2019 appears magnificent for Healthcare AI researchers. Various studies held by organizations such as Accenture, Google’s Deep Mind AI, PwC, IBM, and others observe that artificial intelligence and machine learning are well poised to streamline the work of healthcare providers, reshape delivery of personalized healthcare services, facilitate better care for patients, and significantly reduce the time to look for information in the coming year.

As 2018 draws to a close, here are two leading AI researchers who have come out with a couple of innovative healthcare AI applications. What have they discovered? Let’s check out.

IBM’s Fingernail Sensor

In its new research published in Scientific Reports, IBM Watson Research Team details a first-of-a-kind “fingernail sensor” prototype to help monitor human health. This wireless, wearable device continuously measures how a person’s fingernail bends and moves, which is a key indicator of grip strength. Grip strength is widely used as a metric to measure several health-related components such as the effectiveness of medication in individuals affected with Parkinson’s disease, the degree of cognitive function in schizophrenics, the state of an individual’s cardiovascular health, and all-cause mortality in geriatrics.

IBM’s new fingernail sensor consists of strain gauges attached to the fingernail and a small computer that samples strain values, collects accelerometer data and communicates with a smartwatch. The smartwatch runs machine learning models to rate bradykinesia, tremor, and dyskinesia which are all symptoms of Parkinson’s disease. The sensor uses signals from the fingernail bends such as the tactile sensing of pressure, temperature, surface textures etc.

IBM’s Watson Research Team remarks that “by pushing computation to the end of our fingers we’ve found a new use for our nails by detecting and characterizing their subtle movements. With the fingernail sensor, it is possible to derive health state insights and enable a new type of user interface”.

MIT’s AI System to Detect Brain Haemorrhages

Another significant contribution to healthcare AI comes from the prestigious Massachusetts Institute of Technology (MIT), whose research team has developed an artificial intelligence system to quickly diagnose and classify brain haemorrhages, and to provide the basis of its decisions from relatively smaller image datasets. MIT’s researchers feel that such a system could eventually become inevitable for hospital emergency departments evaluating patients with symptoms of a potentially life-threatening stroke, allowing rapid application of the correct treatment.

In their latest research published in Nature Biomedical Engineering, a team of researchers from MIT describes how the AI system was trained to detect brain haemorrhages. The team used 904 head CT scans to train the AI system, each one consisting of around 40 individual images labelled by a group of five neuroradiologists, as to whether they depicted any of the five haemorrhage subtypes or not, based on the location within the brain. To improve the accuracy of this deep learning system, the research team incorporated steps, imitating the way radiologists analyze the images. The team tested the model system on two separate sets of CT scans: (1) a retrospective set taken prior to developing the system, and (2) a prospective set taken post the model was created. The AI model system was accurate in detecting and classifying intracranial haemorrhages, and proved to be even better than non-expert human readers, while analyzing both the retrospective and prospective sets.

“The next step will be to deploy the system into clinical areas and further validate its performance with many more cases,” says Shahein Tajmir, Radiology Resident at Massachusetts General Hospital.

Artificial Intelligence initiatives like the above point out that 2019 will be a reality check for the hyped technology in healthcare. Across clinical and non-clinical use cases, AI’s growth in healthcare space will be strong and pervasive.

Related read: Integration Challenges in the IoMT Sector.