Interview with veterinarian Dr Bläcker about the AI-supported diagnostic software PicoxIA

The topic of artificial intelligence has long been on everyone's lips in veterinary medicine too. However, veterinary-specific AI tools have not yet been implemented in everyday practice. We spoke to Dr. Dominik Bläcker from the Hamm veterinary practice, who has already successfully established AI-supported software in his practice for a few years.
Since autumn 2022, Dr. Dominik Bläcker has also been using our PicoxIA diagnostic software, which evaluates veterinary X-ray images using AI. "Generally, I take a very positive view of the use of AI-supported software. I think that this type of diagnosis helps to avoid sources of error, check your own diagnosis and support young colleagues in finding a diagnosis. As we also successfully use AI-supported analytics in our in-house laboratory, we see the expansion of AI-supported treatment tools as an opportunity to improve our treatment options and simplify our work processes."

We asked how Dr Bläcker implemented the software and integrated it into his workflow. "The software installation went without any problems thanks to the transfer via remote maintenance. The image transfer takes place in the EXAMION X-AQS software itself. By clicking on the X-AI button in the X-AQS menu under "Tools", the PicoxIA software analyses the X-ray image. The workflow is uncomplicated and fast.
The software analyses the images within a few seconds. The results are displayed on the right-hand side of the screen. The findings of the X-ray image are displayed in the upper white box, which in my opinion should be much larger, as the raedability is not optimal. The diagnosis(es) are listed on the right-hand side in an easy-to-find and legible manner and highlighted in colour according to the probability of the diagnosis."
By the "upper white box" on the user interface, Mr Bläcker means the preview of the findings report. This report can be exported and edited in a writing programme. The report created by PicoxIA contains the stored contact details and the logo of the practice or clinic as well as the corresponding patient information. The exported data is clearly summarised in the sections radiographic findings, conclusion, recommendations and X-ray images. The findings report can be customised, supplemented and further processed at any time.

We were interested to know for which cases Dr Bläcker uses the PicoxIA software."I like to use PicoxIA as a self-monitoring tool, i.e. I look at the relevant X-ray image and make a diagnosis. Then I check my diagnosis with the help of the PicoxIA software. The PicoxIA software is most frequently used for chest X-rays. We are a team of very experienced colleagues, so we tend not to experience any major surprises when interpreting X-ray images. However, I find the reliability and accuracy of the software astonishing. We do occasionally have seriously incorrect results in the HD diagnosis, but we can recognise these very easily by analysing the X-ray image. Overall, however, the software is very reliable and precise and, in my opinion, hardly contains any errors."
The HD diagnosis mentioned by Dr. Bläcker happened with a Norberg-Olsson angle calculation automated by PicoxIA. In order to avoid such incorrect findings, the measuring points defined by the software should be checked and adjusted manually if necessary.

PicoxIA is currently only intended for analysing cat and dog X-rays. Abdominal, thorax and hip images can be analysed. The software already recognises 34 different lesions in these regions."We would also like to be able to use PicoxIA for other regions (e.g. elbows)."
We agree, which is why we are already working hard to expand the scope of the software. Further lesions are to be added and an expansion to other animal species is also planned.

Of course, we also wanted to know whether Dr. Bläcker would recommend the software to his veterinary colleagues and who he thought would benefit the most from PicoxIA."The biggest advantage of using the AI-supported software is certainly the minimisation of incorrect findings and self-monitoring. For young colleagues in particular, the tool is therefore enormously helpful in recognising structures and quickly learn how to read X-ray images successfully. In the training of young colleagues, the AI makes training easier as many things that we would have to explain at great expense are now provided by an AI. I would recommend the PicoxIA software to any colleague without reservation."

We are delighted with this conclusion. We would like to thank Dr. Bläcker for his time and are sure that we will soon be able to fulfil the request for an extension of the regions and lesions covered.