Unmet Need:
Ultrasound imaging is often performed at the lungs as a point of care test for various scenarios, such as analyzing traumatic injury, diagnosing lung pathologies, and monitoring disease progression, or as a guidance technique for certain interventions. Such imaging requires a highly skilled operator in order to generate precise and accurate scans, especially when imaging complex or deep organs, which can place a burden on clinicians and limit its availability to patients. Besides, bedside lung ultrasound involves close contact between clinicians and patients which would increase risk of exposure to infectious diseases, such as COVID-19. Thus, a method is urgently needed to reduce the need of clinical personnel to perform lung ultrasounds in order to ease clinical burden and meet the high demand for lung ultrasounds in the case of an epidemic.
Technology Overview
Researchers at Johns Hopkins University have developed a robot guided system which allows for the automated ultrasound scanning of patients, focused specifically on the lungs. The system can be operated with or without prior CT scans from patients, as it leverages deep learning methods and a force-displacement profile to precisely guide the ultrasound probe to optimal landmark positions for imaging. The system has been validated on full torso ultrasound phantoms, with 100% interpretability when using force-feedback with prior CT, and 87.5% with landmark estimation.
Stage of Development
Preliminary feasibility of system has been shown in phantoms.
Publication
Al-Zogbi et al., Front. Robot. AI, 25 May 2021. https://doi.org/10.3389/frobt.2021.645756
