Originally published in New Venture

The Artificial Intelligence Medical Research Center at Taipei Medical University, led by Vice President Dr. Cheng-Yu Chen, has developed the AI-powered multi-modal imaging precision health platform “Deep-Lung.” With just a single low-dose chest CT scan, the proprietary AI model can simultaneously predict the risks of osteoporosis, lung cancer, chronic obstructive pulmonary disease (COPD), and coronary artery calcification, achieving an accuracy rate of over 90%. The system also automatically generates recommendation reports that meet international standards, making it the first of its kind in the world.

Dr. Yen-Ting Chen, who oversees platform integration, stated that during hospital validation, the platform successfully identified lung nodules that clinical physicians had missed or failed to annotate multiple times…

As Taiwan’s population continues to age at an accelerating pace, the National Development Council estimates that by 2025, more than 20% of Taiwan’s population will be aged 65 and above, officially entering a super-aged society. According to statistics from the Ministry of Health and Welfare, cancer is the leading cause of death among individuals aged 65 and older, with lung cancer ranking first, followed closely by heart disease, chronic obstructive pulmonary disease (COPD), and fall-related accidents.

In response to this, Professor Cheng-Yu Chen of Taipei Medical University has led the university’s AI Medical Research Center team in developing the AI-powered multi-modal imaging precision health platform “Deep-Lung.” Designed specifically for the medical imaging needs of middle-aged and elderly individuals in Taiwan, this all-in-one screening system enables fast and accurate assessment of risks related to lung, heart, and bone diseases.

※ The image shows Professor Cheng-Yu Chen, CEO of the DeepRad.AI team. (Photo courtesy of Taipei Medical University)

Taipei Medical University Hospital’s “Deep-Lung” engine consists of four major modules: the LungRads module, CAC module, BMD module, and COPD module. These modules enable comprehensive, one-time assessments of lung cancer screening, emphysema, coronary artery calcification, and spinal fractures in middle-aged and elderly individuals. Additionally, they assist radiologists in diagnosis, report generation, and treatment planning.

In an interview, Professor Cheng-Yu Chen highlighted that this platform integrates artificial intelligence technology, significantly improving the efficiency and accuracy of traditional CT scan analysis, which previously relied on radiologists’ visual assessments. He emphasized that this advancement aligns with the future trend of developing personalized precision medicine.

Traditionally, conducting separate tests and diagnoses for lung, heart, and bone conditions could take up to six hours and expose patients to significant radiation. However, with Deep-Lung’s All-in-One engine, a single low-dose chest CT (LDCT) scan, analyzed by its AI model, can simultaneously assess lung, heart, and bone conditions in approximately 10 minutes. The system also automatically provides clinical treatment recommendations, reducing the workload on medical institutions while significantly minimizing patients’ radiation exposure.

※ The image illustrates the concept of the Deep-Lung AI multi-modal imaging precision health platform. (Photo source: Taipei Medical University)

In addition, the platform’s simulator is built using the EBM Technologies Incorporated system, making it compatible with most platforms and allowing direct integration with hospital PACS systems. Physicians can use the generated reports without altering existing clinical examination workflows or requiring additional training. In the future, the platform will continue expanding to radiology departments in major hospitals both domestically and internationally while also developing an online cloud-based version for public use. Currently, its collaborations span the globe, including partnerships with Shuang Ho Hospital and Changhua Christian Hospital in Taiwan, as well as clinical validations at Hokkaido University in Japan and UC Irvine Medical Center in the United States.

However, with the rapidly aging population, the future healthcare workforce is expected to decline by more than 20%. Cheng-Yu Chen acknowledges that the role of artificial intelligence will inevitably become even more crucial. He noted that Taipei Medical University has planned three major initiatives to address population aging. The first focuses on using AI-assisted screening during the sub-health stage. The second initiative targets lung cancer, where, in addition to the Deep-Lung platform, the TMU team has developed the world’s first Clinical Decision Support System for Lung Cancer. By analyzing clinical imaging with AI and integrating genetic data, the system provides personalized medication recommendations, which is expected to significantly improve treatment outcomes.

The third major initiative focuses on dementia. According to Brodmann area mapping, the brain consists of more than 40 functional regions, and cortical thinning occurs in different areas due to degenerative diseases. By using 3D imaging analysis combined with genetic parameters such as age and gender, it is possible to determine the specific type of degeneration and predict whether dementia is likely to develop in the future, as well as estimate the potential onset time.

In addition, Taipei Medical University has collaborated with the medical technology startup AetherAI Technology and several other enterprises to develop an AI-assisted, fully automated whole-slide imaging interpretation model for lung cancer digital pathology, eliminating the need for manual annotation and visualizing cancer lesions. They have also created the “Pathology Report Natural Language Processing (NLP) Automated Drug Recommendation System” and the “Comprehensive Genetic Drug Recommendation Model for Lung Adenocarcinoma.” Once a patient’s pathology report is generated, NLP technology analyzes 50 clinically relevant features from the text and identifies the most suitable treatment recommendations based on patients with similar conditions and the best prognoses.

Professor Cheng-Yu Chen stated that the establishment of the center originated from TMU’s decision to transform from a research-oriented university into an innovation-driven university, aiming to help patients through clinical translation. The center was founded with the initial intention of “beginning with the end in mind,” meaning that research is conducted with a clear goal from the outset. Since its establishment in 2019, the center has nurtured over 200 artificial intelligence research teams and has founded five AI startups. These companies focus on areas such as intelligent medical imaging, smart medication safety, sleep care, precision health, digital pathology, smart critical care, and bone density detection systems.

In addition to receiving frequent support from national projects, the center actively participates in industry-academia collaborations and international partnerships, with the goal of becoming the most distinctive AI medical research center in Asia and even establishing an AI-driven hospital. The team is currently striving to secure funding from the Ministry of Science and Technology’s Value Creation Program, aiming to further expand future innovation capabilities.

However, he also acknowledged that Taiwan’s current regulatory framework poses significant challenges, making it difficult to apply patient data for commercial purposes. Combined with fundraising difficulties, the overall startup environment still has room for improvement. He believes that by leveraging Taiwan’s ICT industry capabilities and strengthening international connections, it will be possible to cultivate more successful AI startups, similar to Korea’s Vuno and Lunit, which have successfully transitioned from university research teams to commercial enterprises.

※ The image shows the DeepRad.AI team. Third from the right is the team’s CEO, Professor Cheng-Yu Chen, and fourth from the right is COO, Dr. Yen-Ting Chen. (Photo courtesy of Taipei Medical University)