ABSTRACT

In this experiment 3 operators at a large pharmaceutical company were tasked with capturing measurements of subcutaneous tumours on mice with both callipers and BioVolume on the same day, for the duration of the study protocol. 

This was then repeated over 2 further experiments (3 experiments in total). The primary goal was to compare the inter operator variability of the 2 techniques to see if BioVolume outperformed callipers in the repeatability of measurements between operators. 

A secondary goal of the experiment was to see if collecting image data (thermal,  3D and RGB photographic image sets) using Biovolume provided additional traceability benefits to the operators.  

METHOD

3 Operators anonymised with labels 101, 102 and 103, were chosen at random to take measurements of single subcutaneous tumours implanted on the flank of 40 C57 Black/6 mice. 

The experiment was conducted on 2 further studies containing 64 and 32 mice respectively, this time on Balb/C mice. ​

​

2 different techniques were used for measurement during this standard tumour growth study (with no compound or dosing of the animals).

RESULTS

GROWTH CURVE ALIGNMENT

After all of the data was collected, growth curves were plotted for both BioVolume and calliper tumour volume measurements obtained by the 3 operators.

The results can be seen in Fig. 1, and visually demonstrates reduced variability between users during the initial study, when utilising BioVolume.

VARIABILITY ANALYSIS

Intra-class correlation analysis is used to assess the variance across operators whilst accounting for other sources of variation. High ICC corresponds to high agreement
(1 is perfect agreement). 

​

For the 3 experiments, there is a statistically significant difference between the 2 techniques. The BioVolume ICC score is in the ‘excellent’ group, while the result for callipers falls into the ‘good/moderate’ group.

BioVolume ICC: 0.96 (0.95- 0.96)
Category: Excellent 

Callipers: 0.86 (0.73 - 0.91)
Category: Good/ Moderate

DATA TRACEABILITY & VALIDATION

BioVolume records 3D models, RGB and thermal imagery for each rodent measurement, allowing interrogation at an individual rodent level. Tumour progression can be tracked visually, providing confidence in measurement output.

CONCLUSION

ICC and CoV analysis show statistically significant improvement in inter-operator variability across all 3 users when utilising BioVolume within the three individual experiments. This allows us to conclude that BioVolume can offer an improvement in study repeatability, as well as additional associated benefits.

​

The secondary goal of determining whether the collection of image data provided the user with additional benefits by way of allowing interrogation of raw data was also found to be true. BioVolume’s thermal, 3D and RGB image sets provided traceable visualisations of physical tumour growth over time, resulting in greater user confidence of the study outcome.