While Microsoft does not need an introduction as a key player in the software industry, the giant has quietly been taking strides over many years to solidify its footprint in healthcare. This effort was rather close to home to one of Microsoft’s own employees – who was in the midst of a complicated pregnancy showing signs of a fetal abnormality in the form Posterior Urethral Valve (PUV) syndrome.
Affecting nearly 1 in 5000 to 1 in 8000 male babies per year, the condition creates blockages in the baby’s bladder and is the most common cause of chronic kidney disease (CKD) due to urinary tract obstruction in children. The condition is often caught during fetal ultrasounds, but can be missed without proper training - an inefficiency that could potentially lead to lifelong health issues, or premature death, for children effected with the condition.
As part of a longstanding collaborative relationship between InterKnowlogy and Microsoft, a partnership was arranged to build a model that could serve as a potential aide to doctors in their early detection efforts. Leveraging Microsoft’s Azure Cognitive Services Custom Vision platform, InterKnowology employed Computer Vision and Machine Learning methodologies to train a model simply using existing ultrasound images that showed signs of the condition.
The model was able to decisively differentiate PUV ultrasounds from those that did not show symptoms of the condition, and also provide a degree of certainty of its findings to guide doctors in their attempts to diagnose the condition. It demonstrated the technology’s potential to assist doctors with timely diagnosis and avoid the life altering complications often associated with late detection, while saving costs for healthcare providers.
The results were shared during the 2017 Microsoft Inspire Conference. You can read more about how it transpired on their website: “How one Microsoft mom inspired health care companies to embrace the life-saving potential of AI”.