Oxipit's AI ChestEye imaging suite receives CE certification
02/02/2019

Leading AI-based medical imaging solutions provider Oxipit received CE certification for ChestEye radiology imaging suite.

Online PR News – 02-February-2019 – Vilnius, Lithuania – Leading AI-based medical imaging solutions provider Oxipit received CE certification for ChestEye radiology imaging suite. Oxipit ChestEye provides analysis and preliminary reports for 75 most common radiological findings - the largest scope of diagnosis currently available on the market - with an average area under curve metric of 93%. CE mark ensures that the software complies with medical device regulations and paves the way for commercial deployments in 32 European countries.

"The burden on radiologists has been constantly increasing. This has led to decline in quality of service for patients and bottlenecks to access good quality radiological reporting. The CE mark brings us a step closer in helping radiologists to harness the capabilities of Deep Learning in order to multiply their productivity and provide excellent service around the clock" - notes Oxipit CEO Gediminas Peksys.

ChestEye is the first AI-based full workflow medical imaging suite to be certified by a CE mark. The solution has received numerous praises from industry professionals and won multiple prizes including the 1st place at the EIT Health InnoStars Awards in 2018.

ChestEye imaging suite encompasses a fully automatic computer aided diagnosis (CAD) platform which supports 75 radiological findings. The software localizes these features on a radiograph as a heatmap. It also generates a standardized preliminary text report that incorporates all the radiologically relevant information present in a chest X-ray image.

"ChestEye currently covers over 90% of radiological cases presented to radiologists on a daily basis. It aids the specialists to detect hard-to-catch edge cases and offers a second opinion. In addition, the automatic generation of standardized medical reports significantly reduces the workload of a radiologist, allowing more time for radiogram analysis and less for case description. An internal trial showed 30% time saved per patient and reduced error rate by up to 50%" - notes Gediminas Peksys.

The Search module of ChestEye incorporates a search engine for finding similar-looking chest X-rays in a given database empowering the user to quickly find retrospective cases with similar radiological appearance. The similarity is identified by a neural network, which judges on the pathology present as well as the location of the pathology, its severity and other features.

"Most of our research data comes from tertiary level hospitals where complicated cases are analyzed and some individual images present multiple issues. The dataset represents real-life images that a medical institution encounters on a day-to-day basis, allowing our engine to achieve high precision in everyday applications" - notes Gediminas Peksys.

The suite also includes a patient prioritization solution. After receiving input of ChestEye scan results the platform prioritizes potentially unhealthy patients inviting urgent specialist attention. By automatically arranging scans by urgency it reduces time-to-treatment for time sensitive conditions such as pericardial effusion, pneumothorax, catheter or intubation malposition.

The ChestEye suite is available for deployment on premises as well as a cloud-based software. It seamlessly integrates into the workflow of any radiology department.

"Radiologists are already bombarded with too many screens and software applications. From the ground up we have built ChestEye to easily integrate into PACS (automatic image input) and RIS (automatic report output) infrastructure of a radiology department. We strongly believe that in addition to diagnostic precision, computer-aided diagnosis solutions should increase radiologist productivity and make their work less tiresome. ChestEye supports DICOM protocol for image exchange and our the team is working to incorporate HL7 and FHIR protocols in the upcoming versions of the product" - notes Gediminas Peksys.