Terra Quantum has developed a groundbreaking hybrid quantum neural network (HQNN) model that outperforms traditional methods in identifying healthy livers suitable for transplantation. Collaborating with medical experts, the company designed a system combining quantum computing with classical machine learning, achieving a 97% image classification accuracy for diagnosing non-alcoholic fatty liver disease (NAFLD). This represents a 1.8% improvement over existing techniques.
Published in the journal Diagnostics, the study, titled Hybrid Quantum Image Classification and Federated Learning for Hepatic Steatosis Diagnosis, highlights the potential of HQNNs to provide precise diagnoses while protecting patient privacy. The model leverages federated learning, enabling data sharing among hospitals without compromising patient information, aligning with stringent privacy regulations.
This advance could revolutionize liver transplantation by improving donor liver selection and potentially increasing transplant rates. The Terra Quantum algorithm not only enhances diagnostic accuracy but also reduces false positives, ensuring better outcomes for patients.
Additional information can be found in a press release which can be accessed here.
August 13, 2024