Computerized quantification of psoriasis lesions with colour calibration: preliminary results

Artificial Intelligence-Based Psoriasis Severity Assessment: Real-world Study and Application

BACKGROUND

Psoriasis is one of the most frequent inflammatory skin conditions and could be treated via tele-dermatology, provided that the current lack of reliable tools for objective severity assessments is overcome. Psoriasis Area and Severity Index (PASI) has a prominent level of subjectivity and is rarely used in real practice, although it is the most widely accepted metric for measuring psoriasis severity currently.

OBJECTIVE

This study aimed to develop an image-artificial intelligence (AI)-based validated system for severity assessment with the explicit intention of facilitating long-term management of patients with psoriasis.

METHODS

A deep learning system was trained to estimate the PASI score by using 14,096 images from 2367 patients with psoriasis. We used 1962 patients from January 2015 to April 2021 to train the model and the other 405 patients from May 2021 to July 2021 to validate it. A multiview feature enhancement block was designed to combine vision features from different perspectives to better simulate the visual diagnostic method in clinical practice. A classification header along with a regression header was simultaneously applied to generate PASI scores, and an extra cross-teacher header after these 2 headers was designed to revise their output. The mean average error (MAE) was used as the metric to evaluate the accuracy of the predicted PASI score. By making the model minimize the MAE value, the model becomes closer to the target value. Then, the proposed model was compared with 43 experienced dermatologists. Finally, the proposed model was deployed into an app named SkinTeller on the WeChat platform.

RESULTS

The proposed image-AI-based PASI-estimating model outperformed the average performance of 43 experienced dermatologists with a 33.2% performance gain in the overall PASI score. The model achieved the smallest MAE of 2.05 at 3 input images by the ablation experiment. In other words, for the task of psoriasis severity assessment, the severity score predicted by our model was close to the PASI score diagnosed by experienced dermatologists. The SkinTeller app has been used 3369 times for PASI scoring in 1497 patients from 18 hospitals, and its excellent performance was confirmed by a feedback survey of 43 dermatologist users.

CONCLUSIONS

An image-AI-based psoriasis severity assessment model has been proposed to automatically calculate PASI scores in an efficient, objective, and accurate manner. The SkinTeller app may be a promising alternative for dermatologists' accurate assessment in the real world and chronic disease self-management in patients with psoriasis.

Estimation error of the body surface area in psoriasis: A comparative study of physician and computer-assisted image analysis (Image J)

BACKGROUND

Assessing the area involved, i.e., the body surface area (BSA), is essential in diagnosing the disease severity of psoriasis. However, the BSA estimated by physicians tends to be over-estimated and has shown high inter- and intra-rater variabilities. Furthermore, no reports suggesting the cause and clinical significance of overestimating the BSA results exist in psoriasis patients.

OBJECTIVES

We aimed to investigate the errors in estimating the BSA affected by psoriasis by comparing the physicians' results with those of computer-assisted image analysis (CAIA) and provide suggestions regarding the clinical implications of such errors.

METHODS

Using 43 images, 36 physicians visually estimated the BSA of psoriasis. Subsequently, the images were evaluated using CAIA (Image J). The BSA values determined by the physicians and CAIA were compared and matched. The BSA percentage was also graded from 0 to 6: no lesion (0), 1-9% (1), 10-29% (2), 30-49% (3), 50-69% (4), 70-89% (5), and 90-100% (6). Each grade range was divided, with the bottom and top 50% defined as the "first half" and "second half," respectively.

RESULTS

The average proportion of correct assessments by physicians was 49.4%. Physicians tended to overestimate the area involved by psoriatic lesions by 8.76%±8.82% compared to CAIA. The largest estimation error (proportion incorrect: 75.7%) was observed in Grade 3. Estimations located in the "second half" of the range demonstrated a higher proportion of inaccuracies compared to those in the "first half." An overestimating error occurred in certain morphological characteristics of the psoriatic lesions.

CONCLUSIONS

Inaccuracy of BSA estimation by physicians may be related that information from the human eye is perceived to be exaggerated compared to the actual size. Further research of using artificial intelligence (AI) technology is needed to reduce quantification error and develop an ideal BSA assessment system. Additionally, education and training are needed for physicians to accurately measure the BSA.

Image-based automated Psoriasis Area Severity Index scoring by Convolutional Neural Networks

Background

The Psoriasis Area and Severity Index (PASI) score is commonly used in clinical practice and research to monitor disease severity and determine treatment efficacy. Automating the PASI score with deep learning algorithms, like Convolutional Neural Networks (CNNs), could enable objective and efficient PASI scoring.

Objectives

To assess the performance of image‐based automated PASI scoring in anatomical regions by CNNs and compare the performance of CNNs to image‐based scoring by physicians.

Methods

Imaging series were matched to PASI subscores determined in real life by the treating physician. CNNs were trained using standardized imaging series of 576 trunk, 614 arm and 541 leg regions. CNNs were separately trained for each PASI subscore (erythema, desquamation, induration and area) in each anatomical region (trunk, arms and legs). The head region was excluded for anonymity. Additionally, PASI‐trained physicians retrospectively determined image‐based subscores on the test set images of the trunk. Agreement with the real‐life scores was determined with the intraclass correlation coefficient (ICC) and compared between the CNNs and physicians.

Results

Intraclass correlation coefficients between the CNN and real‐life scores of the trunk region were 0.616, 0.580, 0.580 and 0.793 for erythema, desquamation, induration and area, respectively, with similar results for the arms and legs region. PASI‐trained physicians (N = 5) were in moderate–good agreement (ICCs 0.706–0.793) with each other for image‐based PASI scoring of the trunk region. ICCs between the CNN and real‐life scores were slightly higher for erythema (0.616 vs. 0.558), induration (0.580 vs. 0.573) and area scoring (0.793 vs. 0.694) than image‐based scoring by physicians. Physicians slightly outperformed the CNN on desquamation scoring (0.580 vs. 0.589).

Conclusions

Convolutional Neural Networks have the potential to automatically and objectively perform image‐based PASI scoring at an anatomical region level. For erythema, desquamation and induration scoring, CNNs performed similar to physicians, while for area scoring CNNs outperformed physicians on image‐based PASI scoring.

Evaluation of a simple image-based tool to quantify facial erythema in rosacea during treatment

Background

Facial erythema is a common symptom in rosacea. To overcome subjectivity in scoring erythema severity, objective redness quantification is desirable. This study evaluated an image‐based erythema quantification tool to monitor facial erythema in rosacea patients during treatment and compared these values to clinical scores.

Materials and Methods

Twenty‐one rosacea patients were treated with topical ivermectin for 16 weeks. Clinical erythema scores and clinical photographs were taken at week 0, 6, 16 and 28. Using ImageJ, RGB images were split into red, green and blue channels to measure the green/red ratio of lesional skin compared with a green sticker. With CIELAB colour space, a* (indicating colour from green to red) of a lesional and non‐lesional facial site was measured, calculating ∆a*. Interobserver concordance and correlation between quantitative and clinical erythema values were determined.

Results

Treatment resulted in reduction of clinical erythema scores. No significant changes in red/green ratios were measured. Lesional a* and ∆a* significantly decreased from baseline to week 16 and 28 (P < .05). A weak correlation existed between clinical scores and lesional a* (R_s = 0.37), and between clinical scores and ∆a* (R_s = 0.30), with a clear trend towards higher a* and ∆a* for higher clinical scores. Interobserver correlation was high (R² = 0.82).

Conclusion

ImageJ is a simple, rapid, objective and reproducible tool to monitor erythema in rosacea patients during treatment. The photographs allow retrospective analysis, evaluation of large and small lesions, and discrimination of subtle redness differences. We recommend using lesional a* to monitor erythema of inflammatory dermatoses in clinical practice.

Objective measurement of erythema in psoriasis using digital color photography with color calibration

BACKGROUND

Traditional metrics for evaluating the severity of psoriasis are subjective, which complicates efforts to measure effective treatments in clinical trials.

METHODS

We collected images of psoriasis plaques and calibrated the coloration of the images according to an included color card. Features were extracted from the images and used to train a linear discriminant analysis classifier with cross-validation to automatically classify the degree of erythema. The results were tested against numerical scores obtained by a panel of dermatologists using a standard rating system.

RESULTS

Quantitative measures of erythema based on the digital color images showed good agreement with subjective assessment of erythema severity (κ = 0.4203). The color calibration process improved the agreement from κ = 0.2364 to κ = 0.4203.

CONCLUSION

We propose a method for the objective measurement of the psoriasis severity parameter of erythema and show that the calibration process improved the results.

Computer-aided analysis with Image J for quantitatively assessing psoriatic lesion area

BACKGROUND/AIMS

Body surface area is important in determining the severity of psoriasis. However, objective, reliable, and practical method is still in need for this purpose. We performed a computer image analysis (CIA) of psoriatic area using the image J freeware to determine whether this method could be used for objective evaluation of psoriatic area.

METHODS

Fifteen psoriasis patients were randomized to be treated with adalimumab or placebo in a clinical trial. At each visit, the psoriasis area of each body site was estimated by two physicians (E-method), and standard photographs were taken. The psoriasis area in the pictures was assessed with CIA using semi-automatic threshold selection (T-method), or manual selection (M-method, gold standard). The results assessed by the three methods were analyzed with reliability and affecting factors evaluated.

RESULTS

Both T- and E-method correlated strongly with M-method, and T-method had a slightly stronger correlation with M-method. Both T- and E-methods had a good consistency between the evaluators. All the three methods were able to detect the change in the psoriatic area after treatment, while the E-method tends to overestimate.

CONCLUSION

The CIA with image J freeware is reliable and practicable in quantitatively assessing the lesional of psoriasis area.

Objective measurement of erythema in psoriasis using digital color photography with color calibration

Traditional metrics for evaluating the severity of psoriasis are highly subjective, which complicates efforts to identify effective treatments in clinical trials. We propose a method for the objective measurement of the psoriasis severity parameter of erythema (redness). This procedure is standardized for different camera systems and lighting environments through the usage of a color card with predetermined color values in order to calibrate the images. Quantitative measures based on the digital color images are shown to correlate well with subjective assessment of psoriasis severity collected using a standard numerical scale by a panel of dermatologists. Additionally, the color calibration process is shown to improve results.