The Chinese University of Hong Kong recently announced that its research team used artificial intelligence image recognition technology to interpret medical images of lung cancer and breast cancer with accurate rates of 91% and 99%, respectively, and the identification process took only 30 seconds to 10 minutes. According to the researchers, this technology can greatly improve the efficiency of clinical diagnosis and reduce the rate of misdiagnosis. The Chinese University of Hong Kong recently announced that its research team used artificial intelligence image recognition technology to interpret medical images of lung cancer and breast cancer with accurate rates of 91% and 99%, respectively, and the identification process took only 30 seconds to 10 minutes. According to the researchers, this technology can greatly improve the efficiency of clinical diagnosis and reduce the rate of misdiagnosis. The Chinese University of Hong Kong recently held a press conference. Wang Ping, a research team leader and a professor of computer science and engineering at CUHK, introduced these two studies using artificial intelligence systems: automatic screening of early lung cancer and rapid detection of breast cancer metastasis. According to reports, early lung cancer often occurs in the form of small pulmonary nodules. Doctors mainly use chest computed tomography (CT) images to check for the presence of pulmonary nodules, and each examination has as many as hundreds of tomographic images. The doctor only judges with the naked eye, which is time consuming and laborious. The team used deep learning techniques to interpret CT scan images, and the deep neural network automatically detected the location of possible small pulmonary nodules. It took 30 seconds and the accuracy was as high as 91%. Wang Pingan said that deep learning technology can enhance technical sensitivity and eliminate suspected and false positives. The team will also jointly develop related products in conjunction with several hospitals in Beijing to optimize technology and identify lung nodules as early as possible, providing a reliable basis for early diagnosis and treatment of lung cancer. For the detection of breast cancer, doctors usually scan the position of the hard block by mammography or magnetic resonance imaging. When detecting lymph node metastasis, the doctor will take a small piece of living tissue as a sample, examine whether the lymph nodes metastasize under the microscope, and whether the tumor is benign or malignant. The resolution of a digital bio-sliced ​​full-slice image is very high, and the file size can reach 1GB, which is equivalent to the storage capacity of a 90-minute HD movie, which makes the detection process complicated and time-consuming. In this regard, Wang Pingan said that the team developed a new type of deep convolutional neural network to process slice images of breast cancer in stages. First, use a modified version of the full convolutional network (a fast-predictive model that performs coarser but more sensitive images) to reconstruct more accurate and accurate predictions, then locate and select images containing lymph node metastases. Compared with the results of manual testing by senior pathologists, the accuracy of this automated test is 2% higher, reaching 98.75%, and it takes only 5 to 10 minutes. It is reported that the team started relevant experiments five years ago. Wang Pingan said that it is expected that this automated monitoring technology will be widely used in the medical profession in Hong Kong in the next 1-2 years. Melamine Board,Whiteboard With Stand,Mini Whiteboard,Desktop Whiteboards Dongguan Aoxing Audio Visual Equipment CO.,Ltd , https://www.aoxing-projectorscreen.com
February 17, 2023