The field Semantic Biomedical Computing addresses biomedical information present in various forms including, but are not limited to, structured records, images, texts, devices and systems. It connects such contents and the intentions (of users/designers) via: 1.Semantic Analysis, that analyzes and converts signals such as pixels, words, equations, etc. (contents) to meanings (semantics); 2.Semantic Integration, that integrates the contents and semantics from different sources; 3.Semantic Applications, that utilize the contents and semantics to solve biomedical problems; and 4.Semantic Interface, that interprets users' intentions expressed in natural language or some other form(s) to create, access and manipulate the contents. The ultimate success of Semantic Biomedical Computing requires new, synergized technologies be developed from biology, chemistry, physics, medicine and computing. The Center for Semantic Biomedical Computing (CSBS) is established to facilitate the success of this emerging field of great promise.
Director Dr. Jeffrey J.P. Tsai, Chair Professor
Associate Director Dr. Phillip C.Y. Sheu, Visiting Professor"
Semantic algorithms are based in content and driven by natural language. Semantic algorithms connect human intent and calculable content to obtain, utilize, and process existing content according to the will of the user (to act in a way the user desires). The Semantic Biomedical Research Center strives to utilize semantic algorithm technology to analyze biomedical information and issues related to biomedicine. Current semantic research projects include:
- Semantic Search Engine – Search engines such as Google and Yahoo are based on keywords and lack accuracy. Semantic search engines are based on natural language and utilize questions/answers to quickly and accurately meet user needs. Semantic search engines emphasize the consistency of human thought to fully express the concept of “will”. Simultaneously, search results will not be a long string of index results as seen in standard search engines, but will be based in the results that users want.
- Semantics Service Engine – Web services are an expansion of shared resources in computer applications. Traditional web service engines require that users deconstruct a demand into a series of operations that can be identified by applications. Semantics service engines allow users to utilize will to drive services through natural language (e.x. analyze the genetic relationship between stroke patients and heart attack patients). Semantics service engines operate similarly to semantics search engines but it instead cooperates with web services. It must understand user will (problem), search for services that can provide a solution to the problem, and when necessary must combine these services to solve complicated problems.
- Semantic Medical Image Processing – Medical images are a medium often used in medicine to explore problems and diagnose diseases. There are many types of medical images but they are often straightforward and static; critical information is often hidden within and cannot be directly observed by the naked eye but require analysis through medical image processing technology and algorithms to discover new information and offer a new interpretation to images. Primary image processing technologies include image splicing, image recovery, image augmentation, image matching, and other algorithms such as morphological imaging. This is combined with machine learning, statistics, and related mathematical theories to supplement by exploring biomedicine clinical trials and related foundational research. When combined with semantic algorithms, users can utilize natural language to easily analyze biomedical images to quickly achieve desired results.
- Semantic Drug Development – Drug design is based on molecular structure. Computers create models that predict, select, or design micro molecular (called ligands) that work with drug target molecules to destroy or affect target drug molecules. Drug targeting often relates to critical ligands related to a metabolic path or signal of a specific disease, or the proteins that interact or survive with microorganisms of the disease. The simulation designs ligands that can be tested for biological activity to form QSARs with the target molecule’s structural design. This QSAR relationship can be used to design target drug molecules that are more suitable. Semantic algorithms can allow biology scientists to easily apply tools related to drug development.
- Semantic Systems Biology – Systems biology is developing at a much faster pace lately. Scientists utilize the different aspects of systems biology such as relationship or interaction (genetic manipulation, protein manipulation, and metabolic paths, etc.) to study and interpret the relationships between organisms on different levels. Systems biology mainly focuses on the relationship of affect between biological molecules, cells, and individual organisms. Physics and mathematic models are utilized in dynamic simulations of how organisms affect each other so that scientists can study the relationships of organisms through the concept of systems. Semantic systems biology combines with the concept of semantic algorithms helps to construct a semantic platform that assists systems biology researchers to quickly obtain dynamic simulation results and estimated mathematic models.