Today, more than 79 million Americans are living with Human Papilloma Virus (HPV), the leading cause of cervical cancer and the second most common type of cancer for women worldwide (1). What makes this grave problem even worse is that most women are not aware that they are infected. The mortality rate due to cervical cancer, however, has steadily been reduced by 50% since the 1950s (2). This is because the increased use of screening tests allowed people to detect cancer earlier. Cancer found and treated at an early stage has a much higher chance of being treated successfully (3). Thus, early diagnosis often serves as the key factor to saving lives. Many ailments and diseases can be cured by simply taking the first step of defense against alarming signals by getting routine checkups.
However, many factors - including but not limited to varying socioeconomic situations such as age, income, and occupation - make early detection virtually impossible for many people. The primary reason may be due to the high cost associated with accessing health care services. Every medical service provided to a patient, ranging from a simple prescription to major surgery, requires a service fee. The cost of health care is markedly more expensive in the United States, compared to that of most other countries. According to a report by Milliman, the global consulting firm, in 2011, the annual healthcare cost for an average family of four in the United States was a staggering $19,393 (4). Furthermore, the number of times an average American receives regular checkups is four times a year, which is less than one-third of that of an average Japanese, which is 15 times a year (5). While social and cultural factors play a role in creating the disparity of figures in the two countries, one cannot deny the paramount significance of inaccessibility to medical care.
Another reason for low detection is the constraint of time. In South Korea, for example, more than 80% of all major and local hospitals operate only during 9 AM to 6 PM (6). This rigid operation schedule leaves no time for people with regular “9-to-6” jobs to make necessary visits to local doctors or hospitals. Highlighting the severity of this problem in the United States as well, are recent studies conducted by David Houle and Jonathan Fleece, authors of The New Health Age: The Future of Health Care in America, who revealed that the average time patients have to wait for hospital emergency rooms is approximately four hours (7).
Wrongful deaths and other injuries directly stemming from grave human medical errors and malpractice also serve as serious issues facing the health care system. The Journal of the American Medical Association reported that almost 100,000 people die annually from medical errors in hospitals in the United States alone (7). This figure essentially means that one out of every 370 people admitted to a hospital die due to inaccurate clinical decisions.
In light of this background, one solution is a comprehensive, yet portably accessible control panel. A bioinformatics system called Artificial Intelligence Personal Operating Panel (AI-POP) could play a pivotal role in preventing the aforementioned factors from adversely affecting health by using artificial intelligence and vital signs to analyze and predict patients’ health.
Vital signs can identify the existence of acute medical problems, are a means of rapidly quantifying the magnitude of an illness and how well the body is coping with the resultant physiological stress, and are a marker of chronic disease states. Vital signs are, by definition, important biological indications that display the status of a body’s life-sustaining functions. They include body temperature, blood pressure, heart rate, and respiratory rate, often noted as BT, BP, HR, and RR. These measurements have been useful in detecting, assessing, and monitoring general health by giving clues to possible diseases and showing progress towards recovery (8).
The composition and analysis of vital signs can serve as a rough but meaningful snapshot of one’s state of health. Fever, hyperthermia, heatstroke, and hypothermia are just a few of the diseases that a simple body temperature change can indicate. Fever, for example, is an upward shift in the body’s baseline temperature to generally 38°C or above. Similarly, hyperthermia is a result of an overload of the body’s thermoregulatory mechanisms. If a patient’s body temperature rises to 40°C or above, it would most likely mean that the patient is suffering from a heatstroke—a dangerous condition caused by prolonged exposure to heat. In contrast to the temperature elevations, the indication for hypothermia would be a drop in body temperature to 36°C or below (9). Every component of a vital sign contains vast information about the state of a patient’s health. Thus, AI-POP, which uses body temperature and three other vital signs, would give accurate information about the patient.
Then, an important question inevitably arises: just how can AI-POP actually distinguish between fever, hyperthermia, and heatstroke after the vital signs are measured and the temperature elevation is detected? This is where the Artificial Intelligence (AI) part of AI-POP comes into the scenario. First, it is essential to understand the basics of AI and the method of AI used in the system: Artificial Neural Network (ANN). Although there is no single definition of AI, which was first introduced at a conference at Dartmouth College in 1956, it is broadly and consensually characterized as “machines that can learn” (10).
ANN is a mathematical model that consists of a connected network of artificial neurons, known as “perceptrons,” which are organized in layers to simulate biological neural networks. There are three different types of layers in an ANN. The first layer is the input layer and the last layer is the output layer. Between the input and output layers, there may be an additional layer called the hidden layer. By adjusting the values of the connections between perceptrons, or weights, ANN can learn and generalize data (11).
The applications of ANN to medicine appeared in the early 1990s and have rapidly gained popularity ever since (12). ANNs have been applied to malaria, cardiology, orthopedics, low back pain, cancer, Alzheimer’s, pulmonary diseases, and heat stress (11). In order to diagnose a disease by using AI-POP, a patient’s vital signs would first be entered into the ANN as input nodes. The outputs would be the names of the possible diseases and the suggestions for treatments the patient could take.
The ANN in AI-POP offers accurate medical diagnosis through deep learning, and bases its learning and categorizing from a licet medical cloud database. In other words, the “everywhere, anytime” would serve as the key point in the sharing of vital information from patient to patient, patient to medical professional, and finally, from medical professional to medical institution for further diagnosis and treatments. The availability of AI-POP in every household could provide another tangible benefit of health care: accessibility unparalleled by any other system in modern society.
The idea of diagnosing diseases with only four components may still seem like science fiction. However, our world today already has all the technology needed to turn the abstract idea of AI-POP into a concrete reality. Take one of the medical apps for Apple Watch called AirStrip, for example. During the ‘Apple Special Event’ on September 9th, 2015, AirStrip co-founder Dr. Cameron Powell said that AirStrip for Apple Watch will be a “game changer for healthcare” (13). As long as both doctors and patients have the Apple Watch app, doctors are able to check their patients’ vital signs, get information about their patients’ diagnoses, and view their appointment schedules on their wrists from anywhere in real time (14). In addition to AirStrip, “Pepper”, an Artificial Intelligence Robot conceived by SoftBank, a Japanese multinational conglomerate whose core business is based on telecommunications and the Internet, exemplifies the advancement of AI in our world today. Introduced in 2014, Pepper uses AI platform to form an emotional attachment with its human owners. With AI robots already capable of distinguishing human emotions, SoftBank CEO Masayoshi Son predicts that “Artificial intelligence will overtake human beings not just in terms of knowledge, but in terms of intelligence. That will happen this century” (15).
Buckminster Fuller’s “Knowledge Doubling Curve” shows that human knowledge had doubled every century until 1900 and every 25 years by the end of 1945. Today, different types of knowledge on average double every 13 months. In the future, International Business Machines Corporation (IBM) stated that “internet of things” will cause knowledge to double every 12 hours. One of the types of knowledge today, the clinical knowledge, is doubling every 18 months and is continuing to grow exponentially (16). This is problematic for human medical professionals, because the average human brain only has about 86 billion neurons and is limited to store a certain amount of information. This means that medical professionals will not be able to be up to date for the increasing new clinical knowledge of patients, potentially leading to more medical errors. However, an AI can hold an infinite number of neurons, or perceptrons, in its wirelessly connected network and be able to manage all of the stored information.
It may be tempting to assume that our future will be governed by the impending vast presence of artificial intelligence, but this cannot be further from the truth. In fact, thanks to the breakthroughs in artificial intelligence, humans should be in a better position to dictate and control our own bodies more than ever. Specifically, the integrated approach of conjoining various elements in the fields of bioinformatics, computer science, design, and biomedical engineering, and, of course, artificial intelligence has ushered in a new era. The beauty of bioinformatics and the vital information derived from a combination of vital signs is not the automated system, as great as it is. But it offers the real-time and comprehensive snapshot of the totality of the “big picture”, which the AI-POP will provide, that should serve as the cornerstone of human decision making. With a greater amount of information and more refined data, we can reach more accurate diagnoses at our most convenient time and place, the unparalleled advancement from the current position in medical health care.
To say that our future has limitless possibilities is a gross understatement as well as an irresponsible overstatement, given that the unparalleled advancement in technology expects humanity to make difficult decisions in the convoluted piles of data. Nevertheless, it has always been our job to sort through the overabundance of data to reach the most suitable consensus for humanity.