Tag: Taiwan

As dengue fever continues to rise globally, accurate data on disease burden is essential for informed public health planning and resource allocation. A recent study led by Professor Wei-Cheng Lo of Taipei Medical University examines discrepancies between the Global Burden of Disease (GBD) estimates and reported dengue case data in 30 high-burden countries, calling attention to the need for improved methodologies in disease modeling.

Understanding the Gaps in Global Estimates

The study compared GBD’s model-generated dengue estimates with official surveillance data from countries including Brazil, India, Indonesia, China, and Taiwan. The findings revealed substantial differences: in some instances, GBD estimates were several hundred times higher than reported cases. For instance, in China and India, the GBD estimated 570 and 303 times more cases, respectively, than national health data indicated.

In countries like Taiwan and Argentina, where dengue outbreaks vary dramatically by year, GBD figures showed relatively steady trends, potentially overlooking the episodic nature of epidemic spikes.

Modeling Assumptions and Their Limitations

The observed discrepancies are linked to how the GBD constructs its estimates. These models account for underreporting by adjusting data based on known limitations in surveillance systems. However, many of these adjustments rely on data collected before 2010. In locations where diagnostic tools and case reporting have significantly improved in recent years—such as Taiwan—current estimates may not reflect these advancements.

Additionally, the smoothing algorithms used to illustrate long-term trends may downplay sharp increases in case numbers during outbreak years, especially in regions with cyclic epidemic patterns.

Implications for Public Health Policy

Reliable disease estimates are a crnerstone of health policy and planning. When estimates deviate significantly from local data, they can influence policy decisions and funding allocation. This study emphasizes the importance of aligning global modeling with recent,country-specific data to better support public health decision-making.

Recommendations for Improved Disease Burden Modeling

The authors advocate for more frequent updates to global health models and greater integration of real-time surveillance and diagnostic advancements. They also suggest that future models incorporate the cyclical behavior of diseases like dengue to better capture the reality of epidemic patterns.

Broader Considerations

While this research focuses on dengue, it raises important considerations for global disease burden estimation more broadly. Refining modeling approaches across disease areas will support more effective global health strategies and ensure resources are targeted where they are most needed.

Founded by Professor Cheng-Yu Chen of Taipei Medical University (TMU), DeepRad.AI is transforming the future of radiology through the integration of advanced artificial intelligence technologies and medical imaging. With a mission to bridge the gap between clinical practice and AI innovation, the company leverages the expertise of experienced radiologists and AI engineers to enhance diagnostic precision, reduce interpretation time, and improve patient outcomes.

DeepLung-CAC: Dual Screening with a Single LDCT Scan

DeepRad.AI’s flagship product,DeepLung-CAC, is an AI-powered pulmonary-coronary screening platform that utilizes a single low-dose computed tomography (LDCT) scan to simultaneously assess the risk of pulmonary nodules and coronary artery calcification (CAC). This dual-purpose screening not only improves efficiency but also reduces radiation exposure and costs associated with multiple tests.

The platform is certified by the Taiwan Food and Drug Administration (TFDA) and recognized as the first domestically developed AI pulmonary-coronary screening system. One of its standout features is the LungRads module, which incorporates deep learning models for nodule detection, segmentation, and classification, trained on over 6,000 CT cases. Powered by state-of-the-art 3D deep learning, DeepLung-CAC can complete detailed analyses in under one minute.

Professor Chen emphasizes the significance of early detection in improving lung cancer survival rates. While traditional methods often require 25 to 30 minutes of physician review time, DeepLung-CAC reduces interpretation time to just 5 minutes—enabling radiologists to efficiently analyze images and provide quicker diagnoses. The platform also supports multi-disease risk assessment from a single scan, enhancing its
clinical utility.

DeepBrain-Cognito: Personalized Dementia Risk Assessment

Beyond thoracic imaging, DeepRad.AI has also developed DeepBrain-Cognito, a computational model designed for the early detection of cognitive decline. This platform provides personalized risk assessments for populations with suboptimal health, such as the elderly or those with chronic conditions. By integrating AI with large-scale brain imaging data, DeepBrain-Cognito enables timely intervention for neurodegenerative diseases such as dementia and Alzheimer’s disease.

Recognition and Future Directions

In 2024, DeepRad.AI’s innovations were widely recognized. The company received several prestigious awards, including:

• Future Tech Award
• NBRP Pitch Day Outstanding Team Award
• GenAI Stars Quality Innovation Award
• National Pharmaceutical Technology & Research Development Award

These accolades reflect the platform’s strong potential for real-world clinical applications and its role in driving forward Taiwan’s biomedical AI industry. By combining AI with medical expertise, DeepRad.AI continues to push the boundaries of radiology, aiming to create faster, more accurate, and accessible diagnostic tools for global healthcare systems.

The “Smart Innovation Platform for Preclinical Drug Development”, developed by Professor Shiow-Lin Pan’s research team at Taipei Medical University (TMU), leverages advanced AI technology to significantly reduce the time and costs for drug development.

The platform provides a faster, more accurate solution for developing inhibitors targeting severe illnesses such as cancer and neurodegenerative diseases, and integrates AI models to predict chemical compound characteristicsaccurately, enabling rapid identification of effective small molecule inhibitors. To date, the team has successfully developed nearly 30 inhibitors targeting different protein targets, reducing the average development cycle by 3 to 5 years compared to conventional methods.

Core advantages of the platform:

Interdisciplinary AI Model Development: An expert team in chemical synthesis, artificial intelligence, pharmacology, and toxicology collaborated to build and train the AI model. By combining insights from extensive R&D experience, the platform effectively addresses the challenges of designing new drug candidates that are synthesizable, patentable, and biologically active.

1.Novel and Feasible Drug Structures: The smart innovative platform develop a smart synthesis strategy, which ensures AI-designed drug structures have an 80% synthesis probability, 60% cellular efficacy (IC50 < 10 µM), reduced synthesis costs by 90%, and optimized with real-time validation data from the experimental team.

2.The platform’s advantages lie in its ability to rapidly generate novel chemical structures with higher prediction accuracy, which benefits patients worldwide by significantly shortening the time needed for drug development.

Professor Pan highlighted that the Smart Innovation Platform offers high-success-rate, easily synthesizable, rapid, and precise early-stage drug development services and high-potential early-stage drug products for small-molecule development needs. The platform has been adopted by multiple academic research units and biotechnology companies globally, attracting recognition from international pharmaceutical companies.

Looking ahead, TMU’s team aims to enhance the platform’s capabilities and expand collaborations with pharmaceutical industries in Taiwan and globally. By accelerating the drug development process , the team hopes to bring transformative treatments to patients worldwide.

Professor Shih-Min Hsia‘s research team from Taipei Medical University has collaborated with leading uterine leiomyoma experts, Professors Ayman Al-Hendy and Mohamed Ali from the University of Chicago, to investigate the key mechanisms driving uterine leiomyoma formation and propose innovative therapeutic strategies.

The findings, published in the high-impact journal Redox Biology (impact factor 10.7), highlight the importance and academic value of this research.

Uterine leiomyomas, common benign tumors in women of reproductive age, exhibit a high prevalence and significantly affect women’s health and quality of life. These tumors are characterized by excessive cell proliferation, extracellular matrix (ECM) accumulation, and stem cell-like properties.

The research team identified that Nicotinamide Adenine Dinucleotide (NAD⁺) metabolism and its key enzyme, Nicotinamide Phosphoribosyl transferase (NAMPT), are pivotal in the progression of uterine leiomyomas.

Analysis of uterine leiomyoma tissues revealed that elevated NAMPT expression is positively correlated with increased ECM accumulation and enhanced stem cell-like characteristics. Subsequent experiments using the NAMPT inhibitor FK866 and the vitamin B3 derivative nicotinamide (NAM) demonstrated that these agents significantly reduced uterine leiomyoma cell viability, attenuated stem cell-like properties, and effectively decreased ECM accumulation, highlighting their potential as therapeutic options. Furthermore, the team successfully obtained a patent in Taiwan for the use of nicotinamide as a treatment for uterine leiomyomas, solidifying the translational and clinical impact of their findings.

This study underscores the critical role of NAMPT and NAD⁺ metabolism in uterine leiomyoma development and emphasizes the promise of precision medicine interventions targeting NAMPT as a novel treatment strategy.

This international collaboration with the University of Chicago exemplifies the power of cross-border partnerships in elucidating complex disease mechanisms and developing innovative therapies, paving the way for new advancements in uterine leiomyoma research and treatment.

Colorectal cancer (CRC) stands as the third most common cancer globally, prompting urgent advancements in screening practices to enhance early detection and treatment effectiveness. Conventional population-based screening programs, while effective, often adopt an universal approach, potentially leading to the overuse of medical resources and unnecessary procedures for individuals at lower risk. A study from Taiwan introduces a compelling alternative—tailoring colorectal cancer screening intervals using fecal hemoglobin (f-Hb) concentrations to optimize both efficacy and efficiency.

The need for more personalized screening protocols has become increasingly apparent in recent years, as researchers and clinicians seek to improve patient outcomes while managing healthcare costs effectively. Precision medicine offers a promising path forward, particularly in cancer prevention, where the risk varies significantly from person to person. Utilizing a vast database of over three million participants, this study leads an approach that customizes screening frequency based on biomarkers indicative of an individual’s cancer risk. By aligning screening intervals with personal health data, this method enhances how we approach CRC prevention, making it more targeted and thoughtful in its application.

Conducted using a large-scale dataset from a Taiwanese biennial screening program involving over 3 million participants between 2004 and 2014, this study explores the potential of using fecal hemoglobin concentrations as a marker to personalize the screen-intervals in colorectal cancer screenings. Researchers found that higher f-Hb levels correlate significantly with an increased risk of developing colorectal cancer and higher mortality rates. By leveraging these insights, the study proposes varying the screening intervals based on individual risk levels indicated by f-Hb concentrations.

The research findings include:

• Participants with higher f-Hb levels are at a greater risk for colorectal neoplasia and cancer mortality, suggesting a need for more frequent screenings.
• Conversely, individuals with lower f-Hb levels may require fewer screenings without compromising the screening’s effectiveness.

Implementing this stratified screening approach could reduce the number of fecal immunological tests (FITs) and colonoscopies by 49% and 28%, respectively, compared to traditional biennial screenings. This reduction minimizes patient inconvenience and discomfort as well as significantly cuts down on healthcare spending and resource use.

The implications of such a precision-based screening strategy extend beyond the realm of colorectal cancer:

• Healthcare Efficiency: Streamlining screening processes ensures that resources are allocated where they are most needed, enhancing overall healthcare efficiency.
• Patient-Centric Care: Personalized screening intervals mean that patients receive care tailored to their specific risk, potentially increasing the effectiveness of screenings and patient compliance.
• Reduction in Over-Screening: By reducing unnecessary procedures, this approach minimizes the psychological and physical strain on patients and decreases the risk of complications from excessive interventions.

While the results are promising, the study underscores the necessity for further research, including randomized trials, to validate the practical benefits of personalized screening intervals across diverse populations. Additionally, combining f-Hb measurements with other biomarkers or risk factors could further refine screening accuracy.

This study demonstrates an evidence-based CRC screening example towards personalized medicine, representing a significant shift in cancer prevention and healthcare. It promises a future where medical interventions are reactive and proactively tailored to individual health profiles.

The study offers an innovative solution to the battle against colorectal cancer. By integrating a simple measure of blood in stool into screening protocols, we can make CRC screening more patient-specific, cost-effective, and, ultimately, life-saving. This approach heralds a new era of personalized medicine, where prevention strategies are as unique as the individuals they aim to protect, ensuring that the right patient receives the proper test at the right time.