aidot

[Startup Today 18th. Dec. 2019.15:33] Artificial intelligence (AI)-based biotechnology startup Buzzpole Co., Ltd. announced the completion of the development of a stroke prevention solution jointly researched with Hallym University Chuncheon Sacred Heart Hospital and Gangwon Information Culture Promotion Agency. The name of this software is ‘AI Stroke 

Threat Screening System’, and Buzzpole began objectively verifying system performance ahead of domestic and international market development.

Dr. Dong-hyun Lee, who is in charge of Buzzpole's research, said, "All of Buzzpole's software that incorporates cutting-edge technology uses at least three algorithms."

An official from AIDOT said that “For example, the AI-based ‘cerebral aneurysm diagnosis software’ and ‘carotid artery ultrasound screening software’, which have recently reached the completion stage of development, have a higher level of accuracy at the level of image reading than doctors, and will be developed into intelligent systems that help optimize doctors’ decisions in the future and we will spread solutions that fit the precision medical paradigm to the world.”

Established in June 2014, Buzzpole, a domestic biotechnology startup, emphasized the “philosophy of innovative companies dealing with artificial intelligence” as the background to 

spur innovation.

Dr. Lee said that Buzzpole's innovative technology started from understanding medical personnel and medical practices. He said, “Existing artificial intelligence diagnostic software 

that focuses on the field of image diagnosis focuses on the purpose of identifying normal or abnormal. Therefore, medical diagnosis and judgment in a mechanical way makes it feel as if the speed of human information processing is far behind artificial intelligence. This is the wrong shape and artificial intelligence should develop based on the philosophy of 

helping people.”

Regarding medical engineering, Dr. Lee pointed out, “(Medicine) is not piecemeal work, but mankind’s best technology and study that is created and developed through long-term 

research and clinical results.”

Buzzpole's explanation is that the self-developed algorithm to implement artificial intelligence-based software that helps prevent various diseases has a number of differentiating 

points designed with the goal of increasing clinical utility and accuracy.

Based on data analyzing morphological and biophysical factors that may be involved in the growth and rupture of existing Unruptured Intracranial Aneurysm, it compares and analyzes existing clinical studies in academia, and observes and predicts the natural course by obtaining advice from experts.

Dr. Lee emphasizing the advantage of software and said, “In order to accurately predict the risk of rupture in image data where the location of a lesion appears complicated depending on the imaging angle, such as a cerebral aneurysm, first of all, it is important to require images taken from various angles, and the second important point is detecting lesions and 

measuring the predictive value. Accuracy will be improved when these two things are possible.”  

In addition, he said, “We have overcome the blind spots and limitations of the existing image data classification method, and have our own algorithm that reconstructs the existing 3D network structure by describing the human brain recognition process regardless of the change in the angle or size of the image.”

It turns out that Buzzpole has laid the foundation for pioneering overseas medical markets by continuously securing artificial intelligence engineers from GE Healthcare in Germany.

Buzzpole Headquarters Won-Jun Park said that “This stroke prevention system plans to preemptively target the overseas market. In addition, we have already started exchanges with domestic and foreign university hospitals, academic societies, and research institutes, and are actively discussing with existing diagnostic imaging device manufacturers around the world.”t the core task of artificial intelligence startups around the world is the combination of big data, algorithms, and business models. In response, Buzzpole expressed its 

aspirations, saying, “We will prove it with results in the near future."

Copyright © Startup Today

[Reporter Soo-ah Go]