoplasia, including dysplasia and superficial cancer, is a great challenge for endoscopists. Recently, the application of artificial intelligence (AI) using deep learning in the endoscopic field has made significant advancements in diagnosing gastrointestinal cancers. In the present study, we constructed a single-shot multibox detector using a convolutional neural network for diagnosing different histological grades of esophageal neoplasms and evaluated the diagnostic accuracy of this computer-aided system. A total of 936 endoscopic images were used as training images, and these images included 498 white-light imaging (WLI) and 438 narrow-band imaging (NBI) images. The esophageal neoplasms were divided into three classifications: squamous low-grade dysplasia, squamous high-grade dysplasia, and squamous cell carcinoma, based on pathological diagnosis. This AI system analyzed 264 test images in 10 s, and the sensitivity, specificity, and diagnostic accuracy of this system in detecting esophageal neoplasms were 96.2%, 70.4%, and 90.9%, respectively. The accuracy of this AI system in differentiating the histological grade of esophageal neoplasms was 92%. Our system showed better accuracy in diagnosing NBI (95%) than WLI (89%) images. Our results showed the great potential of AI systems in identifying esophageal neoplasms as well as differentiating histological grades.