 |
Animals in Scientific Research
Science of the Future - Computer & Mathematical Modeling
Advances in technology are dependent upon advances in physics and engineering, but computer science and chemistry also have an integral role. In fact, the new field of bioinformatics represents a truly multi-disciplinary approach to scientific research, since it combines computer science, physics, math, engineering, and the life sciences.
Marvin Cassman, the director of the National Institute of General Medical Sciences (NIGMS) stated, “The future of the biological sciences will be driven by advances in bioinformatics and computational biology.”
Today, computer hardware and software giants like IBM, Sun Microsystems, Hewlett-Packard/Compaq, and Silicon Graphics are helping scientists meet the challenges of processing, managing, and manipulating the huge amount of biological data with supercomputers that can hold a terabyte or more of information. (A terabyte is equal to 1,000 gigabytes.) For example, the genomics firm Celera has 110 terabytes of storage in its computer farm, which is about 11 times the amount of information contained in the print version of the US Library of Congress.
To encourage the use of mathematical tools and approaches to study biology, the NIH has added a new center called the Center for Bioinformatics and Computational Biology (CBCB). The CBCB is not the first NIH center of the future. The National Institute of Biomedical Imaging and Bioengineering (NIBIB) was announced in 2000 and is dedicated to advances in medical technologies. NIBIB will coordinate fundamental research e.g., math, physics, and engineering, as it pertains to the imaging of disorders, diseases, and life processes of the body.
Ruth Kirschstein, then acting NIH director, stated, “While dedicating an institute to medical technologies…may seem novel for the NIH, it is truly a reflection of what science is today and where science will be taking us tomorrow.”
The pioneering work of scientists involved in the development of artificial neural networks (ANNs) is one of the most stunning examples of the integration of multiple disciplines in the advancement of medicine. ANNs, which grew out of an interest in learning and recognition, and involve scientists from the fields of biology, cognition, physics, computer science, statistics, and probability theory, are becoming an invaluable aid for researchers in analyzing and modeling complex data. They are, in essence, highly sophisticated statistical programs based on pattern recognition. The uses of ANNs are evident in virtually every medical specialty, from anesthesiology to neurology, radiology, laboratory medicine, and cardiology.
In cancer management, ANNs are utilized to predict the course of cancer on an individual basis, thus enabling clinicians to customize a treatment protocol based on those predictions.
ANNs are also being used to improve cancer diagnosis. Mammograms have been criticized for being very difficult to interpret; a study in the September 18, 2002 issue of the Journal of the National Cancer Institute demonstrated that the interpretation of mammograms varies widely among radiologist practicing in a community setting, with younger, more recently trained radiologists having two to four times more false-positive interpretations than older radiologists.
When used in conjunction with a radiologist’s interpretation of a mammogram, for example, ANNs allow for more accurate diagnosis of breast lesions.
ANNs are also speeding up the process of determining the type of bacteria that is causing infection in a patient, so that the patient can begin the proper antibiotic treatment as quickly as possible. While the use of cultures currently provides accurate diagnosis, the process can take days—even weeks, during which the patient may be using the wrong antibiotic; the technique using ANNs takes only minutes.
By combining an electronic stethoscope with an ANN, scientists can distinguish harmless heart murmurs in children from those that represent serious cardiac pathology. Scientists hope that in the future ANNs can even be used in “reverse genomics”—that is, quantifying the combined role of genes and the environment in disease incidence. In combination with epidemiology, ANNs can lead researchers to a greater understanding of how to avoid disease.
|
 |
