Use Artificial Intelligence To Help Identify Abnormalities In Medical Images Use Artificial Intelligence To Help Identify Abnormalities In Medical Images
In all branches of medicine, there is an element of exposure to human error, and a report published in BMJ puts medical errors as the third leading cause of death in the US. Every year, radiologists have to comb through an increasing number of body scans, with some estimates suggesting an increase by over 20% per year since 2006. Radiologists not only have to analyze and evaluate a lot of information in a short period of time, but also operate under conditions of pressure and uncertainty. Wrong assessment of medical images increase healthcare costs, lower efficiency, and developing solutions that lower the risk of human errors could be the difference between life and death.
The New York-based startup is trying to make it easier for healthcare practitioners to accurately identify diseases from ordinary radiology image data. They are developing a medical software that uses deep learning technology to identify abnormalities in medical images, helping radiologists make better medical decisions. The technology perform highly complex pattern recognition and highlight abnormalities such as lung nodules, cerebral aneurysms, or evaluate suspicious structures. Over time, the technology become smarter and more robust by incorporating feedback from radiologists, and is able to give suggestions based on similar images previously evaluated. was founded by Per Wakahiu Njenga and Jeet Raut, after a physician misdiagnosed a lump on Raut’s mother’s breast as benign. Fortunately, his family sought out the opinion of a second doctor, who diagnosed the lump as malignant and treatment began immediately. Although everything worked out well in the end, the misdiagnosis led Jeet to think about ways to improve the process of evaluating medical scans and help radiologists.
The innovative technology is not trying to replace what a radiologist does, but is meant to make scanning images faster and more accurate. Today, depending on where you live, radiologists reading your scan can have highly varying levels of knowledge, which could have an impact on the process of accurately diagnosing illnesses. This also means that the technology could have a potential future in developing countries where trained personnel is limited. has a huge potential in more accurately diagnosing illnesses, ultimately lowering healthcare costs, and improving clinical outcomes. It helps radiologists make better medical decisions, increasing efficiency and reducing uncertainty. It lowers the risk of human error, providing more confidence in the results for both the patient and the healthcare provider.
The startup has been part of Gener8tor, a startup accelerator in Wisconsin that recently concluded its 12-week program. They expect to raise a seed round by the end of September, and already has funding commitments from multiple investors. The startup is also seeking to pilot its software at multiple hospitals during the next 12 months, and discussion have been held with large hospital systems based in the Midwest.