Medical imaging is used by medical specialists to create images and sets of data about humans and animals. This information is then analyzed in the hopes of preventing, diagnosing, or examining diseases. The objective of this article is to give an overview of medical imaging to readers who may not be familiar with the process.
As a field, medical imaging incorporates many disciplines, such as medicine, radiology, engineering, science, mathematics, and statistics. In order to discover innovative new findings in this field, knowledge and experience from different disciplines is necessary. Therefore, research teams consist of people from these diverse disciplines. It should be noted that with the increasing demand and interest in medical imaging, there are expected to be more job opportunities in the near future.
Medical imaging can be broken down into two steps: 1) Image acquisition; and 2) image analysis. In the last decades, there have been important improvements in both fields. Innovators working on the image acquisition side have been developing imaging machines to obtain better resolution images while also trying to minimize the radiation doses applied to, or drawbacks suffered by, patients.

It’s likely most people are familiar with the most common forms of medical imaging: Computed tomography scans (CT scans), magnetic resonance imaging (MRI), and ultrasounds. These are not the only forms of medical imaging, but they’re certainly the most well-known.

Each form of imaging has been produced for specific purposes. For instance, MRIs are usually used to view soft tissue, such as internal organs. CT scans are usually used to view more solid parts of the body, such as in the case of bone injuries, chest and teeth imaging, and cancerous tissues [1].

Each form of imaging has its own pros and cons. For example, in CT scans, exposure levels (X-ray tube amperage and peak kilovoltage), slice thickness, and volume of interest (VOI) affect the resolution of the images. Higher exposure levels, bigger VOI, and smaller slice thickness produce better resolution images. However, optimum parameters may not be used in order to limit the radiation dosage, which can be harmful to humans. Another difference is that CT scans take a shorter time than MRI scans.

On the other side of the process, innovators working on the image analysis have been proposing new ideas and methods to assist doctors and radiologists in diagnosing diseases in their early stages. The earlier a disease is detected, the better the patient’s long term prognosis. For instance, early detection of cancer tissues in organs, calcium plaque in arteries, or osteoporosis in bones, improves the odds of successfully treating a patient.
How can diseases or disorders be diagnosed using medical images? Important features (such as color, shape, movement, volume, etc.) of the organ or object of interest should be clear enough to analyze. The doctors or specialists analyzing the images obviously need knowledge about the normal and abnormal behaviors of the object of interest. Let’s examine what some abnormal situations are and how medical imaging can diagnose them.

1. Early detection of the body rejecting a kidney transplant

According to the National Health Institute (NIH), 17,600 kidney transplants were performed in the United States in 2013 [3].  Many of these diseases lead to kidney transplants, which can lead to the body rejecting the transplanted organ. Early detection of the rejection can be a matter of life and death.

Currently, rejection is diagnosed via biopsy, but a biopsy subjects a patient to risks like internal bleeding and infection. Moreover, the relatively small needle used in a biopsy may lead to over- or underestimation of the extent of inflammation in the entire graft [11]. These problems can be compounded by the fact that transplanted kidneys face a number of surgical and medical complications anyway. Therefore, a noninvasive and repeatable technique is not only helpful, but also necessary for early diagnosis of acute renal rejection.

Researchers have been introducing automatic methods to determine both normal kidney function and symptoms of kidney rejection by using dynamic contrast enhanced magnetic resonance imaging (DCE-MRI). In DCE-MRI, a contrast agent called Gd-DTPA is injected into the bloodstream, and as it permeates into the organ, the kidneys are imaged, rapidly and repeatedly. During the permeation, Gd-DTPA causes a change in the relaxation times of the tissue and creates a contrast in the images. As a result, the patterns of the contrast give functional information. MRIs can also provide good anatomical information, which help in distinguishing between diseases that affect different regions of the kidneys.

1. Diagnosing osteoporosis

Osteoporosis is a bone disease characterized by a reduction in bone mass, resulting in an increased risk of fractures [4]. With osteoporosis, a patient’s bone tissue has less than the normal amount of calcium. Low bone mass and osteoporosis occur more frequently in women. Without diagnosis and prevention, a woman can lose 20%-30% of her bone mass during the first 10 years of menopause [5]. Based on a Surgeon General’s report [6], there were approximately 10 million people over 50 with osteoporosis, and an additional 34 million with low bone mass or osteopenia, in the United States in 2002. Unfortunately, the total number is expected to increase to 61.4 million by 2020.

To diagnose and properly treat osteoporosis, doctors need the bone mineral density (BMD) measurements of the vertebral bones. The BMD measurements are strong predictors of fracture risk. In the Surgeon General’s report, it is stated that the relationship between the BMD score and future fracture is stronger than the relationship between cholesterol levels and future heart attack [6]. The BMD measurements are also used to track bone changes in treated and untreated individuals, allowing doctors to assess the effectiveness of certain drug therapies.

Dual x-ray absorptiometry (DXA) and quantitative computed tomography (QCT) are the most widely used method for taking the BMD measurements. QCT measures the true volumetric density in units of grams per cubic centimeter rather than the areal density measured by DXA.

QCT has an advantage in that it can be performed on any commercial CT scanner with the use of specially designed calibration phantoms and analysis software. Improvements have reduced the initially high radiation dose [4].

Once the test is done, a patient’s BMD is compared to the mean value in a reference population of young, healthy adults. The difference between the scores is referred to as a “T-score.” In the BMD measurements, each T-score decrease corresponds with a 1.5-2.5 times greater risk of fracture [6].

1. Identifying symptoms of autism

Autism is a neuro developmental disorder characterized by impaired social interactions and communications, and also by restricted, repetitive, and stereotyped patterns of behavior. It occurs in all ethnic and social groups, and affects every age group.  Experts estimate that 1 out of 88 children aged eight might have autism, and males are four times more likely to have autism than females [7].
To detect autism in the brain, the intensity, shape and volume of the corpus callosum region are analyzed via medical imaging. Data sets are classified into two categories: 1) training and; 2) testing classes. Training classes consist of information about patients that is already known. The objective is to detect the differences between a normal and an autistic corpus callosum in the testing data sets by means of information extracted from the training data sets. During the past two decades, studies of autism's neuropathology have increased dramatically. One can find an example study in [8].

1. Virtual colonoscopy

A virtual colonoscopy can be used to check for symptoms of pre-cancerous growths (called polyps), cancer, and other diseases of the large intestine and colon [9]. Since colorectal cancer is preventable, it is crucial to detect and treat the cancer in its earliest stage.
Most colorectal cancers begin as polyps. As polyps get bigger, they are more likely to develop into cancer, which then has the ability to spread throughout the body.
Polyp size can help to distinguish benign polyps from cancerous ones [10]. In a virtual colonoscopy, a computer puts the images together to create an animated, three-dimensional view of the inside of the large intestine [9]. Certain processes help to make the image clearer, enabling a doctor or specialist to make determinations about the different polyp types [10].

1. Low radiation dose

Unfortunately, the CT scans that reliably identify tumors also expose the patient to an X-ray dose. “It’s known that a radiation dose can increase the risk of cancer, but nobody knows exactly how much,” said Jeffrey Fessler, a professor of electrical and computer engineering at University of Michigan. “In repeat scan situations, it’s crucial that the dose be very low,” he said [13]. Based on the report in [13], Fessler’s team has been investigating methods to reduce the dose from around 2 mSv to between 0.24 and 0.4 mSv. The drawback is that the images taken at these low X-ray doses do not have good resolution as higher-dose. The advance would also benefit children and adolescents, who are thought to be more sensitive to radiation. Now, the National Institute of Health (NIH) has provided Fessler and his team $1.9 million to achieve low dose CT scans with high resolution [13].

Recourses to do discoveries
As we can see, medical imaging plays an important role in diagnosing and treating many diseases – and the above list is just a snapshot of the many ways medical imagining is used. To maintain these many projects, there are important national offices supporting these projects. In the US for instance, there is the National Institute of Health (NIH), National Science Foundation (NSF), and other national and private institutes. In 2010, the total budget of the NIH was approximately $31 billion, which covers all its expenses [12]. The NIH devotes around 10% of its funding to research within its own facilities, whereas 80% of its funding in research grants is given to outside researchers. The NIH spent $10.7 billion on clinical research, $7.4 billion on genetics-related research, $6.0 billion on prevention research, $5.8 billion on cancer, and $5.7 billion on biotechnology.

REFERENCES

1.      http://www.diffen.com/difference/CT_Scan_vs_MRI.

2.      J. L. Prince and J. M. Links, Medical imaging signals and systems, Prentice Hall, 2005.

3.       https://www.niddk.nih.gov/health-information/health-statistics/

4.      G. M. Blake, H. W. Wahner, and I. Fogelman. Dual energy x-ray absorptiometry and ultrasoung in clinical practice. Martin Dunitz, 1999.

5.      S. Tapp. A markov model of secondary prevention of osteoporotic hip fractures. Ph.D Dissertation, 2003.

 6.     Department of Health and Human Services. A report of the surgeon general: Bone Health and Osteoporosis. U. S. Public Health Service, 2004.

7.      http://www.ninds.nih.gov/disorders/autism/detail_autism.htm.

8.      HossamAbd EL Munim, Aly A. Farag, and Manuel F. Casanova, “Frequency-Domain Analysis of the Human Brain for Studies of Autism,” Proceedings, 7th IEEE International Symposium on Signal Processing and Information Technology, ISSPIT 2007, Cairo, Egypt, December 15-18, 2007, pp. 1198-1203. 

9.      http://digestive.niddk.nih.gov/ddiseases/pubs/virtualcolonoscopy.

10.    Dongqing Chen, Aly A. Farag, Robert L. Falk, and Gerald W. Dryden, "Variational Approach Based Image Pre-processing Techniques for Virtual Colonoscopy," Book Chapter of Biomedical Image Analysis and Machine Learning Technologies: Application and Techniques, Editors: Fabio Gonzalez and Eduardo Romero, 2009. 

11.    Farag, A. A., El-Baz, A., Yuksel, S. E., El-Ghar, M., &Eldiasty, T. (2006). A framework for the detection of acute renal rejection with dynamic contrast enhanced magnetic resonance imaging. In Proceedings of the 3rd IEEE International Symposium on Biomedical Imaging, (pp. 418-421), Washington, DC: IEEE Press.

12.    http://www.wikipedia.org.

13. http://bme.umich.edu/faster-image-processing-to-fight-lung-cancer/