Skip to main content
eScholarship
Open Access Publications from the University of California

Dermatology Online Journal

Dermatology Online Journal bannerUC Davis

Using cellphones for teledermatology, a preliminary study

Main Content

Using cellphones for teledermatology, a preliminary study
Pil Chung, Thomas Yu, Noah Scheinfeld
Dermatology Online Journal 13 (3): 2

St.Luke's-Roosevelt Hospital Center, NYC NY

Abstract

We compared live and instantaneous digital image examination of hospitalized dermatology consultation patients and to assess the reliability and the practicality of a teledermatology system equipped with the cellular phone cameras. A dermatology resident examined ten hospitalized dermatology consultation patients. Then a medical student took digital pictures of these lesions from consenting patients with a cell phone camera. The pictures were forwarded to a supervising dermatologist who was able to review them using the identical phone. The differential diagnoses given by the resident and the supervising dermatologist were recorded. Identifiable facial images were never taken. The concordance of the diagnoses between resident and off site attending was 80 percent (8/10 cases) when a diagnosis could be made remotely. In one case no diagnosis could be made because of poor photographic quality and small screen and in one case the attending diagnosis was incorrect. The practicality of this cellular phone camera teledermatology system depended on the Sprint wireless network, which was sometimes unreliable. Cellular phones with cameras offer a means to deliver bedside supervision of dermatology residents but transmission issues, small screen size, and photographic quality complicate their use.


Most studies using digital images in dermatology focus on teledermatology outpatient consultations or on the analysis of pigmented lesions, perhaps because the vast majority of dermatology patients are seen as outpatients and melanomas are associated with significant morbidity [1-10]. Hospitalized patients in need of dermatology consultations often have complex problems and also need prompt attention. Frequently, common dermatologic conditions are missed by the referring service; dermatologic consultation could change the diagnosis and the treatment in more than 60 percent of the patients [2].

The use of digital images is shown to be accurate and reliable method that allows dermatologists to make correct diagnoses [1]. However, the technical difficulty of uploading images onto a hospital's computer network in order to provide remote access has been an obstacle to implementation in many places.

In this study, we decided to assess the utility and the practicality of instantaneous digital image examinations using the camera-equipped cell phone in order to perform dermatology consultations on hospitalized patients.


Methods

Ten consultations were received when a resident and medical student were together performing inpatient dermatology consultations. All ten patients consented to imaging and definitive diagnostic procedures (biopsies, potassium hydroxide preparations) were performed on eight patients. The skin lesions were imaged with a Dual-Band Sprint Smart Camera SCP-5300 PCS Phone (Sanyo Electric Co., Ltd, Japan) using the default lens. The time that required to obtain necessary images and to forward them successfully to the supervising dermatologist was recorded. After obtaining the digital images, the best network signal was located by ambulating with the cellular phone camera and then the transfer of images was performed.


Camera settings

In order to standardize the imaging process, specific camera settings were used. The Dual-Band Sprint Smart Camera SCP-5300 PCS Phone automatically sets flash, brightness, white balance, and color tone, although it is possible to modify these options. The shutter speed and lens aperture were not controllable in this cellular phone camera. The Dual-Band Sprint Smart Camera SCP-5300 PCS Phone is provided with 3 image settings (low, medium, and high) and the quality was set to high to preserve the intricate details and textures of skin lesions. The image of compression was preset by the phone-camera and no external flash unit was used to obtain images.


Data Collection

Images were taken under different (available) lighting conditions; the patients were in different locations in the hospital. Several images were taken of each patient. One image included the lesion (lesion location based on referring physician's description of the location and distribution of the lesions) viewed at the closest possible focal length. Larger lesions or lesions with a large distribution were imaged in two or more views. Another image that encompassed the whole patient was acquired whenever possible. If the rash involved more than one area, regional images were acquired as well.

The imaging was performed by a medical student who was neither a dermatologist nor a photographer and it occurred on the same day as the in-person examination by a dermatology resident. The images were then transferred to the phone of the supervising dermatologist for examination.

A standardized history was obtained from each patient, including demographic information, past medical history, medications, allergy, history of present illness, associated symptoms, and previous treatments. Patients also were asked comprehensive questions about their skin lesions including location, duration, pain, pruritus, and systemic symptoms such as fever, fatigue or weight loss.

After examining the patient the dermatologic resident reported his differential diagnosis (up to 3 diagnoses) to the medical student who recorded them. Subsequently the medical student proceeded to take pictures with the phone camera. The digital images were first stored in the online picture album accessed through the cellular phone and then they were transmitted to another identical phone for examination by the supervising board certified attending dermatologist. If the obtained images were underexposed or undesirable for other reasons as determined by the medical student, the image was erased and another image was taken until a desired image was acquired. In order to comply with the Health Information Portability and Accountability Act (HIPPA) regulations, the images that could potentially reveal the identity of patients such as complete face or body were discarded or avoided.

The time required to obtain necessary digital images and to forward them successfully to a remote cellular phone camera was measured for all 10 cases. When the network signal was either weak or unavailable, the cellular phone camera was moved to different locations until the network signal was strong enough to complete the transfer of digital images.


Reliability assessment

For our primary outcome, we were interested in reliability or interobserver agreement. But in this study we were not able to accommodate two attending dermatologists and only one attending dermatologist examined all of the patients.

Accuracy was our secondary outcome. When clinically warranted, skin punch biopsies (3-4mm diameter) were obtained and the results were used as the standard to which the clinical dermatologic diagnoses were compared.


Results

It took on average of 5-15 minutes to acquire, download, and send three high-quality images. The concordance of the diagnoses between resident and off-site attending was 80 percent (8/10 cases) (Table 1), when a diagnosis could be made remotely. In one case no diagnosis could be made because of poor photographic quality and small screen size and in one case the attending diagnosis was incorrect (Table 2).


Discussion

Other studies have reported the value of hand-held devices for teledermatology. A 1998 Austrian study was very similar to ours in design (similar phones, pictures taken by a medical students) but had a few differences (the images were viewed on a full size computer screen, ten times the number of patients were imaged, no trained dermatologist accompanied the medical student, and only one diagnosis and one differential diagnosis were accepted) [3]. The concordance rate in the Austrian study 70 percent compared with 80 percent in our study. In 2004 the dermatology department at State University of New York at Brooklyn equipped residents with cell phones so that images could be sent to an attending for consultation; the concordance data associated with this system was not reported. In 2006 the Graz group reported that images acquired with and displayed on a personal digital assistant (Sony Cliè PEG-NZ90; Sony, Tokyo, Japan) with a built-in camera (2-megapixel charged couple device with a Sony lens with a focus range of 0.1 m to infinity, flash, macro mode and x 2 digital zoom) by a medical student resulted in a global concordance of three consultants compared to face to face diagnosis of 79 percent [9].

Larsen et al. underlined some of the technical issues regarding the feasibility of Universal Mobile Telephone System mobile phones for video consultations in the home of patients with diabetic foot ulcers [10]. This report found that consultations took from 5 to 18 minutes. Connectivity problems occurred in seven of the fifteen consultations. The off-site consultants felt confident about the diagnoses and care plans reached in fourteen of the fifteen consults. A study done by Braun et al. reported that off-site wound-care consultants felt comfortable making a diagnosis with cell phones in 82 percent of cases. Their image resolutions were identical to those of the cell phone camera we used in our study.

In 2004 Tsai studied teleconsultation using a low-resolution mobile-camera phone for remote management of extremity wounds. Compared to onsite diagnoses, three surgeons were able to achieve 80 percent, 76 percent, 66 percent, and 74 percent agreement, respectively, in remote diagnosis regarding presence of gangrene, necrosis, erythema, and cellulitis or infection [12]. Recognition of gangrene had the highest agreement percent (80 %), sensitivity (85 %), and specificity (93 %); from 68 to 90 percnet of image sets provided equivalent diagnoses of wound descriptors. Tsai reported that 83 percent of wounds were remotely managed with treatment recommendations regarding whether antibiotics or debridement were appropriate. Other reports exist that have found that cell phone cameras can provide off site surgeons with images that can effectively be used to determine surgical care [13,14].

The ability of dermatologists to provide reliable diagnoses and recommend appropriate treatment based on computer images has been studied. The Graz studies evaluated the the use of images on a computer screen in an outpatient settings [3]. Another teledermatology study focused on hospitalized dermatology consultation patients [1]. There is good interobserver and intraobserver agreement [3] and Whited et al [3, 4]. The concordance of live and digital differential diagnoses averages 80 percent but can approach almost 95 percent.

The patients' responses to the use of digital imaging were positive. All 10 patients agreed to have the medical student take digital images of their skin lesions and a few of the patients were very impressed with the cellular phone camera technology. Another study noted that 85 percent of patients felt comfortable with a dermatologist taking patient images during a face to face consultation for discussion with a supervisor [5].

Our study exemplifies the ease of using digital images for diagnosing dermatologic diseases in hospitalized dermatology consultation patients. This type of teledermatology could be applied to urgent-care dermatology clinics where daily supervision by an attending does not occur. In this study, we used the Dual-Band Sprint Smart Camera SCP-5300 PCS Phone, which could be obtained at the time of this study (August, 2003) at a retail cost of approximately $150. This camera phone provided 3 different resolutions (low, medium and high). The acquired images were stored in an on-line picture album accessed through the cellular phone. Then they could be transmitted to other identical or compatible phones within the Sprint cellular phone network. Economically this teledermatology system also has an advantage over one using digital cameras that are very expensive to set up and require extra computers or broad-band internet connections.

One drawback of this system was that the strength of network signal necessary to transfer the images was neither uniform nor consistent throughout the hospital. When the network signal was too busy or weak, the medical student moved to different locations to receive the best network signal. The feasibility and the practicality of this particular teledermatology system equipped with the cellular phone cameras did depend essentially on the reliability of the Sprint wireless network. The strongest network signal was located randomly in different parts of the hospital but usually was close to the windows or elevators. We did not experience any major problems with sending the digital images and it took on average of 5 to 15 minutes to sends 2 to 3 of the phone's highest quality images.

In our study the data and the image collection occurred in the hospital, where lighting and other conditions were variable compared to the clinic. Initial patient interviews, examinations, and image collections were performed in different parts of the hospital such as hospital wards or the emergency department. Similar to a previous study done by Scheinfeld et al. [1], the results of this study also indicate that is possible to obtain quality images in a variety of lighting and environmental situations without expert or extensive knowledge of photographic technique or use of expensive equipment. We tried to generate multiple image sets and select the best, a technique that has been used by Kvedar et al. [6].

The skin conditions encountered in hospital patients differ in many respects from those found in outpatient clinics. Many hospitalized patients presented with generalized vascular reaction pattern eruptions (e.g., drug eruptions, urticaria). In our study, they were found to be difficult to view in a digital image due to its two-dimensional pattern with subtle or minimal texture or surface change. Digitally diagnosing these eruptions that included generalized vascular reaction patterns or papulosquamous patterns resulted in the greatest variances between instantaneous digital and definitive diagnoses. For example, it was easy to misdiagnose a drug eruption as tinea versicolor or vascular reaction. The use of multiple image sets of patients aided accuracy of diagnosis and helped to counteract differences in environment; as our study progressed, the ability of the medical student to capture clinically useful images improved progressively.

Distinct eruptions and three-dimensional lesions such as nodules or cysts rarely resulted in variations among the instantaneous digital, live, and the definitive diagnoses. Discoid lupus erythematosus, seborrheic dermatitis, psoriasis, and inflamed cyst were easy to recognize. There were no consultations in which basal cell, squamous cell, or melanoma cancers were suspected. Some outpatient teledermatology studies have suggested that discrete lesions such as basal cell cancers present challenges for digital diagnosis [7].

The digital examiner's eye also needs to be retrained to see subtle palpable objects in two dimensions, perhaps partially by using other visual cues [1]. For example, urticaria can be accurately interpreted if the eye is trained to look for a zone of pallor associated with erythema. Even seasoned dermatologists may have difficulty in assessing generalized eruptions and complex dermatologic reaction patterns in digital images captured by a cellular phone camera. Even face to face dermatology consultations in a hospital do not always result in an immediate diagnosis, which is why biopsies are done. Sometimes even with histopathology report in hand a definite diagnosis may elude a dermatological consultant.

In sum, teledermatology supervision of residents using cell phones appears to be feasible. Cellular phones with cameras offer a means to deliver bedside supervision of dermatology residents. However, transmission issues, small screen size, and photographic quality complicate their use. This study did not assess if those unschooled in dermatology could take pictures form which an attending dermatologist could make a diagnosis. The importance of the ability of a dermatology resident to focus attention on a specific area and to image that area should not be underestimated. As cell phone consultations probably will not be equated with in person examination by credentialing authorities, payors, and the legal system, they remain a curiosity for now. However, as cell phone cameras are improved and specific attention is directed toward developing skills in non-dermatologists who might image patients, it is possible that they could be an adjuvant for providing and supervising the provision of hospital bedside dermatology care[15].

References

1. Scheinfeld N, Kurz J, Tepplitz E.A comparison of the concordance of digital images, live examinations, and skin biopsies for the diagnosis of hospitalized dermatology consultation patients. Skinmed. 2003;2:14-19.

2. Falanga V, Schachner LA, Rae V, et al. Dermatologic consultations in the hospital setting. Arch Dermatol. 1994;130:1022-1025.

3. Lowitt MH, Kessler II, Kauffman CL, et al. Teledermatology and in-person examinations: a comparison of patient and physical perceptions and diagnostic agreement. Arch Dermatol. 1998;134:471-476.

4. Whited JD, Hall RP, Simel DL, et al. Reliability and accuracy of dermatologists' clinic based and digital image consultations. J Am Acad Dermatol. 1999;41:693-702.

5. Loane MA, Bloomer SE, Corbett RZ, et al. Patient satisfaction with realtime teledermatology in Northern Island. J. Telemed Telecare. 1998;4:36-40.

6. Kvedar JC, Edward RA, Menn ER, et al. The substitution of digital images for dermatologic physical examination. Arch Dermatol. 1997;33:161-167.

7. Krupinski EA, LeSueur B, Ellsworh L, et al. Assessment of diagnostic accuracy using a digital camera for teledermatology. From the American Telemedicine Association Available at: http://www.atmeda.rog/confer/1999/krupinski/telederm.html. Accessed May 25, 2000.

8. Massone C, Lozzi GP, Wurm E, Hofmann-Wellenhof R, Schoellnast R, Zalaudek I, Gabler G, Di Stefani A, Kerl H, Soyer HP. Cellular phones in clinical teledermatology. Arch Dermatol. 2005;141(10):1319-20.

9. Massone C, Lozzi GP, Wurm E, Hofmann-Wellenhof R, Schoellnast R, Zalaudek I, Gabler G, Di Stefani A, Kerl H, Soyer HP. Personal digital assistants in teledermatology. Br J Dermatol. 2006;154(4):801-2.

10. Larsen SB, Clemensen J, Ejskjaer N. A feasibility study of UMTS mobile phones for supporting nurses doing home visits to patients with diabetic foot ulcers. J Telemed Telecare. 2006;12(7):358-62.

11. Braun RP, Vecchietti JL, Thomas L, Prins C, French LE, Gewirtzman AJ, Saurat JH, Salomon D. Telemedical wound care using a new generation of mobile telephones: a feasibility study. Arch Dermatol. 2005;141(2):254-8.

12. Tsai HH, Pong YP, Liang CC, Lin PY, Hsieh CH. Teleconsultation by using the mobile camera phone for remote management of the extremity wound: a pilot study. Ann Plast Surg. 2004;53(6):584-7.

13. Hsieh CH, Tsai HH, Yin JW, Chen CY, Yang JC, Jeng SF. Teleconsultation with the mobile camera-phone in digital soft-tissue injury: a feasibility study. Plast Reconstr Surg. 2004;114(7):1776-82.

14. Hsieh CH, Jeng SF, Chen CY, Yin JW, Yang JC, Tsai HH, Yeh MC. Teleconsultation with the mobile camera-phone in remote evaluation of replantation potential. J Trauma. 2005;58(6):1208-12.

15. Kanthraj GR, Srinivas CR. Store and forward teledermatology. Indian J Dermatol Venereol Leprol. 2007;73(1):5-12

© 2007 Dermatology Online Journal