Almost all of us have experienced the frustration of searching for a document we created in the past. Whether it is a spreadsheet of last quarter’s sales figures, a mission-critical report on 2007’s quality control improvements, or a list of contact names, we are so consumed with creating our documents that we often neglect to think about finding them later.
ADVERTISEMENT |
Enter the era of digital microscopy. It is not uncommon for quality control, manufacturing, or research and development (R&D) facilities to acquire from thousands to hundreds of thousands of images per year from their microscopes and other image-acquisition devices. Many of these images may also carry important data tags, such as supporting spreadsheets, measurement reports, or analytical data derived from these images.
In addition, many industrial organizations must track who created the data, which instrument was used to collect the images, what magnification and contrast method were used, what projects the images were attached to, what happened to those projects, whether an end product depicted in an image passed or failed quality control (QC) inspection, and so on. Although most of us are aware of these parameters when the sample is under the microscope, it is easy to skip over the process of making sure our data can be tracked easily in the future.
Supporting documentation such as reports or spreadsheets are hard to find without a master database; even in some of today’s most high-tech facilities, laboratory employees and QC professionals may still spend hours rummaging through folder after subfolder of images in search of a particular bit of data. Data security for QC images is also an issue. Some companies only consider how they can securely store laboratory images and data after experiencing a dangerous and costly breach.
Images and metadata
Thanks to modern database technology, such issues no longer need to be a concern. A well-structured microscopy-specific database allows users to organize documents and files in a highly structured and intuitive environment, customized to the specific requirements and workflow of the facility. Unlike large-scale, universal database application software, microscopy-specific database solutions typically are applications that directly plug into the image-capture and analysis software and are often sold as an optional add-on module (figure 1).
One benefit of the modular approach is its seamless integration into the image-capture and image-analysis workflow. For example, users can arrange for every captured image to be archived automatically upon acquisition. They can then perform image analysis on that captured image and archive the analysis results with the image. Another benefit of this approach is that important image metadata, such as acquisition time and date, camera name, objective magnification, calibrated pixel size, and more can be recorded at the time of acquisition and viewed in the database application. Finally, such programs can prompt (and in some instances, require) the operator to enter data in various fields. This valuable information is readily available at the time of image acquisition, and entering it then means it is never inadvertently overlooked. The information becomes permanently embedded with the image.
Data commonly recorded in a metallurgical or material-science laboratory environment during the time of acquisition might include:
• Project name
• Product or material type
• Name of the customer for whom the product was manufactured
• Lot number
• Microscope type—e.g., scanning electron, confocal, stereo, or metallograph
• Illumination or contrast techniques used—e.g., darkfield, brightfield, fluorescence, or differential interference contrast
• Sample preparation—e.g., etched, unetched, highly polished, and so on
• Did the product pass or fail QC?
• Operator name
• Additional comments
A designated database administrator can customize these fields in minutes (figure 2). Operators may enter data via drop-down menus or pick lists, restricting them to a library of predefined entries if that is desirable. Dates, integers, simple text, memos, and other input can also be accommodated, and many microscope-image database applications even synchronize field data with third-party database applications such as customer-relationship management (CRM) or laboratory information management system (LIMS) software, eliminating the need for redundant data entry, and making coordination and business management more efficient.
Finding, not searching
Database solutions are designed to help users find what they are looking for quickly and easily. When a microscopy-specific software solution is used, once all the metadata has been recorded and associated with a microscope image, searching for either images or data couldn't be simpler. Operators can perform a database search based on a single field (simple search) or a series of fields (complex search). For example, an operator can search for images acquired between July 7—July 10, failed QC, manufactured by Bob, or for customer ABC. Operators can save often-used search templates for subsequent reuse (figure 3).
In addition to searching, filtering allows an operator to display only data that may be of interest, hiding any nonrelevant data. For example, an operator may filter based on user name, displaying only one operator’s projects while hiding projects created by other operators. Searching inside already-filtered data is possible with many programs as well.
Security and reporting
Archived data may have little value if it is manipulated and tampered with by unauthorized users. As such, security is paramount for developers of modern microscopy-based database and image-analysis software. Most database platforms provide the ability to restrict unwanted users from accessing data. A user with administrative rights may grant individual users or user groups permission to insert and read records, for instance, but not to delete or update records. Some microscopy database software allows previously established user groups, such as those set up in popular operating systems, to be imported for easy transitioning to the new system (figure 4).
It is not uncommon for researchers, QC professionals, or other users to create reports in popular word-processing applications. These documents incorporate images acquired under the microscope, as well as image information, but inserting the data traditionally has been done manually. The best image-analysis and database software can create robust and usable reports automatically, saving time, money, and effort. Report templates can be defined so that when the operator drags an image into the template, all associated fields and metadata are populated in the report automatically. In addition, reports can coordinate multiple sets of measurements and analysis data to create comprehensive graphs and spreadsheets; the software then automatically integrates these into the report (figure 5).
Database implementation
Implementation of a large-scale database solution need not consume excessive resources. Image databases consist of both front-end and back-end components, each of which has its own implementation and support requirements.
The front end pertains to the direct user interface. Front-end components include the database folder structure, unique layouts and views, fields and pick-lists, user account permissions, and related aspects of the solution, and are typically defined individually in conjunction with the software provider. The most successful and efficient front-end implementations require careful advance consideration of what fields will be populated and what kinds of data and reports must be generated. Implementation with the assistance of the software provider can take as little as a few hours, if an organization has done its homework in advance.
The back end deals primarily with information technology (IT) and the server where the bulk data are stored; it is completely transparent to operators. Some microscope-specific, image-database software packages are compatible with the more popular back-end database server products, such as Microsoft SQL Server or Oracle. Many organizations already have a large-scale back-end server implemented for other applications, so corporate IT administrators can often help with implementation, maintenance, and back-up.
Digital microscopy has made tremendous advancements within the past decade and continues to be deployed rapidly in many industrial manufacturing, QC, and R&D facilities. This trend is continuing, and the quantity of images captured and archived from digital workstations is projected to increase dramatically during the next decade. To accommodate the explosive growth of digital imaging and save both time and money, organizations should think now about archiving images and related data in a structured, tamper-resistant environment that allows for quick and easy retrieval. A customized, easy-to-use, full-scale microscope-specific image database solution can be implemented with a surprisingly small investment of time and effort.
Add new comment