Home About us Editorial board Search Ahead of print Current issue Archives Submit article Instructions Subscribe Contacts Login 
  • Users Online: 277
  • Home
  • Print this page
  • Email this page

 Table of Contents  
Year : 2020  |  Volume : 6  |  Issue : 1  |  Page : 1-8

How to make medical information comparable and searchable

Department of Orthodontics, UKSH, Kiel University, Kiel, Germany

Date of Submission09-Mar-2020
Date of Decision27-Apr-2020
Date of Acceptance06-May-2020
Date of Web Publication26-Aug-2020

Correspondence Address:
Wolfgang Orthuber
Department of Orthodontics, UKSH, Arnold Heller Str., 3, House B, 24105 Kiel
Login to access the Email id

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/digm.digm_4_20

Rights and Permissions

For accurate digital representation of original (e.g. medical) information we have to recall that any information is selection from a common set of possibilities. This one dimensional or multidimensional set is called “domain”. Letters and language vocabulary are only special examples of domains which lead to non-reproducible digital representation. Since introduction of the internet, however, it is technically feasible to define domains online. These “adapted domains” can be optimized in dependence of the application (e.g. medical diagnosis). Then all relevant features of the original information in this application can be represented and transported digitally by comparable and searchable “Domain Vectors” (DVs). This would be important e.g. for medical decision support, therefore the introduction of Domain Vectors is urgently recommended. The setup of a comfortable internet presence for definition of adapted domains (and Domain Vectors) would be a first step for this.

Keywords: Domain, domain vector, domain vector, information, standard

How to cite this article:
Orthuber W. How to make medical information comparable and searchable. Digit Med 2020;6:1-8

How to cite this URL:
Orthuber W. How to make medical information comparable and searchable. Digit Med [serial online] 2020 [cited 2020 Sep 30];6:1-8. Available from: http://www.digitmedicine.com/text.asp?2020/6/1/1/293505

  Introduction Top

Medical findings frequently contain complex information, with history data, laboratory data, pictures, etc. For the best medical decision in such a situation we need experience, at best experience from anonymized global statistics from “similar” cases. Every medical practitioner would be interested in such specific advice from global experience, if it is practicable. For this diagnostic specific similarity search within global experience would be necessary, much better guided, and finer than text-based search. Currently, this is out of range, despite introduction of the internet already decades ago. References which describe the current state and the focused standards in medical informatics are given below. Despite these efforts, up to now we have not the necessary globally uniform diagnostic specific, searchable and comparable representation of medical information. However, this aim is so important, that it is worth to focus on it consistently: Obviously, health care could improve very much if we could precisely compare and search within growing global medical experience. Especially in today's age of increasing information overload, targeted neutral search of just the individually relevant information is even becoming more and more necessary for a good therapy. This goes beyond language-based search and requires reconsideration of the fundamentals. First, we need a clear and complete definition of “information.”

  Information Always is Selection from a Domain Top

Any digital information consists of numbers. Every number selects from an ordered set. We call this set domain.[1],[2],[3]

Information always is selection from a domain. (1)

The domain is an ordered set of possibilities and should be identical for all participants of the conversation. Information is transported digitally as number sequence which is a selection from a domain. The definition of the domain is the groundwork of the digital representation of information. It must be clear for sender and receiver of information. An important and up to now unused technical possibility: The domain can be defined online [1],[2],[3],[4],[5],[6] and it can be adapted to the application (without detour through language vocabulary). At this is full freedom for defining the domain and the numbers [7] which select in the domain and so transport digital information. This already indicates that (1) prepares for a very powerful approach to optimize the digital representation of information. Of course it has also the potential of a new important field of research. Its focus is the optimal online (and therefore global) predefinition of the domain of digital information. Currently, this is not done. Up to now typically the bits resp. numbers which build digital information are defined by context in very variable way, for example, using the domain “language vocabulary.” This can be improved very much. To understand this, we first need to know about the advantages and disadvantages of language based information representation.

  Why to Overcome the Restrictions of Language Based Information Representation and Search Top

For language-based conversation multiple (words and phrases as) elements of the domain “language vocabulary” are combined. This increases the count of possible messages nearly exponentially with the count of words. Hence, we get a wide range of meaning using language vocabulary whose size is limited due to the limited capacity of our brain. Our brain is adapted to language vocabulary as domain. It is essential means for oral communication since humans exist. We use it also for written information exchange. This initialized new important ways of communication. Now, we have even digital means and the internet as common source of information. We can use, for example, multimedia applications and see that much more is possible (and often even necessary) than information transport by language vocabulary.

Nevertheless, we still tend to restrict to it in parts of our thinking. For example, global search of information is primarily restricted to search for certain word combinations. At first glance we may be satisfied with the result, because on the web is much text and we can find many short word combinations. However, usually, we search original information. For text-based description, a certain original information “ORGINFO” is mapped to free text in a certain language, and this text is mapped one-to-one (bijectively) to its digital representation “DIGINFO” using, for example, Unicode.[8] If ORGINFO itself is text (e.g., in case of literature or specific names, keywords or titles), it can be efficiently searched by text.

But very often, for example, in case of life science and also medicine, the original information ORGINFO belongs to reality. If it is nontrivial information about reality, for example, a certain medical finding, the description by language usually needs many words and mapping from ORGINFO to its digital representation DIGINFO by word combinations can be done in very variable way. Thus, the probability that we find the most relevant occurrences of original information ORGINFO by a certain word combination DIGINFO becomes small. In addition, the mapping to free text destroys original similarity. There is no general concept which represents “similar” original information ORGINFO by “similar text.” However, search of similar ORGINFO would be important, e.g., in medicine for collection and statistical comparison of groups of patients with “similar” findings. Hence, we need a clear strategy for language independent uniform conversion of ORGINFO to its digital representation DIGINFO which preserves original similarity.

  Global Definition of the Domain of Digital Information Top

After defining a variable or colloquially “after defining a number” we know the meaning of all of its possibilities, i.e., we have defined its domain. Thus, we could define the domain of a certain part of digital information globally by global definition of the number sequence which forms its digital representation DIGINFO. In the current infrastructure its online definition on the internet is the appropriate means for this. We called the defined number sequence together with a global pointer (efficient link) to the online definition “domain vector” (DV).[1],[2],[3],[4] The structure of the DV is

UL plus number sequence (2)

At this “UL” is abbreviation of “Uniform Locator”. The UL is an efficient global pointer (link) to the online definition of the following number sequence. Binary numbers can be designed very efficiently, based on positive self-extending integers [Table 1]. These can be used also in the UL. The UL has a similar addressing function like the URL [9],[10] in a link, but can be designed more efficiently as hierarchical sequence of numbers [Table 2]. To avoid misunderstandings, we used the new term “UL” (”Uniform Locator”).[1],[2] The DV is consequently designed for maximal efficiency according to technical arguments: For transport of as much as possible data using as little as possible energy. There is no discussion necessary. Using the online definitions, editing software can be designed very user-friendly also on binary DVs. Due to the large amounts of digital data, every unnecessary bit is simply a waste of energy. In addition, the online definition should be efficiently designed for quick access to its frequently downloaded parts. Nevertheless, the complete online definition can be large, with additional explanations and also multimedia parts, if wished. Thus, there is a clear distinction between transported data (DVs) and explanatory data (globally uniform online as DV definitions). The latter are not transported with DVs, but are quickly downloadable through UL, if necessary or wished. The online definition should also include a distance function,[1],[2],[11] so that the domain becomes a metric space. This is not difficult for well-defined data, and an objectifiable criterion for comparability. Only if there is a well-defined distance function, DVs with the same UL are comparable and available for precise similarity search. Defining the domain of information in this way online and so globally is a powerful tool for optimizing the representation of digital information in comparable and searchable way. At this, for best usefulness, we have to adapt the domain of the number sequence (2) of the DV to the application.
Table 1: Bits of the 3 short variants of a positive self-extending integer, starting with a half byte. Bit0 and Bit1 represent the count “ADD” of additional half bytes, the other bits (2 bits, 6 bits or 10 bits) represent the mantissa (“M” can be “0” or “1”). If this is not enough, then Bit0=1 and Bit1=1 and further bits will be used for length info. This principle allows arbitrary precision. Floating point numbers can be composed of two such integers with sign bits

Click here to view
Table 2: Exemplary structure of the UL as hierarchical number sequence N0, N1 and further numbers. This allows different subgroup sizes which are completely adaptable to the requirements

Click here to view

  Adapting the Domain of Information to the Application Top

Optimally just the application specific relevant original information ORGINFO is mapped globally uniformly, bijectively (one-to-one) and under preservation of original similarity to its digital representation DIGINFO. Concretely, this needs an application specific and global definition of the number sequence in the DV which forms DIGINFO. For this, we have to ask repeatedly:[1]

Which (additional) independent features are relevant in this application? (3)

Which variants of every feature are differentiable? (4)

If there is a natural order of the variants, we apply it, else we can introduce a useful order, so that most “similar” variants are neighbored. In case of multidimensional order every dimension is regarded as independent variable. After this the variants of all variables are ordered from “small” to “large.” A concrete (selection of a) variant is represented by a number.

Because the features with variants and the representing numbers are applications specific, the experts who know best about this application are most qualified to define these. For this they need a comfortable online presence with interface, where they can provide the online definitions of the numbers which represent the variants of the features which are relevant. The principle has been explained [1],[2],[3] and demonstrated.[7]

  Adapting the Domain of Information in Medicine Top

The better our preknowledge, the more precisely we can think about the most important parameters in a certain situation. In medicine the starting point is an initial rough diagnosis, for example, a selection from the International Classification of Diseases (ICD)-10.[12] Obviously just the professional users, which are specialists of this diagnosis, have the best knowledge for answering the above questions (3) and (4). Therefore, it is important that users can define domains online and also extent the definitions of domains by additional dimensions resp. numbers. For illustration of the procedure we provide an example. [Figure 1] shows a fracture of the left femur. The ICD-10 code [13] of [Figure 1]a is S72.402A: “Closed fracture of lower end of left femur.” It is the initial diagnosis. Which parameters may be interesting in this situation? Besides routine data like date, time, age, body mass index, results of bone mineral density measured by Dual Energy X-ray Absorptiometry (DXA), and laboratory (also genetic) data which inform about general health state, we think about further possibly interesting features. Because it is not an open or comminuted fracture, we do not need complicated anatomical differentiation. For therapeutic decisions, we are mainly interested in the position of the fracture, especially its relative distance from the knee joint. Length of line A in [Figure 1]b shows the shortest distance of the fracture from the lower end of the femur and length of line B serves as measure of the condyle diameter and reference. Such a reference on the same image is useful for comparison purposes. We define FRADISTANCE: = A/B as quotient of length A and length B. The number FRADISTANCE quantifies the relative distance of the fracture from the knee joint. The larger FRADISTANCE, the greater is the distance. This is relevant for therapeutic decisions. In this case, FRADISTANCE was near to 1 and the patient got a retrograde femur nail as therapy, as shown in [Figure 1]c.
Figure 1: X-ray images of fracture of left femur (a) without (b) with auxiliary lines for localization and (c) after therapy

Click here to view

The definition of FRADISTANCE is an example. The general principle is, that the practitioner provides a first rough diagnosis (e.g., from ICD-10) and then automatically gets online the adapted evaluation software with definition of parameters (numbers) which are relevant in this case. These should be unambiguous and objectifiable, so that there is a reproducible one-to-one (bijective) relation of reality to the parameters resp. numbers which build the digital representation. This provides a clear guide for conversion of the decision relevant reality into a comparable and searchable number sequence as digital representation. Creation of the software for (half) automated feature extraction and automatic reproducible generation of the representing numbers is much programming work, but globally necessary only once and the algorithms are globally reusable. Hence, we can avoid redundancy and progress very efficiently. Search can be directly applied for sharing experiences: For example, we can search for therapeutic decisions at patients with 0.8<FRADISTANCE<1.2 and similar routine data (BMI, age, gender) and look for frequency of later possible complications. If the data are available, a universal numeric search engine can provide immediately statistics over different treatment groups. Thus, a single search result could provide more relevant (globalized and individualized) results than today a typical scientific study. The extraction of subgroups can be done e.g. by conditional search or similarity search, as demonstrable by our search engine prototype.[7] Screenshots of it are shown in [Figure 5], [Figure 5] and [Figure 5] in a former publication.[2]

  Sharing Results of Research Top

Strictly speaking, all therapeutic and diagnostic data can be regarded as research data, because they are useful for research. Sharing data for research is in any case important.[14],[15],[16] At this, due to the increasing inundation with data, researchers and clinicians need a searchable data structure and a way to find just the information which is relevant for the current situation (e.g., “FRADISTANCE” in the previous chapter). Comparability is necessary for this. As described above, the online definition makes all DVs with the same UL globally comparable and searchable. Therefore, sharing treatment and research data as online defined DVs would increase efficiency of research and help to avoid unnecessary redundancy. Also, in case of new diseases and urgency, this could help quick sharing of detailed experiences inclusive laboratory results.[17] If these are stored in the DV form (2), they are prepared for global comparability and searchability. This is generally important for data reusage.

  Problem: Perceived Current Level of Knowledge Top

Strangely enough, the above in (1) recalled exact approach to information as selection from a well-defined domain, and the natural conclusion, to define domains online and to use globally defined DVs (2) for information transport, is unusual up to now. A domain resp. “set of possibilities” as precondition of information is an aspect which has been routinely skipped. Perhaps it has been regarded as self-evident fact and not worth to think twice on it. As consequence the explicit (online) definition of a domain for the definition of application adapted information is unusual. However, it is a technical fact that online definition of domains of information is a very powerful and universal tool for definition and generation of globally comparable and searchable digital information. Such information is language independent and can be just adapted to the application where we need it. Already more than a decade ago we started to publish [18],[19],[20],[21],[22],[23],[24] about the unused and large technical potential of user defined number sequences resp. vectors for information transport. We assumed that science is enough alert to see this, but this approach was simply ignored. We noticed other priorities [25] and found unexpected deep problems in understanding this. There have been comments like “While the presented ideas may be interesting from a theoretical standpoint, I find it hard to see any way of applying them in practice.” This indicates a difficult starting position for this approach. If asked experts have no idea how to define number sequences as useful information carriers (2), we need restart teaching fundamental knowledge and a catch-up in education which cannot be done in a short paper. The above first chapters recall the nature of “information” as selection (1) from a domain resp. “common set of possibilities” and the secondary derived nature of belated combinations of information using language or other semantic concepts. Global definition of the domain has primary and fundamental importance. Examples like shown above [FRADISTANCE, [Figure 2] can only touch the general potential of globally defined number sequences. We have to recall that semantic [25],[26] and syntactic [27] a posteriori combinations of information need a basis and that combinations introduce additional variability of digital representation. Also, DVs (2) can be combined a posteriori and such combinations are searchable by combining the ULs in a search. Because DVs can transport very complex information, such combinations can be much shorter than combining small units like terms of language. This reduces combinatorial complexity and resulting freedom for the variability of digital representation. Maximal reduction of variability is possible by usage of only one UL for one application. For this, it is also possible to combine existing online definitions through their ULs in a new combined online definition and to use its UL. Such a combined online definition is like a precise guide for preparation of complex application specific digital information (as number sequence). It is important that experts in information science are familiar with the definition of number sequences (as selection from a domain), with dimensionality, with (by distance function [11]) well-defined similarity and resulting comparability of information. If this is not guaranteed, how should we compare information? There is the temptation to make an arbitrary or biased estimation with corresponding quality. In case of online definition, we could avoid arbitrariness and transparently compare DVs with the same UL in well-defined way using just the distance function which has been online defined for this application. It is clear that this can free us from other cumbersome detours with increased susceptibility to errors. Here, we can only recall that all digital data are number sequences and the potential of this is well known. Furthermore, deep interoperability problems due to the variability of local definition of the digital number sequences by context are well known. We hope that scientists step by step recognize that the DVs (2) (as globally identified and defined number sequences) have the potential to introduce a new era of globally defined and uniformly structured information, which is comparable and searchable using definitions of an (to the application) adapted domain and without detour over language vocabulary or other variable syntax constructions in transported data. At this, the DV structure (2) consciously is optimized for maximal technical efficiency [Table 1] and [Table 2]. Independently of this detailed online definitions used by appropriate editing software can make the efficiently transported information of DVs (2) clear, intuitively accessible and also can show it in detailed explanatory way if wished.

  Current Approaches and Standards Top

Meanwhile, there are many approaches for publishing data on the internet with leading role of the semantic web [25],[28] Rules and approaches with a posteriori combination of information are focused. As mentioned above, this needs a basis. But, where is the definition of the combined information? Where can we define (adapted to the application) the precondition (1) of information exchange: Common sets of possibilities resp. domains? These should be metric spaces for similarity comparison. The metric space concept is not new and books are available about this.[11] However, we have not found the basal application, where users can define adapted domains of information online and so globally and later perform similarity search in such domains according to just those criteria (3) (4) which are important in a certain application. Despite existence of the internet since decades, online definition of adapted domains (1), i.e., definition of adapted and comparable information by users, is not possible. For essential applications libraries with definitions of certain quantitative data have been placed online, in medicine for example LOINC.[29] However, we need general comparability of user defined and user generated information. This is much more and important for real applications. In medicine, the Fast Healthcare Interoperability Resources (FHIR)[30] standard is central. At this, a lot of defining code and syntax is transported together with data. We urgently recommend to place defining information in online definitions of common domains, these need not to be transported together with data and these are surely globally uniform. Hence, these can automatically provide rules for uniform and bijective conversion of original information ORGINFO into its digital representation DIGINFO as DV (2). Such online defined DVs only need to transport the core information. In [Table 3], which shows an excerpt of a FHIR example,[31] the core information is represented by the values of concentration, dates and names of patient and practitioner. This could be efficiently transported as DV (2). It is identified by the UL which also is a pointer to its online definition. Further metadata are not necessary. We recommend to place all defining and explaining code into the online definition. Existing definition libraries for comparison of quantitative data in FHIR, for example, LOINC,[29] could be converted into online definitions with well-defined ULs, so that these are uniformly addressable and directly available. They can be also used for definition of (dimensions of) adapted domains, to get comparable information in dependence of the diagnosis or application. Up to now, FHIR does not focus definition of such domains. But, it allows extensions.
Table 3: Example “hemoglobin” from “Fast Healthcare Interoperability Resources” which is a standard for health care data exchange, published by HL7®. The core information is represented by the values of concentration, dates and names of patient and practitioner

Click here to view

  First Proposed Steps Top

The introduction of DVs (2) as FHIR extension as new efficient user defined data could be a first compromise. At this DVs could be represented as strings using, for example, base64 in RFC4648,[32] where every letter transports 6 bits. For definition of DVs one of the first steps is establishment of a functional and reliable online presence, where users, especially medical experts, can comfortably generate online definitions of DVs in their field of expertise. Step-by-step more and more adapted domains could be defined, so that DVs could take over more and more tasks. DVs are not only very efficient, their online definitions also provide more security, because they are in principle stable and a quickly accessible uniform reference, so that DVs with the same UL are comparable and accessible to precise similarity search. DVs can transport complex structured information globally uniformly directly after online definition for more and more special applications. Such efficient internet controlled automatisms using online definitions are appropriate, human brain is overchallenged by the increasing count of rules from all sides.

  Conclusion Top

The basal definition of information as selection from a domain (1) needs new attention. Domains can be defined online, globally and adapted to any field of interest. The resulting information is language independently globally comparable and searchable. Due to the huge technical potential of online definition of domains of information, information science needs new appropriate focus on the definition of such domains and on the natural terms “similar information”, similarity search and metric spaces as domains. As consequence, there should be familiarity with globally defined and identified number sequences resp. DVs (2) as selections within globally defined domains. A concrete first step for their introduction would be the establishment of a reliable online presence which allows users to comfortably generate online definitions of DVs to their special field. After this, DVs can be defined to represent any digital information in comparable and searchable way.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.

  References Top

Orthuber W. Information is selection – A review of basics shows substantial potential for improvement of digital information representation. Rev Int J Environ Res Public Health 2020;17:2975. Available from: https://www.mdpi.com/1660-4601/17/8/2975. [Last accessed on 2020 Apr 25].  Back to cited text no. 1
Orthuber W. Global predefinition of digital information. Perspect Digital Med 2018;4:148-56. Available from: http://www.digitmedicine.com/article.asp?issn=2226-8561;year=2018;volume=4;issue=4;spage=148;epage=156;aulast=Orthuber. [Last accessed on 2020 Mar 03].  Back to cited text no. 2
Orthuber W. Online definition of comparable and searchable medical information. Rev Art Digital Med 2018;4:77-83. Available from: http://www.digitmedicine.com/article.asp?issn=2226-8561;year=2018;volume=4;issue=2;spage=77;epage=83;aulast=Orthuber. [Last accessed on 2020 Mar 03].  Back to cited text no. 3
Orthuber W, Hasselbring W. Proposal for a New Basic Information Carrier on the Plus Number Sequence. Proceedings of the 15th International Conference; 2016. p. 279-84. Available from: http://oceanrep.geomar.de/34556. [Last accessed on 2020 Mar 05].  Back to cited text no. 4
Orthuber W. Why Informatics and General Science Need a Conjoint Basic Definition of Information. Arxiv Preprint arXiv: 1801.03106; 2018. Available from: https://www.arxiv.org/abs/1801.03106. [Last accessed on 2020 Mar 05].  Back to cited text no. 5
Orthuber W. Uniform Definition of Comparable and Searchable Information on the Web. Arxiv Preprint arXiv: 1406.1065; 2014. Available from: https://www.arxiv.org/abs/1406.1065. [Last accessed on 2020 Mar 05].  Back to cited text no. 6
Orthuber W. Demonstration of Numeric Search in User Defined Data. Numeric Search Viewed; May, 2018. Available from: http://www.numericsearch.com. [Last accessed on 2020 Mar 05].  Back to cited text no. 7
Allen JD, Anderson D, Becker J, Cook R, Davis M, Edberg P, et al. The Unicode Standard 2012. Mountain view, CA. Available from: http://citeseerx.ist.psu.edu/viewdoc/download?doi= [Last accessed on 2020 Mar 05].  Back to cited text no. 8
Berners-Lee T, Masinter L, McCahill M. Uniform Resource Locators (URL); 1994. Available from: https://www.ietf.org/rfc/rfc1738.txt. [Last accessed on 2020 Mar 05].  Back to cited text no. 9
Berners-Lee T. Universal Resource Identifiers in a Unifying Syntax for the Expression of Names and Addresses of Objects on the Network as used in the World-Wide Web. (No. RFC 1630); 1994. Available from: http://www.rfc-editor.org/rfc/pdfrfc/rfc1630.txt.pdf. [Last accessed on 2020 Mar 05].  Back to cited text no. 10
Zezula P, Amato G, Dohnal V, Batko M. Similarity Search: The Metric Space Approach. Vol. 32. New York, USA Springer Science & Business Media 2006.  Back to cited text no. 11
World Health Organization. ICD-10: International Statistical Classification of Diseases and Related Health Problems: Tenth Revision; 2016. Available from: http://apps.who.int/classifications/icd10/browse/2016/en. [Last accessed on 2020 Mar 05].  Back to cited text no. 12
The Web's Free 2019/2020 ICD-10-CM/PCS Medical Coding Reference; March, 2020. Available from: https://www.icd10data.com/. [Last accessed on 2020 Mar 05].  Back to cited text no. 13
Borgman CL. The conundrum of sharing research data. J Am Soc Informat Sci Technol 2012;63:1059-78. Available from: https://onlinelibrary.wiley.com/doi/full/10.1002/asi.22634. [Last accessed on 2020 Mar 05].  Back to cited text no. 14
Walport M, Brest P. Sharing research data to improve public health. Lancet 2011;377:537-9. Available from: https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(10)62234-9/fulltext. [Last accessed on 2020 Mar 05].  Back to cited text no. 15
ACMG Board of Directors. Laboratory and clinical genomic data sharing is crucial to improving genetic health care: A position statement of the American College of Medical Genetics and Genomics. Genet Med 2017;19:721-2. Available from: https://www.nature.com/articles/gim2016196. [Last accessed on 2020 Mar 05].  Back to cited text no. 16
Lim J, Jeon S, Shin HY, Kim MJ, Seong YM, Lee WJ, et al. Case of the index patient who caused tertiary transmission of coronavirus disease 2019 in Korea: The application of lopinavir/ritonavir for the treatment of COVID-19 pneumonia monitored by quantitative RT-PCR. J Korean Med Sci 2020;35:e79. Available from: https://www.jkms.org/DOIx.php?id=10.3346/jkms.2020.35.e79. [Last accessed on 2020 Mar 05].  Back to cited text no. 17
Orthuber W, Fiedler G, Kattan M, Sommer T, Fischer-Brandies H. Design of a global medical database which is searchable by human diagnostic patterns. Open Med Inform J 2008;2:21-31. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2666959/. [Last accessed on 2020 Mar 05].  Back to cited text no. 18
Orthuber W, Sommer T. Searchable Patient Record Database for Decision Support. MIE; 2009. p. 584-8. Available from: https://www.ncbi.nlm.nih.gov/pubmed/19745378. [Last accessed on 2020 Mar 05].  Back to cited text no. 19
Orthuber W, Dietze S. Towards Standardized Vectorial Resource Descriptors on the Web. Vol. 2. In GI Jahrestagung; 2010. p. 453-8. Available from: http://cs.emis.de/LNI/Proceedings/Proceedings176/453.pdf. [Last accessed on 2020 Mar 05].  Back to cited text no. 20
Dietze S, Benn N, Domingue J, Orthuber W. Blending the Physical and the Digital through Conceptual Spaces; 2009. Available from: http://oro.open.ac.uk/23019/1/Dietze-BlendingReal&Virtual.pdf. [Last accessed on 2020 Mar 05].  Back to cited text no. 21
Orthuber W, Papavramidis E. Standardized vectorial representation of medical data in patient records. Med Care Compunetics 2010;6:153-66. Available From: http://www.ebooks.iospress.nl/volumearticle/13127. [Last accessed on 2020 Mar 05].  Back to cited text no. 22
Orthuber W. How to Make Quantitative Data on the Web Searchable and Interoperable Part of the Common Vocabulary. In GI-Jahrestagung; 2015. p. 1231-42. Available from: http://cs.emis.de/LNI/Proceedings/Proceedings246/1231.pdf. [Last accessed on 2020 Mar 05].  Back to cited text no. 23
Orthuber W. Collection of Medical Original Data with Search Engine for Decision Support. MIE; 2016. p. 257-61. Available from: http://www.orthuber.com/SHTI228-0257.pdf. [Last accessed on 2020 Mar 05].  Back to cited text no. 24
Decker S, Melnik S, Van Harmelen F, Fensel D, Klein M, Broekstra J, et al. The semantic web: The roles of XML and RDF. IEEE Internet Comput 2000;4:63-73. Available from: https://ieeexplore.ieee.org/abstract/document/877487. [Last accessed on 2020 Mar 05].  Back to cited text no. 25
Jackendoff R. Semantic Structures. Vol. 18. Cambridge, Mass., United States: MIT Press; 1992.  Back to cited text no. 26
Benferhat S, Dubois D, Prade H. Syntactic combination of uncertain information: A possibilistic approach. In: Qualitative and Quantitative Practical Reasoning. Berlin/Heidelberg, Germany: Springer; 1997. p. 30-42.  Back to cited text no. 27
Ristoski P, Paulheim H. Rdf2vec: Rdf graph embeddings for data mining. In: International Semantic Web Conference. Cham: Springer; 2016. p. 498-514. Available from: https://madoc.bib.uni-mannheim.de/41307/1/Ristoski_RDF2Vec.pdf. [Last accessed on 2020 Mar 05].  Back to cited text no. 28
McDonald CJ, Huff SM, Suico JG, Hill G, Leavelle D, Aller R, et al. LOINC, a universal standard for identifying laboratory observations: A 5-year update. Clin Chem 2003;49:624-33. Available from: https://academic.oup.com/clinchem/article/49/4/624/5641953. [Last accessed on 2020 Mar 05].  Back to cited text no. 29
Bender D, Sartipi K. HL7 FHIR: An Agile and RESTful approach to healthcare information exchange. In: Proceedings of the 26th IEEE International Symposium on Computer-Based Medical Systems. IEEE; 2013. p. 326-31. Available from: https://ieeexplore.ieee.org/iel7/6607262/6627745/06627810.pdf. [Last accessed on 2020 Mar 05].  Back to cited text no. 30
HL7 FHIR Specification v4.0.1, Observation-Example-f005-Hemoglobin, Generated on 06th March, 2020. Available from: https://www.hl7.org/fhir/observation-example-f005-hemoglobin.xml.html. [Last accessed on 2020 Mar 05].  Back to cited text no. 31
Josefsson S. The base16, base32, and base64 data Encodings. RFC 4648; 2006. p. 1-18. Available from: http://www.hjp.at/doc/rfc/rfc4648.html. [Last accessed on 2020 Mar 05].  Back to cited text no. 32


  [Figure 1]

  [Table 1], [Table 2], [Table 3]


Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
Access Statistics
Email Alert *
Add to My List *
* Registration required (free)

  In this article
Information Alwa...
Why to Overcome ...
Global Definitio...
Adapting the Dom...
Adapting the Dom...
Sharing Results ...
Problem: Perceiv...
Current Approach...
First Proposed Steps
Article Figures
Article Tables

 Article Access Statistics
    PDF Downloaded57    
    Comments [Add]    

Recommend this journal