The FDA Invests in Clinical Trial Quality and Efficiency

See photo explanation below.

See photo explanation below.

I reacted with excitement to the FDA’s recent announcement that $38 million had been granted to the Clinical Trials Transformation Initiative (CTTI) to increase the quality and efficiency of clinical trials.  CTTI is based out of Duke University and includes nearly every major government healthcare agency, pharmaceutical company and trade group as members, as well as several medical device manufacturers and academic medical centers.

It’s encouraging to see that the FDA is investing in initiatives that can be paradigm changing in clinical trials. I believe that software-based image analytics can be a key component in this strategy. By leveraging the power of software analytics in clinical trials, where the protocols are well-defined and carefully executed, software-based quantitative measurements of the images provide a cost-effective and objective enhancement to traditional medical image observers. We believe that this initiative is a signal that objective, quantitative analytics in clinical research will become an industry standard in FDA data acceptance. Image analytics techniques may be among the most advanced to deliver that data.

ImageIQ has seen a rise in acceptance of consistent, objective, quantifiable data in the clinical trial sector via software-based image processing and analysis, as well as its effectiveness in reducing costs and accelerating research timelines. For example, in orthopedic spine flexion-extension assessments, we’ve seen variability in measurements of simple angles of up to +/-10%. That level of subjectivity and variability in the measurements resulted in a larger cohort size than necessary to power the study, which resulted in increased costs and trial length. That situation, in the worst case scenario, could lead to less accurate data and weaker product efficacy claims, which undoubtedly causes problem in FDA clearances and eventually post-market acceptance.

There are groups planning their clinical trials that don’t even realize that imaging and software-based image processing and analysis can be a powerful alternative to supporting and strengthening their product safety and efficacy claims. We are very pleased that the FDA has recognized the importance of alternatives to traditional methods that will provide objective, quantifiable and consistent data, sending a clear message to medical device and pharmaceutical companies that the bar has been raised when it comes to clinical trial data.

P.S. Photo description for our science fans:
Osseointegration of Resorbable ACL Screws
Project Summary: Bioresorbable interference screws are theorized to fill in with bone tissue over time as they dissolve, but still provide the same initial anchoring support for the ACL as traditional metal screws. Potentially, such an implant could delay/eliminate revisions related to loosening that is observed with metallic implants. To evaluate efficacy of a resorbable ACL screw, computed tomography (CT) scans were performed and subsequently analyzed to determine the rate of screw resorption, graft remodeling, and new bone formation in the tibial tunnel. In a cohort of 16 patients with multi-timepoint CT scans (post-op, 6 months, 1 year, 2 years), baseline volumes for each patient were reoriented such that the tibial tunnel was aligned perpendicular to the transverse axis. Subsequent timepoints were spatially co-registered to the baseline volume and the tibial tunnel was splined for each timepoint. Using custom tailored segmentation software, the screw, bone graft, and new bone was segmented for each tunnel and pseudocolored to provide a visual assessment of each component over time. Extracted quantitative measures included tunnel volume/diameter, graft volume/density, trabecular bone volume/density, and screw volume/density. For assessment of algorithm accuracy unimplanted screws were imaged on both clinical and micro-CT (high resolution standard) scanners (600 um vs 20 um resolution) and segmented using algorithms described above. Despite their drastically different resolutions there was only a 1% difference in analyzed volume between the clinical and preclinical scanners. Additionally, metal screws present in the femur of a number of patients were analyzed across multiple timepoints using the segmentation routines above. Even though voxel dimensions and scanning protocols were not consistently maintained across all timepoints, there was only a 4% variation in analyzed volume.

P.S.S.
There are hundreds of similar images in the Capabilities section of our website. Browse around.

{ 0 comments… add one }

Leave a Comment