CancerZap!: Battleship Meets Where’s Waldo?
We were delighted to read the editorial by BioPhotonics
Group Publisher Karen A. Newman on the introduction of games such as MalariaSpot
as a means of recruiting people to help solve outstanding scientific and medical problems (May 2012
); two other games with this goal include FoldIt
Despite more than 40 years of the war on cancer, we are still dying of the disease at a substantial rate. It has astounded us that a simple approach – find and destroy the tumors at an early stage before they have a chance to spread1
– has received so little research. Search-and-destroy is a common mode in computer games, so we would like to formulate the problem as a game and solicit the help of game developers to make it a reality.
Let’s start with the popular game Battleship
, available in a plastic version from toy stores and as an online electronic game from many sources. When it’s your turn in Battleship
, you drop a bomb, hoping to hit one of your opponent’s ships. You get a simple, one-bit answer: yes or no. With these answers, you can build a picture of the battle scene, avoid bombing the same place twice and, more importantly, better guess where to drop the next bomb, based on your knowledge of the targets, of your opponent and of your previous hits and misses. The player who blasts all of the opponent’s ships out of the water with the fewest bombs wins.
In x-ray imaging, although the pictures are quite useful in detecting cancer, the x-rays themselves may have a small probability of causing cancer. Therefore, we would like to find tumors with as few x-ray photons (our “bombs”) as possible. Once our targets are found, bigger bombs (concentrated x-ray, radio wave, ultrasound, laser heating, electric cauterization or other ablation methods, or surgery) can be used to destroy the tumors.
But what we call CancerZap
! is played in a bigger, deeper ocean. The board for Battleship is a grid of 10 x 10 squares; the board for real-life CancerZap
! is easily 1000 x 1000 x 1000 cubes or more, each representing a small volume of tissue in the body that may or may not be part of a tumor.
Another difference is that the real “ocean” of our body is full of many harmless “fish and boats” that we do not want to destroy: normal bones and soft tissues that surround, contain or camouflage the tumors we are seeking to destroy. It is much like looking for Waldo in a sea of other objects.2
The opponent in Cancer-Zap
! is not the friendly Waldo, however, but deadly cancer itself.
Standard algorithms for generating x-ray images mindlessly blast photons everywhere, giving the person a hefty dose. We would prefer to seek and destroy tumors, while reducing the chance of starting new ones, minimizing “friendly fire” casualties.
So here’s what we propose: We have designed and written a C++ program that accurately describes the geometry of x-rays going through a volume divided into cubes as part of our research on improving computed tomography algorithms.3
We will make this available. Others have created 3-D computer simulations of tissues and tissue data sets, such as Virtual Man
and Virtual Woman
to which simulated tumors could be added.
A game needs to be programmed that allows the user to aim and shoot x-ray photons through a volume5
to try to locate the tumor with as few photons as possible. The source of the photons could emit photons at random, as tried by former undergraduate Carol Foo,6
or one at a time on command (“turnstile” photons7
). To simulate tumors that appear between checkups, data sets with and without simulated tumors could be provided. A platform could be written that allows gamers to develop effective strategies rather than just use their direct gaming skills. Tumor-detection problems ranging from easy to hard, 2-D to 3-D, would be included.
We are both retired, so we are in no position to lead a group to develop the CancerZap
! game. Newman has promoted ways for retired scientists to share their ideas with younger people (BioPhotonics, May 2011
), and she encouraged us to address this letter to game-savvy readers: Join us in the problem of finding small tumors with the minimum number of x-ray photons.
Embryogenesis Center, Gulf Specimen
Marine Laboratory, Panacea, Fla.;
Micro- and Nanotechnology Institute,
Department of Mechanical and Aerospace Engineering,
Old Dominion University
Glen D. Colquhoun
Retired scientific programmer,
Winnipeg, Ontario, Canada;
Thanks to Steve P. McGrew of New Light Industries for comments. Dedicated to the memory of Anne E. Rudloe.8
1. R. Gordon (2011). Stop breast cancer now! Imagining imaging pathways towards search, destroy, cure, and watchful waiting of premetastasis breast cancer. In Breast Cancer – A Lobar Disease
, Tibor Tot, ed. Springer-Verlag London Ltd., pp. 167-203.
2. R. Gordon and R. Sivaramakrishna (1999). Mammograms are Waldograms: Why we need 3D longitudinal breast screening [guest editorial]. Appl Radiol
, Vol. 28, No. 10, pp. 12-25.
3. G.D. Colquhoun and R. Gordon (2005). The use of control angles with MART (Multiplicative Algebraic Reconstruction Technique). Tech in Cancer Res and Treatm
, Vol. 4, No. 2, pp. 183-184.
4. U.S. National Library of Medicine (2012). The Visible Human Project. http://www.nlm.nih.gov/research/visible/visible_human.html
5. R. Gordon (1979). Feedback control of exposure geometry in dental radiography workshop, University of Connecticut, May 16, 1978. Appl Opt
, Vol. 18, pp. 1769 and 1834.
6. C. Foo (1999). X-ray imaging via intelligently steered x-ray microbeams (B.Sc. thesis; supervisor: R. Gordon). Department of Electrical & Computer Engineering, University of Manitoba.
7. C. Melvin et al (2002). A simulated comparison of turnstile and Poisson photons for x-ray imaging. Canad Conf Elect and Comp Eng. IEEE CCECE 2002
, pp. 1165-1170, IEEE.
8. J. Rudloe and A.E. Rudloe (2011). Chicken Wars
. Woodstork Press, Panacea, Fla.
Share your thoughts
Send us an email at email@example.com
. Submission of a letter constitutes permission to publish it in any form or medium. Letters may be edited for reasons of space and clarity.