By Kasturi Datta |
Biotech organisation Pangaea Data has built an AI platform on Microsoft AI Azure and Microsoft Fabric that helps clinicians identify untreated and under-treated patients by analysing patient records in real time.
The AI platform uses the Azure service ‘Microsoft Text Analytics for Health’ to identify and categorise medical terms from unstructured patient records and uses Azure OpenAI for contextual reasoning to determine whether a clinical pathway needs to be taken.
The platform can surface identities of at-risk patients, giving clinicians the opportunity to improve health outcomes for patients who otherwise would have remained invisible.
“When we’re talking about finding untreated and undertreated patients based on existing information from patient records, we have to keep two things in mind,” said Vibhor Gupta, CEO and founder of Pangaea Data. “Number one, it’s got to mimic what a physician would do manually, and number two, we’ve got make it available at the point of care.”
The platform uses Power BI – the analytics and data visualisation platform in Microsoft Fabric – to present AI insights within familiar workflows, eliminating the need for clinicians to switch between applications.
“Microsoft Fabric connects data from different sources – clinical notes, lab data, vital signs, medication information, imaging – so that our systems can use it as a unified interface to capture and read all of this data together, rather than us having to connect with each individual interface ourselves,” said Jingquing Zhang, head of AI at Pangaea Data. “
This unified approach enables clinicians to carry out earlier assessments and intervention, reducing the long-term economic burden associated with patient re-admissions.
In a US health system study, Pangaea identified previously uncoded patients and demonstrated higher accuracy than traditional ICD code-based approaches, supporting earlier intervention and improved outcomes.
Gupta envisions the integration of Pangaea’s AI platform within healthcare facilities on a global scale, enabling every clinician to have the right intelligence at hand regardless of a patient’s identity. “The impact is going to resonate for the entire life of that patient and their family, and perhaps for generations to come,” said Gupta.
Read more: Pangaea Data uses Azure AI to identify untreated and under-treated patients.