Advantages and Limitations of Physical and Virtual Dose Mapping

Dose mapping is a test performed for medical devices that are sterilized by radiation, in order to map the distribution of dose within product. The objectives of dose mapping are to determine whether a product can be irradiated within the specified minimum and maximum dose range and to establish the parameters for routine processing.

This commentary provides a comparison of virtual and physical dose mapping. It highlights the advantages and use cases of each method, as well as discusses situations in which replacing physical dose mapping with virtual dose mapping would improve quality and efficiency of product development and validation.

Physical versus Virtual Dose Mapping

Physical dose mapping involves placing radiation sensors throughout the product and measuring dose using special laboratory equipment to generate direct experimental data on the dose received by the product. Producing results that are representative of routine processing requires careful planning and dosimeter placement to identify and test the minimum and maximum dose to product. The features and limitations of physical dose mapping are comprehensively documented in industry standards and publications.^1–4

Virtual dose mapping is performed by software that uses a digital model of the radiation source and product to calculate radiation dose throughout the product. Similar to physical dose mapping, the quality of virtual dose mapping relies critically on an accurate definition of the inputs (radiation source, treatment conditions, materials, product and packaging configuration, and load configuration, as applicable) to produce an accurate output.

This key characteristic of virtual dose mapping drives several notable advantages (and limitations) that could be considered when deciding between performing physical and virtual dose mapping for a particular study. The advantages and limitations of each method are summarized in Table 1.

Table 1.

Advantages and limitations of physical and virtual dose mapping.

Speed, Cost, and Quality Considerations

The most noteworthy difference between physical and virtual dose mapping is that the latter does not require physical product, dosimeters, or an irradiator. This results in the following important benefits.

Speed

Virtual dose mapping removes the time and cost of building test samples, quoting, shipping, and competing with production time on the irradiator schedule. With no need for physical product and packaging, it is possible to compare different product and packaging configurations to optimize packaging efficiency for loading the irradiator or dose distribution for radiation-sensitive products much earlier in the product development process.

Cost

With no physical product or dosimeters required, the dose mapping study can be performed with no destructive testing. Iterations of product design and configuration require less setup time and resources compared with physical dose mapping.⁵ Virtual dose mapping at early stages of product development can generate actionable data for planning later verification and validation tests (e.g., using knowledge of dose distribution to plan maximum dose testing for a product to ensure that it can be irradiated within its specified dose range).

Facilitating Transfer among Irradiators

Virtual dose mapping requires extensive definition and validation of irradiator source characteristics (e.g., size, geometry, energy level, distance from the product) to perform calculations. After these characteristics are established, however, the same set of operating conditions could be used to model dose distribution in a wide variety of products. Virtual dose mapping could be used to establish equivalence in operating conditions between two irradiators and accelerate migration of a large variety of products for business continuity or other reasons following transfer guidance in AAMI TIR104:2022.⁶

Quality

In some situations, virtual dose mapping can produce a higher-quality result compared with physical dose mapping by eliminating the sampling variables present in physical dose mapping. Virtual dose mapping also does not involve disassembly and assembly of product for placing dosimeters. Further, the virtual dose map considers all surfaces throughout a product, whereas physical dose mapping can only measure a small number of discrete points at which dosimeters are placed, with the skill and bias of the dosimetrist potentially affecting the process. Physical dose mapping relies on expert knowledge of the dose map creator to identify and place dosimeters in likely low- and high-dose positions, considering radiation physics and the variety of materials and design features of the product.

When to Consider Virtual Dose Mapping

To use virtual dose mapping tools as a replacement for physical dose mapping, it is necessary to define more specifically the situations in which virtual dose mapping tools will produce an equivalent or higher-quality result than physical dose mapping (i.e., be more appropriate from a risk-based perspective).

Statistical process control for irradiation processing typically is established by defining σ_process as a combination of the variability associated with the radiation source/conveyor (σ_mach), product configuration encountered during dose mapping (σ_map), dosimetry system calibration (σ_cal), and dosimeter measurement during dose mapping and routine process monitoring (σ_rep), if used. Virtual dose mapping could be used as a higher-quality replacement for physical dose mapping if the combined uncertainty of the dose calculated from a virtual dose mapping exercise is lower than the combined uncertainty of dose measured during a physical dose map.

Of the components of σ_process (σ_mach, σ_map, σ_cal, and σ_rep), σ_mach should be established during configuration of the virtual dose mapping tool to define the energy spectrum of the radiation source and any variation associated with conveyance, if used, through the radiation field. The definition of σ_mach can be determined from irradiator operational qualification data or allowed variation of critical process parameters.⁴ Physical dose mapping could consider σ_mach to be included in σ_map, as the product and dosimeters included in the study are subject to machine variation; however, any virtual modeling of dose distribution should be careful to specify σ_mach to avoid underestimating process variation.

The components of σ_cal or σ_rep do not exist in virtual dose mapping, as the determination of dose is based on mathematical modeling rather than measurement by dosimeters. With the remaining components of uncertainty, the comparison of virtual with physical dose mapping can be written as shown in Table 2. Understanding differences in σ_map for virtual and physical dose mapping is critical. Dose map variability σ_map includes variation from product load variability, dosimeter measurement reproducibility, and dosimeter placement variability; these features have important differences between physical and virtual dose mapping. (Table 3).

Table 2.

Condition in which virtual dose mapping produces an equivalent or higher-quality result compared with physical dose mapping. *σ_mach could be included as part of σ_map in a physical dose mapping exercise.

Table 3.

Factors that influence σ_map in virtual and physical dose mapping.

Product load variability is driven by product- and process-related variables. Product-related variables include the design, materials, and manufacturing process leading to differences in the arrangement of individual product units. Process-related variables include process conditions, such as the quantity and arrangement of product in the form presented to the irradiation source. Variation from process conditions could also include rearrangement of product between exposures if this is applicable to the product and irradiation process and could cause a change in product configuration (e.g., flipping a box vertically between passes through a gamma cell or electron beam irradiator).

Dosimeter positioning variability in physical dose mapping consists of placing dosimeters on or within product features and can be challenging for device features such as curved surfaces or features that are small relative to the size of the dosimeter. Dosimeter positioning variability can be further challenged by radiation sources with lower penetration of the product, considering the manner in which the product is presented to the irradiation source.

The presence of dosimeters altering the measurement of absorbed dose to product describes a phenomenon in which the dosimeter (and any tape or other material used to attach the dosimeter to the product) absorbs energy that otherwise would be deposited within the product or other dosimeters if unobstructed. This is unlikely to be an important variable in gamma or X-ray irradiation, as the penetrating ability of high-energy photons is not a challenge for most dosimetry. However, substantial measurement error can occur under certain circumstances in electron beam, where low-density product is more common, if the presence of dosimeters substantially alters the quantity or distribution of mass within the product being mapped.

By eliminating the dosimeter measurement reproducibility (σ_rep) and dosimeter placement variability components of σ_map, virtual dose mapping could produce a substantially more accurate determination of dose distribution compared with physical dose mapping, particularly for certain product configurations in electron beam in which the limitations of physical dosimetry methods introduce considerable measurement error. Gamma and X-ray irradiation could also benefit from virtual dose mapping for products in which the design features prevent measurement by dosimetry, such as internal features that are inaccessible without altering the device.

One potential weakness of virtual dose mapping is the challenge with assessing product load variability. For products with substantial variation in their arrangement within the package or that are capable of rearrangement during the irradiation process, physical dose mapping has the advantage of measuring the product “as is” without relying on prediction of the most likely arrangement of product. This could include products with poorly constrained packaging configurations, such as boxes of bulk items or certain products that are low density on average but contain a variety of materials of different density for which individual units could overlap. The risk of product reorientation is increased for products that are vertically flipped as part of the irradiation process without being constrained within the package.

As described in Tables 3 and 4, the ease of use and accuracy of virtual dose mapping tools depend critically on the ability to make representative models of the product and processing conditions, within an acceptable uncertainty budget. With product configurations that are difficult to model accurately, risk of the virtual dose map producing unrealistic results is increased. With greater definition and control of the product configuration, virtual dose mapping could determine the dose distribution with greater accuracy and lower uncertainty of σ_map compared with physical dose mapping with dosimeters.

Table 4.

Example product features related to σ_map likely to be more accurately measured in virtual versus physical dose mapping.

Conclusion

For many products and purposes, using virtual dose mapping tools earlier in the product development process or for change control provides substantial advantages. These benefits include faster product design iteration, more efficient packaging configurations, and lowering risk of unexpected failure to meet dose requirements after committing to the product and packaging design. In terms of process capability and quality, virtual dose mapping could predict dose distribution with greater accuracy or lower uncertainty compared with physical dose mapping in certain situations by eliminating uncertainty from measurement error and the dosimetry system.

Resources for Virtual Dose Mapping.

In support of virtual dose mapping technology, the ASTM Committee E61 on Radiation Processing published ASTM E2232-21, Standard Guide for Selection and Use of Mathematical Methods for Calculating Absorbed Dose in Radiation Processing Applications.⁷ For users considering development of modeling tools, ASTM 2232-21 includes a list of various mathematical modeling methods and source code.

In addition, the Food and Drug Administration currently is updating its guidance for General Principles of Software Validation. The agency's new draft guidance is titled Computer Software Assurance for Production and Quality System Software.⁸