The study by Penn State University, which possibly came as a by-product of advanced techniques that facilitate early-stage diagnosis of infants with neuromuscular diseases, maybe the first study that came as the result of the use of advanced technology. The establishment of an engineering scientific mechanical professorship in memory of James L. Henderson, Jr., was a critical development led by the Larry Cheng group’s introduction of novel methodology based on wearables and machine learning, which has allowed an abnormally high accuracy in recognizing and categorizing at-risk newborns.
Innovative technology to combat neuromotor diseases
This recent test case, a study entitled ‘Advancement in Science,’ consists of a network of soft wireless IMUs in combination with a miniature machine-learning algorithm that can be used to monitor the spontaneous activation of infants’ motor functions and signs of their developmental disorders or progressions, like cerebral palsy. This artificial intelligence robotic system is designed to detect abnormal movement patterns that are indicative of cerebral palsy and autism spectrum disorder, which is the objective of this technology.
Dr. Cheng consciously disclosed that these movements are an indicator of a human’s general neural development from 0 to 20 weeks after birth. Conventional methods in which human experts usually make the diagnosis by visual assessment of symptoms are often impoverished by human error, extreme weather conditions, and the complexity of video setup. Regarding this sensor technology provided by Cheng, we can talk about two friendlier improvements pertaining to objectivity and practicality. Those improvements can reduce the time required for earlier and more accurate diagnoses.
Early detection of neuromotor issues with new sensor tech
Sensors have been trending in the fitness industry for some time, a tendency that will continue in the coming years, considering their capacity to measure activity levels in real-time. This data is incredibly valuable for users who want to monitor their progress and achieve their fitness goals.
This sensor system developed by Cheng’s team has mechanical properties similar to human skin, making it ideal for handling an infant’s delicate skin. The infant’s forehead, wrists, and ankles were chosen as the places where the first five IMUs were installed, making a network that is sparse but captures the whole scope of the infant’s movement without being intrusive. This raw data is then processed by the team’s machine-learning algorithm that utilizes a custom algorithm to identify movement patterns as either ‘Normal,’ ‘High Risk,’ or ‘Abnormal.’
Hence, it not only improves the accuracy of the diagnoses but also drastically reduces the cost required and the materials needed by contrasting the methods of diagnosis accustomed previously,” Cheng said. Small machine-learning algorithms are less useful in areas where resources are limited. The focus is on rapid results from the smaller algorithms, with the determination that elaborate AI frameworks are not necessary.
Advancing expert systems for enhanced healthcare outcomes
Pilot research of 23 babies, although promising, is not enough to confirm protean results as some degree of error element is involved in such studies. Therefore, large-scale studies have to be carried out for validation. Along with Cheng and other team members, working with medical professionals is among their goals, as well as shedding more light on the product and refining the technology. Developing this sensor network will facilitate research that is not only limited to neuromotor evaluations but embraces cardiopulmonary assessment, vocal cord vibration, and sports training, too.
The inclusion of artificial intelligence and wearable technology in medical diagnostics opens a new window of opportunities in the provision of care to young children, which may improve the establishment of early interventions and a child’s survival rate in the future for individuals at risk for neuromotor diseases. As the research continues to be carried out thoroughly, this technology may become the most necessary tool for physicians and parents to monitor the newborn on a preventative and developmental basis.
This breakthrough from Penn State also underlines the necessity and usefulness of collaboration between different fields, in this case, medicine and computer science. It also brings the shift to the fact that such applications are accessible nowadays. With the help of compact and efficient artificial intelligence-based systems combined with noninvasive wearable sensors, medical diagnostics, and patient care may evolve into a new way of approaching this issue, where technology will be used to make easy detection and treatment strategies.
This article originally appeared in News Medical
