At the March 2012 annual meeting of the Association of University Radiologists, the Association of Medical Students in Radiology (AMSER) Lucy Squire and Association of Program Directors in Radiology (APDR)/American College of Radiology (ACR)’s keynote lecture was delivered by Dean Buonomano, professor of neurobiology and behavioral neuroscience at UCLA. Buonomano is the author of the 2011 book, Brain Bugs: How the Brain’s Flaws Shape Our Lives . In it, readers learn that the human brain dwarfs even our most sophisticated technology: the internet now contains 20 billion web pages connected by 1 trillion links, but the brain boasts 100 billion neurons connected by 100 trillion links, a vastly greater degree of complexity.
The title of Buonomano’s lecture was, Features and Bugs of the Brain’s Architecture—Effects on Memory, Decisions, and Pattern Recognition. Professor Buonomano argues that, as a computational device, the brain exhibits both strengths and weaknesses. These strengths and weaknesses have important implications for the performance of individual brains, groups of people, organizations, and whole societies. They also bear important implications for research, education, and clinical practice in radiology. Radiologists who do not know how to accentuate their strengths and compensate for their weaknesses will fall short of their potential.
According to Buonomano, the brain is very good at pattern recognition. For example, people can quickly locate the lone numeral 5 among a group of 2s. But if the numerals are tipped on their sides, the task becomes more difficult, in part because the figures no longer resemble numerals with which we are very familiar. Yet our natural aptitude for pattern recognition is not equaled by our computational ability. Humans will never outperform computers at determining the logarithms of 5-digit numbers. In short, we are much more adept at pattern recognition than computation.
The brain and digital computers differ from one another in important respects. The brain contains approximately 100 billion neurons, and each neuron has approximately 10,000 connections with other neurons. By comparison, a computer may have billions of transistors, but each one has fewer than 10 connections. The brain is very good at parallel processing, though it tends to be somewhat noisy and unreliable, and it operates at a frequency of less than 1 kHz. A computer excels at serial processing, exhibits much greater reliability, and can operate at speeds greater than 1 GHz.
To understand the strengths and weaknesses of the brain, Buonomano calls our attention to 3 of its specific features: pattern recognition, association architecture, and its 2 decision systems. In terms of pattern recognition, the superiority of the brain to computers is revealed by the widespread use of CAPTCHA—Completely Automated Public Turing test to tell Computers and Humans Apart—to verify that a visitor to a website is a human being and not a computer. Wavy and uneven words that are relatively easy for a human to read can prove all but impossible for a computer to decipher.
Though innately gifted at pattern recognition, the human brain can also improve with practice. Such perceptual learning is often feature- and orientation-specific. For example, radiologists tend to outperform novices at detecting characteristic radiologic findings, but perform no better at “Where’s Waldo?” . Such learning tends to be slow and occurs over weeks or months, not minutes or hours. As we learn, the properties of individual neurons actually change. For example, these neurons begin to respond to figures oriented in new directions. Such changes can also be seen across larger stretches of the brain. Portions of the sensory homunculus, once thought to be fixed, can expand with training.
Get Radiology Tree app to read full this article<
Get Radiology Tree app to read full this article<
Get Radiology Tree app to read full this article<
Get Radiology Tree app to read full this article<
Get Radiology Tree app to read full this article<
Get Radiology Tree app to read full this article<
Get Radiology Tree app to read full this article<
Get Radiology Tree app to read full this article<
References
1. Buonomano D.: Brain bugs: how the brain’s flaws shape our lives.2011.WW NortonNew York
2. Nodine C.F., Krupinski E.A.: Perceptual skill, radiology expertise, and visual performance with NINA and WALDO. Acad Radiol 1998; 5: pp. 603-612.
3. Gunderman R.B.: Biases in radiologic reasoning. Am J Roentgenol 2009; 192: pp. 561-564.
4. Manning D.J., Finst P., Gale A., Krupinski E.A.: Perception research in medical imaging. Br J Radiol 2005; 78: pp. 683-685.