Medical Informatics Representation and Computation on Medical and

Medical Informatics Representation and Computation on Medical and Health Care Information

Medical Informatics • Health records in paper Problems: • High chances of damage of patient records. • Automated processing of records is impossible. • Increased use of electronic media, devices. Advantages: • Less chances of damage, ease of maintenance. • Data insertion and retrieval is simple. • Easily interoperable. • Migration of healthcare industry to electronic domain.

Targeted at Design and Development of Information System for Medical Science and Technology Health Care Services Business world involving health care

Health Care Informatics Suppliers’ Enterprise Systems Pharmacies Medical suppliers Insurance providers • • • E X T R A N E T S Providers’ Enterprise Systems ERP GDSS CDSS • • I N T R A N E T S Physicians And Specialists P R M Patients Internet Src: K. Siau, “Health Care Informatics”, IEEE trans. On Info. Tech. In Biomedicine, 7(1), March, 2003, pp. 1 -7

Enterprise Resource Planning Full integration of an organization's information from pay-roll and human resources to accounting and finance. Database integration (Information Sharing) Information logged once and accessed by different modules maintaining the data consistency. Track inventory, order information and delivery requirements Determine equipment usage and maintenance schedule

Decision Support System (DSS) Conventional DSS: financial and scheduling. Clinical DSS (CDSS): diagnosis, pharmacy, emergency and nursing practices. CDSS used to send alerts and reminders to patients about preventive care.

Patient Relationship Management (PRM) Primary focus on determining and meeting patients' needs. Tracking patients’ information from diet and exercise data to past diagnosis information from family history and allergy information. Send E-mails satisfying queries, informing newly published health care studies and reminding about preventive measures.

Medical Informatics and Tele -consultation • In early days tele-consultation means • Sending all hardcopies of patient records. • takes significant amount of time. • Send the patient to a remote center. • nearly impossible for emergency patient. • Using medical informatics tele-consultation means • Sending the medical records only. • Most of the time, no need to send the patient. • Online consultation among doctors. • Large scope of knowledge sharing.

Standards HL 7: Health Level Seven. It is an international healthcare standard for medical data exchange between computer systems in healthcare. http: //www. hl 7. org/ LOINC: Logical Observation Identifiers Names and Codes. These identify the test results or clinical observations uniquely. http: //www. loinc. org/ ICD-10: International Statistical Classification of Diseases and Related Health Problems. ICD provides codes to classify diseases and a wide variety of signs, symptoms etc. Every health condition can be assigned to a unique category and given a code. http: //www. who. int/classifications/icd/en/ ICD-10 -PCS: ICD-10 Procedure Coding System. This is a system of medical classification used for procedural codes which is developed as a replacement of ICD-9 -CM volume 3 (contains inpatient procedures). DICOM: Digital Imaging and Communications in Medicine. This is a standard for handling, storing, printing, and transmitting information on medical imaging. http: //medical. nema. org/

Message Transmission HL 7 Standard Sender Data HL 7 Message Creation HL 7 -enabled system Hospital A MSH|^~&|REGA EVN|A 05|199901 PID|1||191919^ NK 1|1|MASSIE^E NK 1|2|MASSIE^I … HL 7 Message HL 7 Standard Receiver HL 7 Message Parsing HL 7 -enabled system Hospital B Data

HL 7 Message Structure Event Message N Message 1 Seg. N Seg 1 Field 1 Comp 1 Terminated by <CR> Comp. N Separated by '|' Separated by '^'

Message Encoding Sequence of segments separated by '|' Compulsory and optional segments Segments as sequence of fields separated by '^' Compulsory and optional fields A field is described by a data type (e. g. AD data-type denoting an Address)

An example of HL 7 message for patient admission MSH | ^~ & | Clinic || Central|Reg ||| ADT^A 01 |MSG 00005 | P | 2. 3 EVN | A 01 | 199601051530 PID ||| 2 -687005 || Evans^Carolyn || 19620324 | F ||| 903 Diane Circle^^Phoenixville^PA^19460 || (610) 555 – 1212 || S | C ||156 – 96 – 2542 PV 1|| E | Emergency |||| 0148^Addison^James ||| SUR

Admit/Visit Notification 1. Message Header (i) From: Clinic (ii) To : Central 2. Event (i) Date: 1996 -01 -05 (ii) Time: 15: 30 3. Patient Identification (i) Internal Patient ID Number: 2 -687005 (ii) Family Name: Evans (iii) Given Name: Carolyn (iv) Birth Date: 1962 -03 -24 (v) Sex: F

Admit/Visit Notification (Contd. ) 3. Patient Identification (contd. ) (vi) Street Address: 903 Diane Circle (vii) City: Phoenixville (viii) State of Province: PA (ix) Zip or Postal Code: 19460 (x) Phone (Home): (610) 555 – 1212 (xi) Phone (Office): (610) 555 -1212 (xii) Marital Status: S (xiii) Religion: C (xiv) Social Security Number: 156 -96 -2542 4. Patient Visit (i) Patient Class: E (ii) Point of Care: Emergency (iii) Attending Doctor's ID: 0148 (iv) Family Name: Addison (v) Given Name: James (vi) Hospital Service: SUR

File as Reference Pointer (RP) of OBX MSH|^~&|TELEMEDICINE||200808141246||ORU^R 01|SUR|P|2. 4| EVN|R 01|200808141246| PID|||SUR 05032008000||Mandal^Pulin^Bihari||198003050000|M|||Block - D^VSRC, IIT Kharagpur^West Bengal^721302^India|91|754123||||Hindu| NK 1|1|Sumita Mandal|Mother|Parikpur, Hansda West Midnapore| OBR|4|||ZIITKGP 9903^Patient Images^HL 7 IITKGP|||200808141246| OBX||RP|ZIITKGP 9903 -1^Sample blood slide^HL 7 IITKGP||SUR 0503200800005032008 BLD 00. JPG^TELEMEDIK 2005^IM||||||X|||20080209||||||20080305| OBX||RP|ZIITKGP 9903 -2^Routine Blood tests and Grouping {Blood R/E}^HL 7 IITKGP||SUR 0503200800005032008 i 0000. JPG^TELEMEDIK 2005^IM||||||X|||20080305||||||20080305| • Multimedia file is not contained within HL 7 message. • Reference path to the file is kept in the message. • Multimedia file has to be sent with HL 7 message.

File as Encoded Data (ED) of OBX MSH|^~&|TELEMEDICINE||200705091251||ORU^R 01|SUR|P|2. 4| EVN|R 01|200705091251| PID|||SUR 17012007000||Kijhari^Punam^||196801170000|F|||aaaa^bbbb^eeee^West Bengal^897454^India|91|7855596||||Hindu| NK 1|1|kkkk|hhhh|gdagsv fg jfsgfgjadsg gdsa g fagsg| OBR|6|||ZIITKGP 9903^Patient Images^HL 7 IITKGP|||200705091251| TXA|1|IM|multipart|200705091251||20040404||||||ZIITKGP 9903 -1|ZIITKGP 9903||||AU| OBX||ED|ZIITKGP 9903 -1^blood slide^HL 7 IITKGP||^multipart^related^A^MIME-Version: 1. 0 Content-Type: multipart/related; boundary="HL 7 -CDA-border-CDA-HL 7" EXT: =JPG TYPE: =BLD ENTRYDATE: =20070117 SIZE: =16601 --HL 7 -CDA-border-CDA-HL 7 Content-Location: SUR 1701200700017012007 BLD 00. JPG Content-Transfer-Encoding: BASE 64 Content-Type: image/pjpeg /9 j/4 AAQSk. ZJRg. ABAg. EASABIAAD/7 RBKUGhvd. G 9 za. G 9 w. IDMu. MAA 4 Qkl. NA+0 AAAAAABAASAAAAAEA DAw. RDAw. MDAw. MDAENCws. NDg 0 QDg 4 QFA 4 ODh. QUDg 4 ODh. QRDAw. M Cg. Ay. ADAAMAAg. ADMANg. Az. AAo. A --HL 7 -CDA-border-CDA-HL 7 --||||||X|||200705091251|

Reference Information Model (RIM) Root of all information models. Provides a static view of the information. A HL 7 -wide common reference model that integrates all Technical Committees’ domain views. Committees and SIGs generally work with a small subset of the RIM - called Domain Information Model or DIM.

Reference Information Model (RIM) contd. . Foundation Classes Acts an intentional action in the business domain of HL 7. Ex: patient observation Participations exists only in the scope of one act. Ex: surgeon Roles a socially expected behavior pattern usually determined by an individual's status in a particular society. Ex: doctor Entities physical thing or organization and grouping of physical things. Ex: a person Act Relationship To relate 2 acts. Role Link To relate 2 entity roles.

LOINC Logical Observation Identifiers Names and Codes. Names & codes uniquely identify observations. Laboratory Observations Clinical Observations Administrative Observations Compatible with HL 7 and SNOMED Represent observation in HL 7 message.

LOINC The fully specified name of a test result or clinical observation has five or six main parts <Analyte / component>: <kind of property of observation or measurement>: <time aspect>: <system (sample)>: <scale>: <method> QRS AXIS representation in LOINC: QRS AXIS : ANGLE : PT : HEART : QN : EKG 2951 -2

LOINC in HL 7 Message generator generates a message with observation results using LOINC. System supports LOINC, parse HL 7 message, retrieve observation result. OBX-3: Observation Identifier OBX||TX|2093 -3^Total cholesterol^LN|0|78|^mg/dl|||||F|||20050223| Code Text Coding System

DICOM Digital Imaging and Communications in Medicine Standard for handling, storing, printing, and transmitting information on medical imaging. Specifies the following protocols for devices claiming conformance to DICOM syntax and semantics of data to be exchanged. format for storing media in DICOM compatible devices etc.

DICOM Composite Image IOD Information Model Patient 1 is subject of 1, n Study 1 1 creates Frame of Reference contains Equipment 1 1, n 0, n series 1 Spatially or Temporally defines contains 0, n Image (Pixels) 0, n Curve 0, n VOIL UT 0, 1 Modality LUT 0, n Overlay 0, n Waveform

DICOM Waveform IOD Information Model Patient 1 is subject of 1, n Study 1 Equipment contains 1 1, n creates 1, n Frame of Reference 1 0, n series 1 contains 1, n Waveform temporally defines

DICOM Waveform Information Model Patient 1 is subjec of Waveform Attribute Waveform 1 1, n Study 1 contains 1, n Series contains 1, n Multiplex Group 1 Time of Acquisition Context Annotation Multiplex Group Attributes Number of Channels Sampling Frequency Timing contains 1 Contains 1, n Waveform 1, n Channel 1 contains 1, n Sample Channel Definition Attributes Channel Source Metric Anatomic Location(s) Function Technique Channel Sensitivity Baseline Skew Filter Characteristics

DICOM File Structure Preamble (128 bytes) DICM (4 bytes) DE 1 DEn

Dicom Structure DICOM Data. Information format Data Element Tag < 4 bytes > Tag Data Element DE VR VL < 2 bytes > …………… Data Element VF Explicit VR VF Implicit VR < 2 bytes > VL < 4 bytes >

An Example of Data Element Patient Information Patient Name 0010 VR VL VF PN 000 C Sridhar Raja Explicit VR 0000 000 C Sridhar Raja Implicit VR

Some vital tags for rendering images in DICOM Standard Tag Description Transfer Syntax Tag Samples per pixel Numbers of Frames Number of Rows Number of Columns Number of Bits Allocated Number of Bits Stored Pixel Data Hex Encoding (Group, Element) (0 x 0002, (0 x 0028, (0 x 0028, (0 x 7 FE 0, 0 x 0010) 0 x 0002) 0 x 0008) 0 x 0010) 0 x 0011) 0 x 0100) 0 x 0101) 0 x 0010)

PACS Picture Archiving and Communication Systems. Goals of PACS are to improve operational efficiency while maintaining or improving diagnostic ability Computers or networks dedicated to the storage, retrieval, distribution and presentation of images. PACS network consists of a central server that stores a database containing the images. Web based PACS system is becoming more and more common. Based on DICOM standard, also accepts other media formats.

PACS (Ref: http: //www. advantech. com. cn/)

ICD-10 International Statistical Classification of Diseases and Related Health Problems. Provides codes to classify Diseases Signs, symptoms Abnormal finding Complaints External cause for injury and disease etc.

ICD-10 Every health condition can be assigned to a unique category and a given code. Can be used for Morbidity, mortality statistics Clinical decision support system. The limitations of ICD-9 -CM Lack of specificity and details. Can’t support transition of IHDE etc.

ICD-10 in HL 7 Segment for diagnosis: DG 1 2 nd field of DG 1: diagnosis coding method (deprecated). 4 th field of DG 1: diagnosis description (deprecated). 3 rd field of DG 1: diagnosis code ICD-10 is used to construct 3 rd field Identifier Code DG 1 -3: Diagnosis Code DG 1|||H 11. 2^Conjunctival scars^ICD 10|||F|||||||20050223| Diagnosis description. Coding System

ICD-10 -PCS (ICD-10 Procedure Coding System. ) Medical classification used for procedural codes. Codes are comprised of seven components. Each component is called a “character”. All codes are seven characters long Individual units for each character are represented by a letter or number. Each unit is called a “value” 34 possible values for each character Digits 0 - 9 Letters A-H, J-N, P-Z

SNOMED Systematized Nomenclature of Medicine. Collection of medical terminology covering most areas of clinical information. Diseases Findings Procedures Microorganisms Pharmaceuticals etc.

Standardization: Complexity in Medical World Vast and dynamic knowledgebase. Close interaction of different complex systems. Patient Management, Diagnosis and Investigations, Treatment and Procedures, Drug and pharmacology, Disease classification etc. Process Standardization. Regional and demographic variations. Adjustment with real-life constraints. Infrastructure, Human resource, Material resource.