Objectives Artificial intelligence (AI) has been an important addition to medicine. We aimed to explore the use of deep learning (DL) to distinguish benign from malignant lesions with breast ultrasound (BUS). Methods The DL model was trained with BUS nodule data using a standard protocol (1271 malignant nodules, 1053 benign nodules, and 2144 images of the contralateral normal breast). The model was tested with 692 images of 256 breast nodules. We used the accuracy, precision, recall, harmonic mean of recall and precision, and mean average precision as the indices to assess the DL model. We used 100 BUS images to evaluate differences in diagnostic accuracy among the AI system, experts (>25 years of experience), and physicians with varying levels of experience. A receiver operating characteristic curve was generated to evaluate the accuracy for distinguishing between benign and malignant breast nodules. Results The DL model showed 73.3% sensitivity and 94.9% specificity for the diagnosis of benign versus malignant breast nodules (area under the curve, 0.943). No significant difference in diagnostic ability was found between the AI system and the expert group (P = .951), although the physicians with lower levels of experience showed significant differences from the AI and expert groups (P = .01 and .03, respectively). Conclusions Deep learning could distinguish between benign and malignant breast nodules with BUS. On BUS images, DL achieved diagnostic accuracy equivalent to that of expert physicians.

No abstract

No abstract

No abstract

3D human body models have been widely used in all fields. For acquiring precise human body models, the research about their recognition and reconstruction has become a major topic. Here we present a new method for modeling 3D human body based on single Kinect from the viewpoint of cost and operability. In this method, human body 3D information is recognized and acquired by only one Kinect while 3D human body models are reconstructed by using the tools of Processing and Point Cloud Library. To achieve the reconstruction objective, the iterative closest point algorithm is adopted for registering the captured upper human body 3D point cloud data with the standard reference human body data. The experiment results demonstrate that this method is feasible.

No abstract

Somatosensory interaction is a kind of man-machine interfacing technique for information acquisition through human postures, which are widely used in digital art nowadays. To create opportunities for the visually impaired to enjoy digital art, two sound art systems named Dynamic Sound and Concrete Sound, which are based on somatosensory technology, were designed in this study for the visually impaired to appreciate with pleasure. The former system emphasizes resonances between humans and sound, allowing the visually impaired user to trigger different sounds by hand gestures which promote the user's physio-pleasure and ideo-pleasure. The latter system, also being controlled by hand gestures, combines sounds with three-state physical phenomena as feedbacks which are explained orally by an accompanying person to the visually impaired user, creating an inter-person communication that promotes the user's socio-pleasure. By public exhibitions, users' feedbacks were acquired via interviews, and evaluated to prove the effectiveness of the proposed systems with the following findings: 1) interactions by hand gestures offers the visually impaired with opportunities to enjoy digital art; 2) sound art provided by the systems allows the visually impaired to gain pleasure via man-machine interactions; 3) the systems innovatively integrate dynamic visual performances with auditory feedbacks in the interaction process; and 4) through the development of gesture operations, more performance techniques can be devised for sound art, allowing gesture motions to replace control interfaces in future designs.

No abstract