Computer Aided Diagnosis in Rosacea Erythema

[Guide]

We may be hearing more of using computer aided diagnosis (CAD) in diagnosing rosacea. This post/thread will continue to add more information on this subject as more papers are released. This may help reduce the number of reports on misdiagnosis. These papers on this subject are coming out of China, showing that this country is way ahead on using AI with computer aided diagnosis in dermatology. 

Wouldn't it be important for the RRDi to start collecting images of rosacea in our gallery from members willing to volunteer for this?  What are your thoughts on this?

"Although computer-aided diagnosis (CAD) is used to improve the quality of diagnosis in various medical fields such as mammography and colonography, it is not used in dermatology, where noninvasive screening tests are performed only with the naked eye, and avoidable inaccuracies may exist." [1]

The above paper concluded that "Lastly, unlike previous studies, our method provides a solution to classify multiple diseases within a single image. With higher quality and a larger quantity of data, it will be viable to use state-of-the-art models to enable the use of CAD in the field of dermatology."

The paper relied on images at "Dermnet NZ, an archive of skin disease information launched and maintained by a group of dermatologists from New Zealand" and "gathered a total of 15,851 images" with "input and output resolution consistent at 304 × 304 pixels so that "zero-padding allowed us to keep the input and output resolutions consistent, thereby allowing the retention of information present on the border of our images."  "Among the images obtained through Dermnet, the erythema of 100 images was masked by dermatologists, to be used as a ground truth. For segmentation, 60 images were used for training, and 40 images were used for testing." Notice below the results of the top level categories shown from this endeavor what resulted in the number one spot. 

Another paper on this same subject states, "We statistically confirmed the distinct clinical features of inflammatory papular dermatoses of the face and proposed a diagnostic algorithm for clinical diagnosis." [2]

A third paper states that using "a novel framework based on deep learning trained by a dataset that represented the real clinical environment" the results indicted "could accurately classify most common dermatoses that we encountered during outpatient practice including infectious and inflammatory dermatoses, benign and malignant cutaneous tumors." [3]

A fourth paper indicated that another group in China "developed a convolutional neural network (CNN) based on two datasets from a consecutive series of patients who underwent the dermoscopy" with results "achieved at least moderate consistency with the reference standard." [4]

"Due to the high incidence of skin tumors, the development of computer aided-diagnosis methods will become a very powerful diagnosis tool for dermatologists. The skin diseases are initially diagnosed visually, through clinical screening and followed in some cases by dermoscopic analysis, biopsy and histopathological examination. Automatic classification of dermatoscopic images is a challenge due to fine-grained variations in lesions. The convolutional neural network (CNN), one of the most powerful deep learning techniques proved to be superior to traditional algorithms." [5]

Another group "developed an artificial intelligence dermatology diagnosis assistant (AIDDA)" focusing on "psoriasis (Pso), eczema (Ecz), and atopic dermatitis (AD)" and the results are that "AIDDA is thus already achieving an impact in the diagnosis of inflammatory skin diseases, highlighting how deep learning network tools can help advance clinical practice." [6]

End Notes

[1] Sci Rep. 2021; 11: 5350.
AI-based localization and classification of skin disease with erythema
Ha Min Son, Wooho Jeon, Jinhyun Kim, Chan Yeong Heo, Hye Jin Yoon, Ji-Ung Park, Tai-Myoung Chung

[2] Predictive Model for Differential Diagnosis of Inflammatory Papular Dermatoses of the Face

[3] A Deep Learning Based Framework for Diagnosing Multiple Skin Diseases in a Clinical Environment

[4] Deep learning-based, computer-aided classifier developed with dermoscopic images shows comparable performance to 164 dermatologists in cutaneous disease diagnosis in the Chinese population

[5] Skin Diseases Classification Using Deep Leaning Methods

[6] A deep learning, image based approach for automated diagnosis for inflammatory skin diseases

[Guide]

"At present, the development of skin imaging technologies [dermoscopy, reflectance confocal microscopy (RCM), optical coherence tomography (OCT), etc.], teledermatology, and artificial intelligence (AI) has profoundly and comprehensively changed the nature, service model, and public recognition of dermatology (1). Skin imaging, as an important technical system in modern dermatology, continues to gain the attention of researchers and the wider community alike....n sum, this special Research Topic of skin imaging technology, teledermatology, and AI in dermatology collects the latest endeavors, progresses, experiences, and challenges in these fields and shows the successful integration of modern information technology into dermatology."

Front Med (Lausanne). 2021; 8: 757538.
Editorial: Progress and Prospects on Skin Imaging Technology, Teledermatology and Artificial Intelligence in Dermatology
Chengxu Li,  Je-Ho Mun,  Paola Pasquali,  Hang Li,  H. Peter Soyer, and Yong Cui