Researchers at the Royal Veterinary College have found that machine-learning algorithms have the potential to improve the diagnosis of Cushing’s syndrome in dogs. 

Historically, diagnosis has been achieved with multiple blood tests. However, this process can be time-consuming, expensive and in some cases does not provide accurate results.

In an attempt to address these diagnostic challenges, the researchers, with funding from Dechra, assessed anonymised data from 939 dogs which had been tested for Cushing’s syndrome from the VetCompass population of 905,554 dogs and 886 veterinary practices across the UK1.

Using structured clinical data to look at the dogs’ demographics, clinical signs at presentation and laboratory results, machine-learning algorithms were applied to predict a future diagnosis of Cushing’s syndrome. Dogs suspected of having Cushing’s syndrome were included in the analysis and classified based on their final reported diagnosis within their clinical records.

The researchers say that the findings indicate that machine-learning aided diagnosis could predict the diagnosis of a practising veterinary surgeon and that using machine-learning methods in clinical practice could contribute to improved diagnosis of Cushing’s syndrome in dogs.

Additionally, further development of these algorithms could lead to earlier, more reliable and cost-effective diagnoses and therefore, better clinical care for dogs with Cushing’s syndrome. This could also create opportunities for this technology to be applied to other clinical problems. 

Imogen Schofield, lead author and PhD student at the RVC, said: “Machine-learning algorithms, like those used in this study, are already widely integrated in our day-to-day lives to help make certain decisions, such as Google or Netflix recommendations. Now this technology can be harnessed to help improve diagnostics in veterinary practice.

"By embracing the use of machine-learning methods, we are a step closer to providing vets in primary-care practice with an easy to use, low cost and accurate test that can support the often frustrating process of diagnosing Cushing’s syndrome in dogs."

Greg Williams, Senior Business Manager at Dechra Ltd. and Industrial Supervisor of the PhD studentship, said: “By funding Imogen's PhD and working with the RVC we have been able to develop validated clinical scoring and quality-of-life assessments to help vets deliver effective control and management of Cushing's syndrome in dogs.


  1. Schofield, I., Brodbelt, D.C., Kennedy, N. et al. Machine-learning based prediction of Cushing’s syndrome in dogs attending UK primary-care veterinary practice. Sci Rep 11, 9035 (2021). 

Whilst you're here, take a moment to see our latest job opportunities for vets.