"Evolution & Transformation"
It was about two years ago when I encountered Infervision. This group of very young scientists impressed me. Their development in China’s market is very promising. During our cooperation, young doctors and scientists from our hospital fully integrated into one with Infervision. We hope that through our joint effort, we will achieve excellence in the next few years.
Infervision’s products are already online and have achieved a good performance. Regarding the outlook of deep learning, in general, that is only applicable to a product that has not been realized. In fact, the image analysis product based on deep learning is already landed, and artificial intelligence in now in progress at Shanghai Changzheng Hospital.
Through cooperation with Infervision, I realized that the significance of artificial intelligence includes the following: First, to improve the efficiency of doctors, departments and hospitals, from the current product perspective, the computer analyzes pictures much faster than people, and therefore helps our patients; Second, the more cases computer sees, the larger the sample size is and the higher the accuracy rate is, which could help reduce misdiagnosis and missed lesions.
"A.I. Supporting Healthcare"
Wuhan Tongji Hospital has several hospital areas, and we have established a set of information management platforms with 300 heterogeneous sub-systems. We indeed need artificial intelligence to provide us support. We now have so many hospital areas, and the total number of beds will gradually expand to 10,000. Therefore, our doctor resources are not enough, including radiologists and pathologists. We really hope that we have an automated A.I. system to assist us in diagnosis.
Infervision’s A.I.—Augmented CT Screening Solution (AI—CT) is applied in early lung cancer screening. With its high paralleling computing power, AI—CT can precisely grasp the core characteristics of lung cancer and efficiently detect suspicious lung cancer lesions in CT scans. AI—CT is a technology that facilitates early detection and early treatment of lung cancer. Through comparison with physicians, AI—CT proves to increase the efficiency of lung cancer screening. AI—CT is especially sensitive to hard-to-detect nodules such as semi-solid and ground glass nodules and hence enhances radiologists’ diagnosis accuracy.
Infervision’s A.I.--Augmented X-ray Screening Solution (AI—DR) can detect more than 20 different kinds of cardiothoracic lesions. It helps screening diseases during both regular physical examinations as well as in-patient and out-patient radiology studies. AI—DR is especially sensitive to lung nodules on X-ray scans. During several months of use at a collaborating hospital, AI—DR helped screen out a few lung cancer patients who were initially misdiagnosed by radiologists. X-ray is often thought as an outdated medical imaging method. However, AI—DR is a technology that revives the significance of X-ray in medical imaging diagnosis.
Deep Learning Research Platform AI—Scholar surveys deep learning models to provide robust and powerful GPU computation for medical data modeling. The platform can process over 100 high-resolution medical DICOM images in one second. AI—Scholar includes the most advanced medical imaging related deep learning models. Doctors can combine and tailor models to investigate their own research questions. AI—Scholar features user-friendly interface. Through several brief clicks, a doctor without exposure to any programming experience can train a deep learning model. From Infervision’s perspective, deep learning should become part of the methodological toolkit in medical research.