It has been several years since AI technologies were formally introduced into medical settings. Coreline Soft's AI-based medical imaging analysis solutions have significantly improved the accuracy of chest CT imaging interpretation and operational efficiency, establishing them as a cornerstone of medical innovation. For instance, Coreline Soft's flagship product, 'AVIEW LCS Plus,' has been utilized in the national lung cancer screening interpretation project for eight consecutive years since 2017 and is currently used by approximately 90 medical institutions in South Korea. Additionally, it participates in lung cancer screening pilot projects in major European countries such as Germany, Italy, and France, expanding its presence in the global market.
A New Challenge of Balancing AI 'Accuracy and Efficiency'
As seen in lung cancer screening, sensitive AI technology proves beneficial for diagnosing severe diseases and is recognized as a means to enhance the overall quality of screenings. In some countries, the use of AI in medical practices has even become mandatory. However, increasing sensitivity to avoid missing abnormal findings can lead to a higher rate of false positives. In clinical settings, the diverse clinical situations contrast sharply with screening cohorts, resulting in more frequent false positives that can impede the efficiency of medical staff. In essence, while AI technology is introduced to assist with tasks, the process of verifying an abundance of identified abnormalities can become burdensome. At the 2024 Radiological Society of North America (RSNA) conference, many AI solution companies expressed that their concern surpassed mere accuracy, focusing instead on genuine workflow improvements. For medical AI to showcase its true value, it is crucial to effectively provide only the necessary information that supports clinicians' judgments.
Coreline Soft's Initiatives: Large Vision Model (LVM) and Large Multimodal Model (LMM)
To address these issues, Coreline Soft plans to develop new algorithms based on Large Vision Models (LVM) and Large Multimodal Models (LMM). Inspired by the large language models (LLMs) that made headlines with ChatGPT, where traditional AI imaging analysis solutions primarily rely on labeled data, LVM will apply unsupervised learning from vast clinical datasets to identify patterns; subsequently, it will be fine-tuned with labeled data to optimize it for specific medical domains. Meanwhile, LMM will incorporate other forms of data, such as clinical information and test results, alongside imaging data during the learning process. This approach is expected to enable AI to derive more refined and flexible answers and provide technologies capable of accurately identifying only those lesions that require careful attention from physicians in actual clinical settings. This will help prevent the distraction of medical professionals from unnecessary information and maximize the efficiency of patient care.
Coreline Soft's Strengths and Insight-Driven Technological Innovation
Coreline Soft's strength lies in the feedback obtained from actual users through early market entry. By consistently reflecting the discomforts and demands experienced by medical professionals in the field, we aim to establish ourselves as a company setting a new standard for thoracic AI imaging analysis products within the next decade. Insights garnered from these reflections translate into core competencies that go beyond mere technology development, fulfilling the practical needs of healthcare providers. Based on these insights, we will continue to evolve towards more refined and intuitive interpretations. This capability will position AI as a trusted partner in the medical field, transcending its role as a mere analytical tool.
A Vision Driving the Future Competitiveness of Medical AI
AI technology must now go beyond simple technological innovation and move toward creating tangible value in medical settings. Coreline Soft seeks to secure new competitiveness for its AI imaging analysis solutions through improved algorithms, aiming to lead the future of medical AI in the global market. By maximizing accuracy and efficiency, and transforming data into insights, Coreline Soft will continuously strive to provide better solutions in a changing healthcare environment.
About the Author
Junmin Park, Chief Product Officer (CPO) of Coreline Soft: Junmin Park, an emergency medicine physician, combines his clinical experience with the AI-based medical imaging field to plan and manage products that are practically helpful in the field. As the CPO of Coreline Soft, he plays a crucial role as a bridge between the requirements needed in clinical practice and the development of AI solutions, significantly contributing to securing global product competitiveness. He possesses deep insights into the perspectives of actual hospital settings and physicians while also understanding AI medical imaging technology trends and facilitating marketing for innovative medical devices and other products.
List