Issue: July 25, 2011
July 25, 2011
3 min read
Save

Computer-assisted melanoma surveillance may play larger role in clinical practice

Issue: July 25, 2011
You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact customerservice@slackinc.com.

HEMONC TODAY Melanoma and Cutaneous Malignancies.

NEW YORK -- Computer-assisted surveillance will be an increasingly important tool in the future to help clinicians diagnose melanoma, according to a speaker here.

Allan C. Halpern, MD, MSc, chief of dermatology service at Memorial Sloan-Kettering Cancer Center in New York City, spoke at the HemOnc Today Melanoma and Cutaneous Malignancies meeting about the technology’s potential to identify suspicious lesions, and as an aid to diagnose individual concerning lesions.

Watson, an IBM program/computer was a “watershed moment in computer evolution. Because now, for the first time, computers can understand natural language and can query an entire huge database to come up with the best possible answer,” Halpern said, noting the computer’s defeat of human rivals on Jeopardy.

IBM is now using Watson to interpret radiologic images, Halpern said. Until such a system is capable of diagnosing melanoma images, there are other systems that can assist in identifying and diagnosing lesions.

Acknowledging that it is impossible for clinicians to keep up with every possible piece of information in the medical literature, Halpern spoke about Isabel – a diagnostic tool that scans PubMed and can point out differential diagnoses that a clinician might not have considered – as a jumping off point to using computers to assist in diagnosing melanoma.

In terms of identifying images and following changing lesions, a computer would need pictures from every patient visit, Halpern said. “Logistically that sounds hard, but given that every cell phone has a camera in it, it’s not hard to believe that you could create a system in which a patient walks through the door and gets a full body picture that’s not that expensive.”

Over time, one problem with using multiple images is registering the images perfectly one on top of the other, because patients’ bodies change. Another problem is curved surfaces and 2D images, which make registration even more difficult.

On the other hand, in the 3D environment, registration is easier. Halpern cited epidemiologic studies in which he tracked moles in children 180· around the body, from neck to toes with 3D resolution, to hone in on depth of lesions below a millimeter. “Once we move to the 3D platform, it will be easy for the computer just as if you had instantaneous capture.”

The technology has to overcome barriers to practice before patients will accept it, however.

“The cultural challenges are the bigger ones,” Halpern said. One of these challenges is the culture at the FDA. One company presented a melanoma diagnostic device to the FDA, MelaFind, but even though it surpassed the FDA goals for specificity and sensitivity, it was not approved on a split decision, he said. “The culture at the FDA as it relates to any diagnostic device related to cancer is ‘We have zero tolerance for risk, we do not want any device that has any chance of missing any cancer.’” Halpern noted that dermatologists miss the diagnosis as well.

Halpern said patients’ acceptance of diagnostic instruments for melanoma detection in clinic will depend completely on whether physicians accept them. Even if the instrument misses the diagnosis 3% to 5% of the time, “if you incorporate the instrument into the practice because you think it is helpful, the patient will accept it,” he said. “It’s your reassurance they’re looking for; you are the clinician, not the device. When the day is done, general physician acceptance will drive both FDA acceptance and patient acceptance.”

New technologies can’t replace good practice, Halpern said. “The best machine for melanoma detection of individual lesions isn’t going to help if we don’t do total body skin exams and if we don’t pick the right patients and we don’t convey that information correctly. On the other hand, if we embrace technology, we should be able to figure out how to use them as adjuncts to care.” – by Carey Cowles

Presented at: HemOnc Today Melanoma and Cutaneous Malignancies Meeting; June 25-26; New York.

Disclosure: Dr. Halpern reported receiving consulting fees from Canfield Scientific, DermTech, Lucid Inc., and Quintiles.

The 2012 HemOnc Today Melanoma and Cutaneous Malignancies Meeting will be held April 13-14 at the Grand Hyatt, New York, NY. Learn more at HemOncToday.com/melanoma.

Twitter Follow HemOncToday.com on Twitter.