DICOM PS3.17 2020a - Explanatory Information |
Page 419 |
UU Macular Grid Thickness and Volume
Report Use Cases (Informative)
UU.1 Introduction
Ophthalmologists use OPT data to diagnose and characterize tissues and abnormalities in transverse and axial locations within the eye. For example, an ophthalmologist might request an OPT of the macula, the optic nerve or the cornea in either or both eyes for a given patient. Serial reports can be compared to monitor disease progression and response to treatment. OPT devices produce two categories of clinical data: B-scan images and tissue measurements.
UU.2 Use of B-scan Images
Prior to interpreting an OPT B-scan (or set of B-scans), users must first determine if the study is of adequate quality to answer the diagnostic question. Examples of inadequate studies include:
•The pathology that needs to be visualized does not appear within the field of the scan.
•The image quality is not sufficient to see the tissue layers of interest (i.e., media opacity, blink, etc).
•The scans are not in the expected anatomic order (i.e., due to eye movements).
Insomecases,inadequateimagescanbecorrectedbycapturinganotherscaninthesamearea.However,inothercases,thepatient's eye disease interferes with visualization of the tissues of interest making adequate image quality impossible. Ideally, when choosing between multiple scans of the same tissue area, physicians would have access to information about the above questions so they can select only the best scan(s).
The physician may then choose to view and assess each B-scan in the data individually. When assessing OPT B-scans, ophthalmo- logists often identify normal or expected tissue boundaries first, then proceed to identify abnormal interfaces or structures next. The identification of pathology is both qualitative (i.e., does a structure exist) and quantitative (i.e., how thick is it). If previous scans are present for this patient, the physician may choose to compare the most recent scan data with prior visits. Due to workflow constraints, it may be difficult for B-scan interpretations to happen on the same machine that captures the images. Therefore, remote image as- sessment, such as image viewing in the examining room with the patient, is optimal.
UU.3 Use of Tissue Measurements
In addition to viewing B-scan image data, clinicians also use quantitative measurements of tissue thicknesses or volumes extracted automatically from the OPT images. As with image quality, the accuracy of automated segmentation must be assessed prior to use of the numerical measurements based on these boundaries. This is typically accomplished by visual inspection of boundary lines placed on the OPT images but also can be inferred from analysis confidence measurements provided by the device software. In ad- dition to segmentation accuracy, it is also important to determine if the region of interest has been aligned appropriately with the in- tended sampling area of the OPT.
The analysis software application segments OPT images using the raw data of the instrument to quantify tissue optical reflectivity and location in longitudinal scan or B-scan images. Many boundaries can be identified automatically with software algorithms, see Figure UU.3-1.
