The Growth of Digital Imaging/PACS: How to Survive the Deluge

Transcription

1 The Growth of Digital Imaging/PACS: How to Survive the Deluge Gary Wendt, MD, MBA - Vice-Chair of Informatics, Enterprise Director of Medical Imaging, University of Wisconsin-Madison John W. Melski,, MD - Clinical Informatics Medical Director, Marshfield Clinic Craig Schiefelbein - CEO & Co-Founder, Paragon Development Systems Sources of PACS Data Explosion 1

2 Implications of Multislice CT and High Speed MR on Enterprise Image Distribution Huge data sets Confusing to clinicians Image overload for radiologists Integrated PACS / 3D workstation 2

3 Imaging Outside of Radiology PACS vs EMR Why Use PACS Beyond Radiology? PACS is the best tool to manage medical images from ALL areas of the enterprise Understands the concept of like priors Head CT compared to Head CT R hand skin lesion compared to the same prior lesion The EMR manages clinical data Not exam centric 3

4 The Ideal Vision Other - Optical Diagnostics Information Generation Cardiology Workspace (new imaging exams) Workflow Information Imaging Generation Managemen (new imaging Radiology Devices exams) Workspace t CR/DR OrdersWorkflow CTInformation ImagingGeneration Acquisition Management MRI (new Devices imaging exams) QC Orders US CR/DR Processing Acquisition Fluoro Workflow CT Imaging Interpretatio QC Nuc Management Med MRI Devices n Processing PET Orders US CR/DR Product Transcriptio = Images Interpretation + Reports Acquisition Fluoro CT n QC Transcription MRI Processing Nuc Med Distribution US Distribution Interpretation PET Fluoro Transcription Nuc Med Product = Images + Distribution Reports PET Product = Images + Reports Institutional Image Storage Long-Term On-Line Networks Clinical Workspace Information Integration PORTAL Orders Results Clinical Notes Image Viewing Product = Patient Care Institutional Information Systems (HIS, RIS, CIS) Potential Image Sources Ear / Nose / Throat, OtolaryngologyO Microscope Imaging, Pathology Respiratory Imaging, Bronchscopy Gastrointestinal Imaging, Endoscopy Intra-Operative Imaging, Laproscopy Orthopedic Imaging, Arthoscopy Imaging of the eye, Opthalmology 4

5 Visible Light (DICOM VL) Enterprise image distribution will improve the speed and quality of patient care: Review prior images and reports Share images over the web for consultation, conferencing and teaching Any endoscope can be a digital modality Intelligent Appliance Plug and play,, installs in seconds Supports most models of all major scope vendors DICOM Clinical procedure specific UI Otolaryngology Endoscopic imaging of the ear, nose and throat Numerous types and styles of endoscopes Patient is conscious and aware Exams tend to be repeated at follow up Video frame rates can reach 200fps Standard 30fps, 10 sec = 300MB 200fps = 2GB for 10sec Requires synchronized audio 5

6 Nose/ Right Left Nasal Cavity Nasal Cavity Frontal Frontal Ethmoid Ethmoid Maxillary Maxillary Sphenoid Sphenoid Cancel [Esc] Nasopharynx Nasopharynx Nose/ Pathology: Now ~ 1867 ~ 1680 Microscope-based Subjective Local (glass slide) Labor-intensive 50x 100x 200x 400x 100:1 400:1 1600:1 6400:1 Search Diagnose 6

7 Pathology: Future Improved Analysis - Multiple fields & resolutions - Memory - Virtual slide(s) - Image Analysis - Annotation Archival & Retrieval Standardization Digital (Virtual Microscopy) - Quantification - Automation Objective Global (virtual slide) Assisted by automation 7

8 Benefits of Single Image Archive and Display System Analogous to working in a system with multiple EMR s Increased cost to maintain, train and support Multiple image silos compromise patient care ENT / CT / PET Correlation 8

9 ENT / CT Correlation Multi-institution and Multi-modality 9

10 Environmental Costs of Expanding PACS Storage Requirements All spinning costs Purchase Maintenance Electricity Cooling Benefits to integrated HSM (hierarchal storage management) Tape (or other nearline media) uses little or no energy to sit in slot Use of DICOM MWL to do intelligent prefetcs in event of disaster recovery Alternative is restoring all data of which only a small percent will ever be used 10