HIMSS Asia Pacific Innovation Hub

by Hanalytics Pte Ltd

What is Hanalytics?

Analysing medical images has been a manual and laborious process by doctors and medical staff, often with blind spots. Given the high volume of medical image data that require analyses by hospital staff and a severe shortage of medical staff to patients, we have invented a machine learning software that can assist medical doctors to identify abnormalities in both 2D and 3D medical images from CT and MRI scanners.

Our software employs our in-house proprietary artificial intelligent (AI) technology (similar to Google’s AlphaGo technology) that mimic the human brain. It not only identifies abnormalities but also predicts possible diagnosis with more than 90% accuracy. The suggested diagnosis can be further verified by the attending physicians who can then provide feedback to the system, further improving its intelligence.

What Makes BOSCH Eyecare Solution Innovative?

1) State-of-the-art algorithms for image analysis: Our proprietary deep learning technology is more sophisticated than machine learning technologies that existed in the industry today, in terms of i) intelligence, ii) speed and iii) scalability. Our AI algorithms are based on large-scale neutral networks and are capable of handling large volume of imaging data with labeled regions of anomaly. On the contrary, many existing machine learning algorithms are largely based on mathematical and statistical methods that are inferior in auto-inference of patterns from data and unable to handle large data efficiently.

2) Multifunctional diagnostic algorithms: Our proprietary technology encompasses three major functions: image segmentation, image classification and image auto-retrieval. These features have been designed with doctors’ requirements so that they can be operationalized directly in their daily work diagnosing patients’ conditions.

3) Self-improving accuracy: Our deep learning software is capable of continuous learning and ingest new data as soon as it is available whereas traditional techniques are not able to do so. The accuracy of the predictive analysis improves over time, empowering the junior doctors and radiologists, and improving the productivity of experienced doctors and radiologists.