A theoretical 52 x 63 Grid Map of the contiguous US, presented using map images, with cells represented by each map measuring approximately 30 mile n-s x 40 mile e-w.

(examples shown below and elsewhere are mapped using a 20×20 or 25×25 mile grid cell area; the illustration above uses a sampling of my 3D maps to depict the theoretical grid)


“Long story short:
Corporate strategists need to get out of their 20th century mindset and into the 21st century.”

Prabhakar Gopalan’s 

Why corporate strategy needs to change with the cloud.  Posted January 1, 2013.



The Values of Imagery for Standard Reporting

Imagery–the value of images in reporting is that one image can reveal 100fold more that a standard report.  The mind, through visualization processes, can understand a drawing or illustration much more quickly that it can a page or half-page of writing.  The former requires the use of automated senses and logic, including the necessary problem solving skills important to survival.  The latter requires time and contemplation, a few milliseconds more than the typical image.  When you multiply this difference by a thousand fold or more, you are talking about second to minutes, and the continuous need for reflection, rearrangement of what you learned, for new discoveries to be made using a mostly left hemispheric learning experience.

The advantage to using maps, be they cognitive memory maps or maps we can relate to like an anatomical chart, an acupuncture meridian figure, the pages of our microbiology textbook, a satellite image, or a real map, is that conclusions may be drawn from these much more quickly.  And these conclusions that we draw are partially common sense and intuitive, basic on imagery not words.  Words need to be defined, and the redefined, and then the meanings of those definitions placed into context.

An illustration can also produce optical illusions so to speak, but excellent images provide us with the truest facts presented in the truest way, and with just a little bit of coaching or teaching on how to interpret an image like a map, we learn to make more sense from a map than we can ever deduce from extended verbal reports.  The left mindset is simply not the best way to go in an information age where spatial data tells you more than linear, word-by-word, line-by-line of report data.  In other words, or images, a page with 15 to 20 illustrations of our results displayed says much more than a twenty to thirty page report replete with endless tables, of which only the first five or ten rows of data are reviewed due to such complexity, tells us more than an entire report can ever tell us when it comes to making good corporate decisions.

This map for example tells you where exactly the problem with childhood immunization is in this country:


Note: A little later on in this page you’ll see that this does not tell us where outbreaks is a problem, which appear in a 9 maps grid below.  Refusals to immunize do not necessarily present a public health issue due to the herding effect.  But it does demonstrate how a 3D detailed map of distributions tells us more than a state or county wide made of the same, and can be used to produce more accurate contour maps than the irregular polygons provided by census block groups and block.

This method of analyzing data is what defines a slow moving company from a fast mover, fast generating company that is very much ahead of its time.   The rules of creativity and genius is that to be that far ahead, there has to be few others like you.  Right now, there are very little companies taking on the complex 3D task of understanding the world and its numbers.  There are companies out there that like to think otherwise though.  They think that by adding new technology, by applying old techniques to more information, that somehow they are ahead.  I can take a picture of every square mile surface of the world, produce a global map of it, and try to claim that map is ahead of its time.  But there is nothing new projected by those more detailed images, or my new way of presenting it.  Today’s companies are applying some new technological ideas to larger amounts of data, rehashing older statistics for the most part.  There are almost no companies that can tell you if their 90 to 100 million sample represents the nation as a whole for US statistics.  That is because they still have yet to find that place in their path where they next to learn how to review and make that decision.

What is the probability that 100 million out of 360 million people are going to represent the country?

If you cannot answer this question, you are already significantly behind in the Big Data market.  And you are behind that first less than one percentile who are way ahead of you in terms of the innovation bell curve.  This means you are essentially a bell curve company, at most a follower, and innovative mostly because you think you are, at least with the knowledge of how to use spatial statistics.

The reason I make all these statements about business intellect is as follows.  We are a left-brained world, a linear data world not a spatial data world, a 2D mapper not a 3D mapper.  Its kind of like the Khan versus Kirk situation Spock eludes to in a scene near the end of the Wrath of Khan movie, when they are moving through space and cannot see what lies next to them until it’s staring them in the face.  Spock says something like: Think of yourself as thinking 3D, whereas he (Khan or the opposition) is thinking 2D.

Who wins in the end?


A standard 2D county map presentation–how well does this answer the question . . . 

Were I to begin my Beacon Program for interventions today, where would I disperse national funds for this?  In other words, if you were establishing hierarchies for funding the most needy first, going down through the tiers over time, in what 5 counties would you begin this activity, followed by where and whom, what ethnic groups, etc.?

The use of 3D imagery in mapping relates to corporate needs in terms of cost savings, competition and preventive care health programs development.  In the above map, 3D imagery tells you where the peak is in similarly classified regions, be it peak in prevalence, counts, cost, density of cases, or any other spatial measurement.   These can all be better evaluated using the 3-dimensional methodology.  All four of the above metrics including maps can be produced in Teradata and the other non-GIS system in less than 1.5 hours (usually 5 minutes each, but let us be on the safe side).  It is the responsibility of corporations to keep up with this aspect of the spatial data movement and with GIS, even if that means changes in management.

Few companies work in an eidetic fashion with their data.  Until recently, GIS was considered being too creative, too “outside the box” by big business.  Companies, especially big companies, have been mostly linear in their methodologies.  Even when spatial attempts are made, due to the inherent non-eidetic nature of most of today’s leadership and workforce, biologically and behaviorally, companies never truly work at an eidetic level.  tables and graphs are used or simple descriptive maps, not any detailed spatial or 3+dimensional spatial interpretation of data useful for prediction spatiotemporal behaviors.

The main thing about an eidetic company is that it is by rule a creator of innovations.  Eidetic companies not only produce what is expected of them.  Internally, they produce many times more information, enabling them to be a generation of time ahead of nearly all of the others they are competing with.  Since eidetic companies explore ways of pulling information together that are not only fast but novel, they have the edge in the competition. They are that industry that one ideally wants to be with if and when he/she is truly of the creative type, and bearing the knowledge base and skills needed to be this way.

Most businesses unfortunately for the most part are two dimensional in nature.  That is they think in terms of flat paper–reports, graphs, tables, and when they enlist maps into their skillset, only 2D mapping done at some descriptive, isopleths level, not at a point specific, high targeted level, designed to make and measure amount of change.  The basics of corporate statistical modeling  have forever been only descriptive in nature, reporting this accurately perhaps, but making horrendous mistakes sometimes when they try to take that leap into the statistical analysis, statistical validity based methods.  Descriptions and measurements of changes in numbers such as average cost per year per individual or product, or the like, are simply not the same as accurate measurements of how, when and where each of these changes, one at a time, were statistically significant, not just grander in nature.  3D logic and 3D mapping is the way to begin your venture out of this narrow path that your competitors like to take in the business world.  2D may be easier, by 3D is better.  And in terms of illustration, use and productivity, 3D+ is the only way to succeed over your competitors.

In terms of health information mapping, OLAP, data mining, and better use of full data content, 3D imagery provides everything that a 2D model can produce, plus at least one addition bit of information that cannot be deduced from the 2D model.  A 3D model tells us where the peaks and valleys are.  With a 3D model we not only know those areas where the largest statistical outcomes may be seen, but also know exactly how each regions compares with its neighbor.  This is important when it comes to equal area analysis.

With Equal area analyses, when the value of a cell is greater than the next, this sets the stage for a diffusion process to ensue.  In the case of diseases and people’s behavior, this means that one area is performing a particular act or activity required for that large statistic to be placed on the map, meaning that these same measured events have a higher likelihood of flowing to the next place where there are less of these examples.

In the business world, items that may engage in such flow behaviors include new discoveries, unique behaviors, innovations, greater availability of money, higher concentrations of psychosocial and biological-ecological influence.  Areas where higher amounts of emission of pollution take place naturally lose that pollution as it dissipates to the surround areas.  Areas where large numbers of behaviors take place are likely to cause neighboring areas to engage in the same behaviors as some form of competition or mimicry, or simply taking advantage of the newly discovered behavior.

The following are methods of presenting data to clients who are in search of information on where the lags are in services, where costs are too high, where greater amounts of efforts need to be made to improve long term performance.  For the top row of images, the first image is an overlay of three metrics compared spatially, the second depicts a rare disease that is heavily clustered in a part of the U.S. just 45 miles wide and involving an major urban center within a high population density location, the third image is a krigged outcome for the data, meaning that it has been adjusted and re-displayed to take into account areal density as well as individual grid cell values.


This shows us that with regard to disease patterns, places where diseases prevail and even are most likely to erupt can behaviorally and ecologically be predicted using this method.  Since disease, human habits, behavioral patterns, do tend to flow within a population based upon density patterns, especially cultural and economic or income density, this means that flow from areas of high concentration to low concentration can be reviewed and predictions made about whereto proceed to prevent such flows from developing, or to stop them at some midpoint.

When that disease in dependent upon ecological causes and barriers, areas next door lacking that disease may remain vacant of an examples, until the means to break these social and ecological barriers develop.  When that disease is population density related, census data can be used to define potential flow patterns.  The first model in the above series demonstrates how different layers can be overlain, to demonstrate where risk lies relative to counts and prevalence or percentage patterns.  This way of reviewing multiple values spatially on just one image provides more information that a 20 or 300 page report on the same detailing all of these geographic findings, listing them in descending order in order to ease this information flow to the clients.

For this reason disease maps in which data is displayed in 3D format using very small areas, over large areas of research and display, provide us some of the most unique insights about a disease we can develop.  These insights are lacking from 2D models.

The value of this methodology can be demonstrated by the first image in the second row–on immunization refusals.  To data no immunization refusal study has been performed on the national data and linked to the development of better intervention programs and the implementation of more cost effective monitoring processes for diseases that normally are immunized against.  The above map demonstrates exactly where the social problem exists regarding refusal by mothers to immunization their children.  In the next image we see what immunizations demonstrate regional distributions that related to these refusal to immunize codes.  For some diseases, outbreaks occur randomly and the pattern of these outbreaks is not directly linked to refusal behaviors.  For others that are more common like mumps, rubella and pertussis, there is a spatial overlap of refusals to immunize with areas where the disease erupts due to population densities.  These two could be overlapped to show you the exact place where an outbreak is likely to occur.  This information can then be used to prevent such a public health problem from actually developing.


The 3D model presented here have the added advantage of being relatively high resolution, meaning we can actually see the transportation routes or natural topographically or climatologically defined routes that a disease or medical behavior can take.  With enough surveillance of examples to draw conclusions from, we can even apply this technique to developing more accurate spatial models of disease patterns than those developed without the use of spatial analytic techniques as their primary surveillance measurement tool.

In theory, the 3D models presented here use a method that can be applied to any statistic. Their presentation do not require any special training or understanding of the data, for the most part.  Some data can be mapped that is normally very virtual in its value, meaning it doesn’t become important to us until we have a better udnerstanding of the meaning of the slope, nearest neighbor value or degree of clustering evaluated and used to provide a p-value and related CI or standard deviation.  These features too can be mapped using the NPHG method, but for the most part any outcome generated using any method out there for evaluating spatial data will be more presentable and immediately understandable using the methods presented here.

The following are more examples of the 3D maps I generated by my method of evaluating the National Population Health Grid Data.


Example 1.  An immunization refusal map series.  The following depicts how these refusals are distributed for nine vaccinations that kids should have.  Some refusals are very much centered on the Pacific Northwest.  Others are more nationally distributed behavioral patterns, for understandable reasons, like the refusal of a tetanus shot.  The reason for such a broad distribution of pertussis (cough/croup) vaccine refusal may related to our perspective of this disease and its “severity” or “seriousness.”  Many of us are probably left with an image suggesting this disease is not very much different from common sore throat and cough, even though its long term consequences can be much more severe.



Example 2.  Comparison Two Disease Types to better understand cause and improve intervention planning processes.  A common paradigm for Japanese Takosubo diagnosis is that it is an example of a cardiomyopathy disease.  However, my findings suggest it is a disease with a similar onset and develop as the more common Munchausen’s Syndrome.    I base this on shared distributions for the two spatially across the U.S..  The 3D multilayer mapping technique effectively demonstrates this spatial relationship, which normally cannot be evaluated at the national level due to the difficulty of obtaining enough data.  Note the first and last population curves have very similar old age onset features, with female cases (pink) > male cases (blue) for this age range.    There is not gender difference in cardiomyopathy syndromes, and the idiopathic, psychogenic cardiomyopathy has a younger age as its onset.



Example 3.  Reporting on an Important Social Issue.  There are a number of diseases or behaviors that society becomes sensitized to in the news.  A good number of these pertain to child neglect and abuse.  With the right coding procedures, metrics can be established to define these social problems, map them out, and then compare them spatially with such variables as median income, areal gender, ethnicity or race patterns, numbers of unwed mothers in a given research area, regional GINI coefficients.   The following for example might be the header of a standard reporting page of important indicators of clustered child neglect or abuse patterns.  Psychological diagnoses, physical diagnoses, V-codes, Emergency Room coded visits, and a variety of long term abuse or neglect indicators can be merged into a report and a scoring method developed to define small area total risk scores.  A properly developed multilayer formula allows for standardized methods to overlay these individual 3D images onto a single image, with the new risk scoring formula used to define the z-axis variable, with a resolution of about 25 miles squared, or less with special programming.



Example 4.  Aggregated Disease Scoring and Analyses.  Some diseases are more naturally defined in their distribution than humanly defined.  In other words, they infect humans as opportunistic victims, and their distribution tends to mimic a population density pattern for a given area.  Such environmentally-linked diseases also often present with well defined borders, like the next disease which is due to a microorganism that likes to reside in perilacustrian (lakeside) settings due to a combination of temperature, humidity and growing medium requirements.  The disease displayed here is very much linked to the local environment, and is reported due to the large numbers of people residing in its natural habitat.



Example 5.  Research Question: Is homelessness distributed evenly across this country by age and place?  Answer: the age range for homelessness with a distribution closest to that of the national population is perhaps the 45-54 and 55-64 year old category.  But notice these two categories do show a more northern tendency.  A similar argument could be made for the 16 to 25 and 25 to 34 year olds, although the higher peaks on the west coast suggest homelessness in west could be more than in the east.  Most important to note in these maps are the clusters of very young (0-15 yo) homeless people documented in the first map, with a peak in the Pacific Northwest.  This peak is even high for the 16 to 25 year olds.  Streetwise homeless people (16-25) aggregate on the west coast due to lower population densities in and around urban settings with more supportive soup kitchens and the like, in addition to a few other social issues that occasionally hit the media (prostitution for example).  Another interesting peak to be curious about is the larger numbers of 65+ homeless people found in the Great Lakes region, close to Chicago.  Normally weather makes a difference, as west coast studies suggest.  This Chicago area peak warrants further investigation.



Example 6.  Infibulation. [for definition go to either wikipedia page (non-offensive) or Equality for Women page (a little more offensive), or finally, disturbing but truthful coverage by a reporter.]  Infibulation is a culturally-bound medical practice that is defined based on traditions and culture, not health care need.  It is usually associated with African and African American Muslim cultural settings.  A basic review of where it occurs provides some valuable information as to where in the nation this behavior exists and/or has been reported in the medical records.  A grant funded program that wants to target as many people as possible with little money needs to know where to begin.  This is how the use of various mapping equations come into play.  The first image is a standard distribution reports map, the second that map compared with the same values squared, in order to emphasize the highest density, high “risk” areas.  The third uses prevalence measures to determine percentages of local populations exist with high rates of reporting, (Muslim population dense areas are not necessarily going to demonstrate these high rates.)  Two other comparisons are made as well for prevalence and N or Nsq.  A hybrid mapping technique has been developed which allows for these data to be merged to produce a single 3D image, based on the logic defined in Example 2 for U.S. versus culturally-linked disease patterns. [Culturally-bound=linked to a culture based on its belief systems; Culturally-linked=more likely to occur in a given culture due to cultural physiology-anatomy-heredity reasons.]



Example 7.  Child Abuse by Kids.  Another important social issue.  The maps below demonstrate N, Nsq and IP (Independent Prevalence=without Kriging performed).  There are four ways to define high risk areas: 1) go by raw numbers because in spite of population density, cases must be prevented, 2) go by N squared to see if that provides you with a region to focus upon, 3) go by IP and use that to assign priorities for national programming of an intervention or education program, and 4) hybridize two of these into a single map to define high risk areas and well as highest need areas, and define priorities based on the way the equation is modeled.



Example 8.  Foreign Born Disease Patterns and Introduction.  Disease penetration as a form of accidental in-migration, unplanned infection and transportation of animal hosts, and through bioterrorism methods are three hot topics in modern disease mapping.  A standardized report for all possible microbial or foreign born agent diseases, in combination with animal born diseases cane be developed.  Nearly 200 such diseases, organismal and non-organismal, were tested using this method.  Some were associated with standard in-migration East Coast and Great Lakes routes, a few took the West Coast-Pacific Rim route of entry, and one or two diseases demonstrated either unexpected or unique penetrations such as via northern and southern borderland routes, or by way of tourism routes linked to in-migration followed by infectious outbreaks.



Example 9.  Occasional Outbreaks of Disease.  The standard in-migration of cholera in U.S. history has been introduction via the southern shoreline or borders.  Cases come in frequently due to certain foodways practiced in Mexico regarding seafood preparation (a pickled or seasoned raw fish recipe is the usually cause, but on occasion raw shellfish).   Outbreaks might also occur along the water edge in estuarine settings.  The standard CDC practice for evaluating disease risk is to map ocean surface temperatures, evaluate shoreline water chemistry, collect and test water samples for pH and form of alkalinity, and to monitor local host populations such as fish, shellfish, etc. in the natural niduses for these bacterial.  Genetic testing of vibrio enables specific strained to be uncovered.  In some cases warning are published regarding eating of potentially contaminated fish and seafood whenever ecological carriers of the disease have been identified.  The following sequence of scan-in images are focused on the nidus of Asiatic cholera’s vibrio species off the coast of Louisiana, within the Mississippi River Delta ecosystem.  Two niduses are identified on these images for the cases reported.



Example 10.  “Hot Spots” mapping.  Some diseases have “hot spots.”  A hot spot is where a disease tends to return to the setting in relatively large amounts due to unique features in that region.  These features can be physiographic, ecologic, or demographic in nature.  An example of physiographic and implied ecologic due to its biotic origin is the fungal disease Rhinosporidiosis.  This condition has two very prominent peaks in the eastern United States.  One in Florida and the other in southern New England.  Brazilian blastomycosis is possible an example of a break out of the condition.  Should this recurrence occur erratically, it is possible some sort of evolution of a natural ecological setting might commence.  The west coast-east coast spatial distribution here (Brazil-California) suggests an overland travel cause, making it more likely that certain ecological features may be assisting in this overland migration process (people and/or host-vector animals, combined with soil effects).  The single peak linked to Infibulation is found of the IP mapping technique, and could be an artifact of this process (see above Example 6).  The Off-road vehicle accidents peaks, were they to recur spatially, would present us with a unique moral and very regional issue to contend with (kids on ATVs) as intervention specialists.  The remaining two conditions demonstrating very narrow peaks are partially genetic and partially biological in nature.  A strong argument would be made for these two examples by accompanying each of these maps with a 2D krigged map of the same, thereby defining true hot spots with neighboring disease diffusion settings.



Example 11.  Foreign Born Disease In-migration Patterns.  Each country or region suspected of accidentally aiding in the in-migration of foreign born disease patterns has its particular sets of diseases to be on the watch for.   The list of ICDs linked to each well defined continental setting can be mapped as a group to see if different continents (or countries) have difference in-migration routes, as was carried out for the following maps.  Two versions were often used.  The first list was short and focused on diseases with well established histories of certain domains they normally occupy, such as Russian born animal-host-vector zoonotic disease patterns.  A second mapping routine was carried out by adding other infectious diseases and culturally-specific or culturally-linked patterns linked to an in-migration case.  The V2 images represent that group.  The two niduses that stand out on these maps and in need of confirmation are the Asia high prevalence/case rates along the Pacific Rim, an expected result, but so too was Russian disease in-migration patterns, which were not represented by the maps above.  The second dilemma pertain to the Middle and South American spike far north of the southern border.  This either suggests some coding problems in which a disease was included that perhaps is not expected to be from the south, or that there is some behavior that is responsible for there being a lack of connectivity between this middle-south American disease pattern and the nidus close to the northern border states.  A review of each of the diseases mapped for this evaluation demonstrated that a very specific Yucatan-Mexican born disease (not to be disclosed here) was responsible for this spike.  (One other Middle America vectored disease showed a similar pattern.)


Example 12.  In-migrating Diseases from the same Region.  Eight very country-specific in-migrating zoonotic-anthroponotic disease patterns are displayed here.   Population density impacts economic travel and trade, including livestock shipping.  So many of these diseases adhere to normal population density related expectations.  Louse-born Typhus and Louping Ill require reviews due to the unique ecological peaks they present.  For both, their peaks in the Pacific Northwest perhaps suggest a Pacific Rim trade routes relationship.  Alternatively, southern migration of these diseases from Canada is not atypical for this part of the country.  There are several well defined inland barriers (topographic/physiographic and phytozoologic) that prevent the spread of these diseases and their domestic animal causes from making their way eastward much due to animal migration.  The eastern range and Oregon desert followed by the Rocky Mountains provide excellent natural ecological borders for these two diseases.