Scoring Methodology

This page provides the technical detail on how Legit.Health computes the automated SCORAD and EASI scores for atopic dermatitis, from image input through to the final severity value. Both indices are built from the same two image-derived measurements, extent and intensity, so this page explains those first and then how each formula combines them.

The example images on this page are synthetic dermatology imagery used for illustration; they are not patient records.

Body surface area (BSA)

Body surface area, the extent of skin affected by eczema, is the foundation of both SCORAD and EASI. It is also the single largest source of variability in manual scoring, and the input the AI measures most precisely.

How the AI measures affected area

A deep learning segmentation model (DeepLabV3+ architecture) analyses each body area photograph and classifies each pixel as:

• Affected skin (eczematous lesion)
• Unaffected skin (normal or uninvolved)
• Non-skin (background, clothing, hair)

The percentage of affected skin relative to total visible skin area is computed per photograph. These per-image percentages are combined across all captured body areas using anatomical weighting to produce the total BSA (0–100%).

$$\text{BSA}\,(\%) = 100 \times \sum_{a} w_a \cdot \frac{A_{\text{affected},a}}{A_{\text{skin},a}}$$

where, for each captured body region $a$, $A_{\text{affected},a}$ and $A_{\text{skin},a}$ are the affected and total visible skin areas (in pixels), and $w_a$ is that region's anatomical weight (the weights sum to 1).

Input: body area photograph

Output: affected areas detected across the body

18.24%Affected surfaceWhole-body extent, measured at pixel level

Why pixel-level BSA matters

In manual scoring, BSA is estimated using the "rule of nines", a coarse visual estimation method. This is the single largest source of inter-rater variability. Pixel-level segmentation replaces that estimation with an objective measurement, eliminating the most variable component of the manual score.

Intensity scoring: the six clinical signs

Alongside extent, both indices grade the intensity of the disease signs on a 0–3 ordinal scale at the representative affected area. SCORAD grades six signs; EASI grades four of them (erythema, swelling, excoriation, lichenification) in each body region.

The six signs and what each one captures:

Sign	Description
Erythema	Redness of the affected skin caused by inflammatory vasodilation, graded by how intense and widespread the red discoloration is.
Swelling	Swelling and small raised papules produced by fluid accumulation and inflammatory infiltrate in the dermis, graded by how elevated and palpable the lesions are.
Crusting (oozing/crusts)	Weeping of serous exudate from damaged skin together with the crusts that form as it dries, a marker of acute or secondarily infected lesions.
Excoriation	Linear erosions and scratch marks left by repeated scratching in response to itch, reflecting the intensity of disease-related pruritus.
Lichenification	Thickening of the skin with accentuated skin markings and a leathery surface, resulting from chronic scratching and rubbing over time.
Dryness	Xerosis, meaning roughness and fine scaling, assessed on skin away from active lesions, reflecting the impaired skin barrier typical of atopic dermatitis.

For SCORAD, the intensity total $B$ is the sum of all six sign scores (0–18).

Subjective symptoms (pruritus and sleep) are separate from these signs: they are patient-reported, and EASI has no subjective component.

The SCORAD formula

SCORAD (SCORing Atopic Dermatitis) combines extent, intensity and subjective symptoms:

$$\text{SCORAD} = \frac{A}{5} + \frac{7B}{2} + C$$

where:

• $A$ = Extent (affected body surface area, 0–100%)
• $B$ = Intensity (sum of six sign scores, each 0–3; total 0–18)
• $C$ = Subjective symptoms (pruritus NRS 0–10 + sleep disturbance NRS 0–10; total 0–20)

The maximum SCORAD is 103. The objective SCORAD (A/5 + 7B/2) has a maximum of 83, and this is what the AI computes. Subjective symptoms are patient-reported.

The EASI formula

The Eczema Area and Severity Index (EASI) is a co-accepted primary efficacy measure for AD, alongside SCORAD. EASI scores four signs (erythema, swelling, excoriation, lichenification) across four body regions, each weighted by regional BSA.

$$\text{EASI} = \sum_{r} w_r \times \text{BSA}_r \times \left(\text{erythema}_r + \text{swelling}_r + \text{excoriation}_r + \text{lichenification}_r\right)$$

The maximum EASI is 72. Unlike SCORAD, EASI has no subjective component, so it is entirely image-derived.

Severity scale

Absent

Per sign0

SCORAD0

EASI0

Mild

Per sign1

SCORAD1–24

EASI1–7

Moderate

Per sign2

SCORAD25–50

EASI8–21

Severe

Per sign3

SCORAD>50

EASI>21

The same four-level classification applies to each individual sign (scored 0–3), the overall SCORAD score (0–103) and the EASI score (0–72). The bands are the standard SCORAD and EASI severity cut-offs used in clinical trials.

SCORAD vs. EASI

Aspect	SCORAD	EASI
Scale	0–103 (objective 0–83)	0–72
What is scored	Extent + six signs + subjective symptoms	Four signs across four weighted regions
Patient-reported outcomes	Yes: pruritus + sleep NRS	No: purely objective
Regulatory acceptance	FDA and EMA accepted	FDA and EMA accepted

SCORAD and EASI are computed automatically by the Legit.Health platform. The choice of primary and secondary endpoints is made during protocol design based on sponsor preference, prior precedent in the indication, and regulatory strategy.

Manual vs. automated scoring

Manual scoring

SCORAD / EASI graded by a clinician

• BSA estimation

  Visual rule of nines, ±15–20% variability

• Sign scoring

  Subjective, 0–3 per sign

• Reproducibility

  Inter-rater variability across sites

• Time per assessment

  5 to 10 minutes

Automated scoring

Automated SCORAD / EASI computed by Legit.Health

• BSA estimation

  Pixel-level segmentation, objective

• Sign scoring

  Objective AI, 0–3 per score

• Reproducibility

  Identical score for identical image, every time

• Time per assessment

  Under 2 seconds, plus 3 to 5 minutes of image capture

Image-based scoring

Because the platform scores from standardised photographs rather than a live examination, signs that a clinician might normally palpate, such as swelling, are read from visual cues. Each AI model is accepted only when it is non-inferior to expert inter-rater variability on the same photographs, so the AI is at least as consistent as dermatologists working from the same images.