Pilot Project Summary:
Data is being generated at an exponential rate, and while this data is useful, it could be more structured. In a clinical setting, developing new treatments and providing physicians with an overview of all the information available on a patient is challenging. To extract information automatically and avoid manual data input, Melax Tech has used natural language processing (NLP) to extract structured data from unstructured patient reports.
The Roche team, led by Stefanie Kaufmann, the lead data scientist, worked with Melax Tech, led by Dr. Xiaoyan Wang, on a pilot project to extract structured data from unstructured patient reports for the last three months. Clinical notes are rich in pathological information, which can be used to guide decision-making, but retrieving this information accurately and at scale is challenging. Melax Tech's solution was to use NLP, starting with annotating pathological notes. The result was extracting meaningful and normalized information encoded in standard terminologies.
Melax Tech has proven to be a capable solution for extracting structured data from unstructured patient reports in a clinical setting. Using NLP, they have solved the challenge of retrieving rich pathological information with accuracy and scale. The pilot project with Roche is a testament to the success of Melax Tech's technology and expertise in biomedical NLP.
To learn more about Startup Creasephere Programs, visit their website here.