Image Diagnostic Support AI to Assist Physicians: EIRL

LPIXEL Inc.

Image Diagnostic Support AI to Assist Physicians: EIRL
Image Diagnostic Support AI to Assist Physicians: EIRL

Supporting doctors by detecting suspected lesions in medical readings, such as X-rays

In chest X-rays and CT scans, physicians must correctly identify lesions and findings from a vast number of images without error. Regardless of their level of experience or physical condition, they are required to maintain consistent quality in their readings.

EIRL integrates seamlessly into existing systems to detect and highlight areas of concern in images, helping prevent oversight by the physicians. This technology is particularly promising in regions with a low physician-to-population ratio or areas where there are significant disparities in the quality of healthcare between urban and rural regions.

System information

Easily incorporate AI by connecting it to existing image management systems

EIRL complies with the international medical imaging standard DICOM and enables seamless transmission and reception of image data from PACS or modalities already implemented in medical institutions. The analysis results can then be output to workstations or other devices used by physicians.

Additionally, a cloud-based service is available, allowing for easy use as long as an online connection is established.
EIRL product lineup

A pioneer of medical AI in Japan

Since its launch in 2019, EIRL has been implemented in over 800 medical facilities across Japan as of June 2024. Being a forerunner of diagnostic support AI in Japan, EIRL is actively utilized in medical practices nationwide.

Even after the product's release, continuous updates and performance improvements are made based on feedback from physicians. The product lineup has also expanded beyond chest X-rays to include solutions for CT scans and colonoscopy applications.
EIRL Chest Screening

Improve detection sensitivity with integration

When using Chest XR, a part of EIRL Chest Screening, for interpretation compared to physician-only readings, sensitivity increased by 11.1% for radiology specialists and by 15.5% for non-specialist physicians with less than five years of experience. The results are based on a reading trial conducted during the regulatory approval process for the EIRL Chest XR medical image analysis software in Japan, involving chest X-ray images from 88 cases with abnormal shadows and 266 normal cases, assessed by 10 physicians. The breakdown of participating physicians included three radiology specialists and seven non-specialists (three with over 10 years of experience and four with less than five years).