- Background
In the current endoscopy industry, electronic endoscopes (such as ureteroscopes and bronchoscopes) generally lag behind rigid endoscopes (such as laparoscopes) in terms of resolution indicators. The resolution of mainstream electronic endoscope sensors is mostly concentrated at the 400×400px (160,000 pixels) level, and the details and information presented are relatively low, making it difficult to meet the needs of refined diagnosis and treatment scenarios such as early gastrointestinal cancer screening (which requires identification of lesions <5mm).
Super-resolution reconstruction (SR) technology can break through the physical limitations of sensors and become an important research direction for improving the imaging quality of electron mirrors.
2. Principle
The original 400x400px low-resolution image is processed by a deep learning-based SR algorithm, achieving a 2x upsampling, resulting in an 800x800px high-resolution image output. Compared to traditional imaging, the super-resolution algorithm’s key feature is its built-in super-resolution module, which performs super-resolution processing on the input image and outputs a high-resolution image. This algorithm improves quantifiable imaging data in multiple dimensions, increasing resolution by a factor of four from 160W to 640W, improving line-of-sight metrics by 50%, and boosting signal-to-noise ratio by 22%.
3. Effect Display
Hikimaging applies super-resolution technology to the 2302 series CCU, supporting the access of a full range of electronic scopes such as ureteroscopes, bronchoscopes, and hysteroscopes, effectively improving the current situation of unclear electronic scope imaging.
(1) Renal pelvis under ureteroscopy: fine structures are clearly visible.
(2) Kidney stones under ureteroscopy: the internal structure of the stone is clear.