Ophthalmology Imaging Software Algorithms Improved in Just 6 Weeks

Summary

In just six weeks, ScienceSoft helped the Customer improve its biomechanical imaging software for eye tissue elasticity diagnostics. We fixed defects and improved report generation functionality and algorithms for device findings visualization.

About the Customer

The Customer is a US-based ophthalmological medical device company. Its flagship product enables biomechanical imaging of eye structures (i.e., cornea, retina).

The CE-marked product provides in vivo diagnostics capabilities and empowers ophthalmologists and surgeons to make informed eye treatment decisions based on biomechanical tissue properties. The efficiency of revolutionary technology is proven by scientific research compliant with 42 CFR Part 50 and 45 CFR Part 94.

The product is accompanied by desktop medical software for device findings analysis and visualization. To upgrade the software functionality within short timeframes, the Customer sought an ISO 13485-certified technology partner.

Rapid Team Augmentation with a Java Developer

After studying the Customer’s healthcare software development needs, ScienceSoft assigned a Java engineer to put the Customer’s vision into action. Throughout the project, ScienceSoft’s talent reported to the Customer’s software engineer and communicated with him daily to implement the tasks set by the Customer’s team of scientists.

Improving the Algorithms for Device Findings Visualization

The Customer’s product uses light pulses to measure reverberations through a depth of eye tissue. To determine tissue elasticity, the software performs a spectral analysis of scattered laser light in several reference points and measures changes in wavelength using the Brillouin principle. The software uses the linear interpolation method to calculate the elasticity measurements between the reference points. The results are presented as a heat map, with the stiffer tissue areas appearing as cool colors and the flexible areas appearing as warmer colors.

The Customer wanted to make the distinction between the areas of different elasticity more discernible so that eye care professionals would spend less time studying the device findings. ScienceSoft’s Java engineer was tasked to research the potential ways to improve the algorithms.

After eliminating the methods that didn’t suit the Customer’s needs (e.g., those used for extrapolation instead of interpolation), he came up with a plan to adjust the scale for measurements-to-colors conversion (i.e., split the measurements into smaller ranges). The Customer was satisfied with this approach and decided to implement it during the subsequent software updates.

Imaging Software Improvements Delivered in Just Six Weeks

Besides the algorithm improvement research, ScienceSoft’s Java engineer fixed the software bugs identified by the Customer’s team and improved the findings report generation functionality.

Initially, only the primary device findings were exported to a PDF report generated by the software. Following the Customer’s requirements, our engineer extended the number of exported report parameters to be fully identical to the initial output displayed by the desktop software. Thanks to these improvements, eye care professionals will have access to sufficient diagnostic information no matter how they access the biomechanical imaging software findings.

Key Value for the Customer

• Rapid team augmentation with a Java developer to meet strict project deadlines.
• Medical device software improvements implemented in just six weeks.
• Full remediation of the discovered medical device software defects.
• Optimal technical design for improving the device findings analysis and visualization algorithms.

Technologies

Java.