Fundus Photography Analysis

Fundus photography is the most frequently used imaging modality because it is non-invasive, well accepted by patients and, above all, because it allows the visible state of the retina to be recorded at a particular point in time, allowing documentation and a deeper and extended analysis.

Based on deep learning algorithms, we provide you with an AI fundus photographic analysis system that can automatically screen and diagnose retinal diseases, such as Diabetic Retinopathy (DR), Retinopathy of Prematurity (RoP) and Age-related Macular Degeneration (AMD), without the need for a physician to interpret the results.

Model Functions

Diabetic Retinopathy Analysis
Based on the deep learning enhanced DR automatic detection algorithm and combined with the International Clinical Classification of Diabetic Retinopathy, our Diabetic Retinopathy (DR) automatic analysis function can filter out DR-free images, divide the DR images into a group, and provide 4 outputs: negative (none or mild DR), referable DR, vision-threatening DR, and/or low exam quality which noted a limitation in analysis or image quality, in addition, the system can not only identify vision threatening and referable DR but also AMD and possible glaucoma in a multiethnic cohort.

Figure 1 The Diabetic Retinopathy automatic analysis function

Retinopathy of Prematurity Analysis
Challenges in retinopathy of prematurity (ROP) screening and diagnosis include diagnostic variability, the paucity of specialists in ROP, and access to care issues. Our automatic ROP screening function based on machine learning can meet these challenges well. Advanced artificial intelligence technology enables our ROP automatic screening function to evaluate vascular changes, features of the disease such as zone or stage, category of disease, and disease progression, whose accuracy is non-inferior to human expert diagnosis, notably when identifying vascular changes such as pre-plus and plus disease.

Figure 2 The Retinopathy of Prematurity automatic analysis function

Age-related Macular Degeneration Analysis
Using multiple convolutional neural networks, we trained an algorithm to perform segmentation, classification, and prediction. Experimental results show that with this algorithm our Age-related Macular Degeneration (AMD) analysis function achieves comparable diagnostic performance in detecting referable AMD when compared to human graders.

Figure 3 The Age-related Macular Degeneration analysis automatic analysis function

Advantages

Detailed
The examination of fundus photographs is earlier than CT and other imaging examinations and requires more detailed observation. Manual reading will inevitably cause omissions because it often results in a tiny bleeding point or exudation point. The accuracy rate can reach 90%.

Fast
A doctor reads dozens of pictures throughout the day, while our AI can complete the analysis and judgment of a fundus photo in 10 seconds.

Reliable
Traditional fundus examinations are mostly ophthalmologists, whose cross-disciplinary knowledge background is rarely poor that can easily lead to misdiagnosis and missed diagnosis, while our AI imaging experts can provide interdisciplinary diagnosis recommendations based on accumulated medical studies.

CD ComputaBio has formed a team of experts excellent in imaging science and clinical domain knowledge, providing AI-driven solutions for fundus photography analysis according to your detailed requirements.