• Dr. Timothy Smith

Detecting the Invisible in Medicine


Photo Source: Pixnio

Advances in computing and the avalanche of seemingly mundane data people produce merely by using their smartphones, tablets, and computers now can provide miraculous insights into human health. Humans continuously produce many signs, both physical and chemical, that reflect the state of their health. Doctors, friends, and parents can pick up these signs of ill health, but some signs of disease evade human observation. To help pick up the signs that humans can miss, we have different technological tools to help pick these signs.

Service dogs perform many essential tasks for people, including the remarkable ability to detect invisible signs of some human diseases. Trained service dogs can detect epileptic seizures up to forty-five minutes in advance. Such an advanced warning gives the epileptic person time to take a medication, call for help, or find a safer place to go. Although science has not yet determined what chemicals the dogs smell, research published in Nature Scientific Reports clearly shows that dog’s powerful sense of smell is at play. French researchers found that people with epilepsy produce an odor from their saliva before a seizure that dogs can pick up but humans can’t. (Nature) The canine sense of smell exceeds a human’s sense of smell by 100,000 times. In the same vein, dogs can detect other maladies with distinctive odors such as strokes, diabetes and cancer. Because dogs have millions of odor detection cells in their noses, a brain more developed to interpret those smells combined, and a profound ability to notice changes in humans, they possess a genuinely superhuman ability to monitor people’s health.

Another human companion, the computer, in the form of cell phones, wearables, and tablets, goes with us everywhere and continuously collects information about us. Elon Musk commented at the World AI Conference in 2019 that we have become so integrated with our phones and computers that we have already become cyborgs. Researchers and inventors have found ways to use the information we generate to better diagnose some types of disease. For example, professors Schwab and Karlen at the Institute of Robotics and Intelligent Systems in Zurich, Switzerland, collected four types of data from smartphones in an attempt to detect Parkinson’s Disease. Their work appeared in 2019 in a research article titled, “PhoneMD: Learning to Diagnose Parkinson’s Disease from Smartphone Data.” (ETH Zurich) Parkinson’s Disease (PD) is a chronic and progressive neurological illness marked by a loss of dopamine production in the brain leading to muscle tremors, loss of balance, and a shuffling gate, among other symptoms. Although no cure exists for Parkinson's Disease, patients benefit from early detection and medical intervention to help mitigate some of the symptoms. However, early detection proves difficult in the clinic, and the authors note that 25% of PD diagnoses are incorrect. Schwab and Karlen collected four types of data from the smartphone including gait, voice, screen tapping, and memory tests. Smartphone data collected from healthy people and patients with Parkinson's were then analyzed by artificial intelligence, and the results clearly showed a significant ability to detect the disease. Such monitoring over time could also notify people of the progressive emergence of Parkinson's Disease they show more extreme . In another study, a group of researchers at the University of Texas collected data on how people type on their computer keyboard and used it to detect Parkinson's Disease. James Moore writes in an article titled, “How You Type on a Keyboard Could be Used to Detect Disease,” that slight differences in typing cadence down to the millisecond observed over time informs a machine learning program of progressing Parkinson's Disease. (Patient Worthy) Moore goes on to write, “Typing analysis has also been used to detect Huntington’s disease, heart problems, and sleep disorders.” The basic action of typing provides an invisible pattern of neurological health that could only be observed by a machine that can see very subtle changes and trends.

People continuously display both visible and invisible signs of their well-being. Symptoms of disease such as fever, change in color, among many others may be easily detectable to the average person, but many symptoms may not be as obvious. For years, doctors have used various tests to determine the presence of a disease. Service dogs can detect through their fantastic sense of smell symptoms of illness that even science has not figured out. For example, service dogs can detect seizures before they happen and also identify some forms of cancer and diabetes. Remarkably, researchers have begun to harvest data that people produce in abundance in the daily use of their smartphones. Subtle changes in gait, voice, memory, typing, and screen tapping can help smart machines to detect the early onset of disease such as Parkinson’s and Huntington’s Disease earlier than a traditional diagnosis. Your smartphone, the data it produces, and artificial intelligence will add another kind of service dog that looks after your health and well-being.

Dr. Smith’s career in scientific and information research spans the areas of bioinformatics, artificial intelligence, toxicology, and chemistry. He has published a number of peer-reviewed scientific papers. He has worked over the past seventeen years developing advanced analytics, machine learning, and knowledge management tools to enable research and support high level decision making. Tim completed his Ph.D. in Toxicology at Cornell University and a Bachelor of Science in chemistry from the University of Washington.

You can buy his book on Amazon in paperback here and in kindle format here.