Beyond 911: Medical applications of mobile phones

At this year’s SPIE Photonics West meeting in San Francisco, gynecologist Christophe Milien of University Hospital in Mirebalais, Haiti, began his talk with a Haitian saying: Tout moun se moun. (“Every person is a person.”)

But “poor people do not have the luxury of being sick,” he continued. Cervical cancer is a leading cause of death in Haiti and other developing countries, because women are never screened for the disease. They have no access to care, particularly cancer treatment centers or specialists.

Cheaper screening methods and medical equipment are necessary to fight just only cervical cancer, but any disease that can be treated if detected early enough. Indeed, the need for affordable medical devices in low-resource settings is well documented.

At Photonics West, academics and entrepreneurs shared their ideas on how to develop medical imaging devices using existing hardware and processing capabilities that already have a strong infrastructure.

Enter the smartphone

Milien’s hospital screens women at mobile clinics using a modified smartphone that enables colposcopic imaging of the cervix. He uploads images to a web portal, enabling off-site clinicians to review and provide advice. “Visually, there is not much noticeable difference between mobile and conventional colposcopes,” he said.

Academic-turned-entrepreneur David Levitz, who is commercializing the mobile phone colposcope through his company MobileODT , can vouch for this.

Levitz knew little about cervical cancer when he won a grant to develop a smartphone colposcope. His background in optical coherence tomography (OCT) led him to the idea of a multimodal imaging system to be used on a smartphone. “The real power is looking for disruption in the epithelial tissues that develop cancer,” explains Levitz. “Optical technologies are best for detecting disruption of layers in scattering media.”

After successfully raising seed money, Levitz was able to research and build prototype lenses and light sources that attach to a mobile phone. However, it wasn’t until a National Cancer Institute (NCI) conference in January 2014 that a doctor encouraged him to make a colposcope product.

“Colposcopy is a weird optical problem,” says Levitz. “Only a small fraction of light reaches the camera. The way to solve this problem is to shine a lot of light on the target so the small amount captured is higher intensity.”

MobileODT’s technology consists of a lens that attaches to a mobile phone camera; a white LED light source; and a handle and case. The setup combines high resolution bright-field imaging to provide the view clinicians are accustomed to working with; polarization difference imaging (PDI) to provide information about the micro-structural patterns in the superficial layer of tissues; and multispectral imaging to provide information about tissue composition. "[These] complement each other and can be implemented inexpensively,” Levitz adds.

For a tenth the cost of current standard colposcopes, MobileODT’s product is promising. But the mobile colposcope must satisfy the FDA’s stringent regulatory requirements before it can be marketed.

In the meantime, MobileODT is partnering with universities and hospitals for long-term R&D in multispectral and PDI imaging. That initiative includes a project with the Scripps Research Institute to conduct a comparison study of old and new imaging technology, while also tracking pathology data, at Pro Salud clinic in Tijuana, Mexico. “This is a way to take technology and leverage it forward,” said Scripps gynecologist Bruce Kahn.

Monitoring arthritis

Caught early enough, cervical cancer can be treated and lives saved. But even for chronic diseases for which there is no cure, mobile health provides a way of providing care to patients who otherwise have no medical resources. Diseases such as arthritis represent a huge cost to society, yet they often have too low a threat level to justify repeated visits to a doctor.

An important step toward managing chronic diseases such as arthritis is to quantitatively monitor disease progression and evaluate treatments. A Vancouver-based company, eTreatMD , has developed a mobile phone imaging technique that allows a patient to do this reliably at home, greatly reducing the cost of diagnosis and treatment management.

For arthritis of the hand, diagnosis can be done by visual inspection. Symptoms are manifested as changes in hand anatomy. The current positive metric for osteoarthritis is an x-ray, but the x-ray is not indicative of actual symptoms. If people can reliably monitor symptoms and track the causes of flare-ups and results of treatments, they can take control of their treatment and choose when to seek intervention.

“What people need is a way to manage their treatment. They are orphaned by the medical system,” says Nick MacKinnon, CEO of eTreatMD. “People don’t want to have chronic disease impact their quality of life.”

MacKinnon’s company is commercializing an image-processing mobile phone app that generates a report on the variability of hand measurements. The user places his or her hand on a sheet of paper and takes a picture. Cloud processing software removes smartphone camera distortion, identifies marker points on the segmented image, and takes anatomical measurements of finger length, width, and joint thickness down to half-mm scale. It also identifies irregularities such as swelling or abnormal angular deviation of a joint. The result is an analysis of anatomy, and a graphical report comparing changes over time.

“In most key areas of anatomical measurements, we’re below 1% error,” adds MacKinnon. “If you can identify where there is swelling or abnormality, these correlate with pain levels.” The eTreatMD app also provides space for recording pain levels and conditions, such as weather and other medication, that might evoke a response.

“We help people not to feel helpless by putting measuring and monitoring in the hands of the people who are suffering,” says MacKinnon. He plans to extend eTreatMD to monitor skin conditions and joint injuries, with a focus on using data to manage effective treatment at low cost.

With cheap, readily available hardware, and the ease of cloud processing for image and data analysis, mobile phones offer meaningful possibilities for medical adaptations. They already have a high pixel count, continuous autofocus, and high dynamic range. Simple lens additions and flash modifications can turn them into medical Swiss Army knives.

Explosive adoption of mobile phones in low- and middle-income countries represents many people’s first access to computing and telecommunications.

It might soon represent their first access to medical care, too.