Frequently asked questions

What is a condition?

In the scope of the domain of programming the integration of the device, the term condition is a pathology, as define in FHIR, in a body site of a patient.

It has two extra properties:

• frequency, which specifies how often an encounter will be created automatically to track the condition.
• isTracked, that allows pausing or following up the condition.

Can I create an EpisodeOfCare without assigning it to a condition?

No, currently the platform only admits creating EpisodeOfCare with an assigned condition. This eases the tracking of the condition inside the application.

Where can I find the codes for specifying a body site or a pathology?

The Swagger contains the full list of codes for selecting a body site or a pathology in the description of the corresponding properties.

Can the application also accept .HEIC formats?

No, we do not accept images in .HEIC format. We accept .JPG and .PNG.

What is the minimum information in an API request? What would you recommend?

Only the image of the skin lesion is required for the API to return a value.

We do recommend also sending bodySite, gender, height, weight and birthdate, because for some features it may improve the output, but it's not really required.

At what threshold do you consider the "probability" of a conclusion to be high enough to be confident?

This way of thinking is not the best way to conceptualise the issue.

The conclusions object of the API returns an array of possible conditions. But you must keep in mind that conclusions is the output of a classification algorithm. In other words, the algorithm knows a list of around 300 classes, and when it looks at an image, it distributes 1 across the 300 classes.

As a result, a condition may have a probability of 0.000000001, and another one 0.9. The important thing is that, if you sum the probability of all conditions, you will get 1, because it is a distribution.

With this in mind, it may be true that, if a condition has a probability higher than 0.5, in a large number of cases, that will be the actual condition. But that is not the correct methodology of analysing a distribution, even if it works.

Classifying and diagnosing are different things. To diagnose, you must classify; but it is a separate process. Diagnosing means looking at the distribution of the probability, but also considering the context and understanding the relationship between the conditions.

In this regard, the API contains an object called preliminaryFindings, that outputs keys that utilise the array of conclusions and has clinical knowledge embedded. We call these 'indexes':

Preliminary findings

{
  // ...
  "preliminaryFindings": {
    "hasConditionSuspicion": 99.2,
    "isMalignantSuspicion": 0,
    "isPreMalignantSuspicion": 0.6,
    "needsBiopsySuspicion": 0,
    "needsSpecialistsAttention": 3.86
  }
  // ...
}

For instance, isMalignantSuspicion and isPreMalignantSuspicion combine the probabilities of conditions that are considered malignant and pre-malignant. On the other hand, hasConditionSuspicion is an index that combines the probabilities of actual conditions, reflecting whether or not a picture of skin tissue has a condition.

What is the performance of the classification neural network for each of the conditions?

To measure the performance of the classification algorithm, we use two metrics: Sensitivity and Specificity. In the following table, you will see the value for each condition on each of the metrics, and also the code of the condition according to ICD-11.

The method we follow to extract those metrics consists on running a set of images that are different to those used to train the algorithm through the algorithm. In other words: we have our training dataset, and our testing dataset. The training dataset is used to traing the algorithm, and the testing dataset is used to test it. Thus, the following metrics reflect the results of the internal testing that we perform with our testing dataset. It is worth noting that we always use the same testing dataset, so as to ensure that the algorithm is consistently improving in the right direction with every new version.

Pathology name	Sensitivity (%)	Specificity (%)	ICD-11 code
Acanthosis nigricans	86.79	93.88	ED51.0
Acne	97.05	97.52	ED80
Acne fulminans	100	99.23	ED80.40
Actinic cheilitis	89.13	99.38	EK90.Y
Actinic elastosis	85.71	99.49	EJ20.0
Actinic keratosis	95.04	90.63	EK90.0, XH36H6
Actinic porokeratosis	90.24	99.39	ED52
Allergic vasculitis	76.19	99.29	4A44.BZ
Alopecia	93.48	98.96	ED70
Alopecia areata	92.86	87.77	ED70.2
Amelanotic malignant melanoma	94.12	97.56	XH3TK1
Amyloidosis	91.67	99.57	5D00
Androgenetic alopecia	100	99.47	ED70.0, ED70.1
Anetoderma of Jadassohn Pellizzari	75	99.63	EE41.1
Angiofibroma	90	97.97	XH1JJ2
Angiokeratoma	90.48	98.75	EF20.1, XH4KP7
Angioma	86.21	99.14	XH5AW4
Angular cheilitis	97.92	96.2	DA00.0
Aphthae	90.91	99.56	DA01.10
Aplasia cutis congenita	87.5	99.54	LC60
Atopic dermatitis	85.6	95.36	EA80
Atrophoderma	87.5	99.18	EE7Y
Atypical nevus	96.49	99.02	2F20.1
Basal cell carcinoma	96.64	91.19	2C32, XH2615
Beau's lines	100	99.62	EE10.1Y
Becker's nevus	96.15	99.24	LC02
Behçet disease	91.3	99.29	4A62
Benign keratosis	72.5	94.27	XH0S03
Blue nevus	95.65	98.91	2F20.Y, XH7QJ7
Bowenoid papulosis	77.78	99.67	1A95.1, 2E64.2
Brachyonychia	100	99.87	EC22.0
Bullous pemphigoid	80.81	97.87	EB41.0
Café au lait spot	87.5	99.49	EC23.0
Calciphylaxis	100	99.6	EB90.42
Carcinoma	86.36	98.92	XH63D2
Cellulitis	75	99.06
Cheilitis	53.85	99.45	DA00.0
Cherry angioma	81.82	99.79	2F25
Chondrodermatitis nodularis helicis	86.84	99.79	AA12
Chromomycosis	81.82	99.69	XN066
Cicatricial alopecia	75	99.66	ED70.5Z
Comedo	81.48	95.52	ED80.0
Compound nevus	90.73	97.12	2F20.0Y, XH27A6
Condyloma acuminatum	96	98.74	1A95
Confluent and reticulated papillomatosis	95.24	98.92	ED51.Y
Congenital nevus	90.63	98.84	2F20.2
Contact dermatitis	83.43	94.37	EK00, EK01, EK02
Cornu cutaneum	86.54	98.73	ME60.3
Cutaenous larva migrans	86.96	99.33	1F68.2
Cutaneous lupus erythematosus	79.71	96.44	EB5Z
Cutis marmorata	62.5	99.84	ME66.Y
Cutis rhomboidalis	77.78	99.74	EJ20.0
Cystic acne	96.67	95.33	ED80.3
Darier's disease	80.28	98.85	EC20.2
Dermal nevus	85.71	99.3	XH2MQ5
Dermatitis	85.81	94.69	EA8Z
Dermatitis herpetiformis	78.43	97.94	EB44
Dermatofibroma	93.47	97.37	2F23.0
Dermatofibrosarcoma protuberans	72.73	99.49	XH4QZ8
Dermatosis palmoplantaris juvenilis	88.89	99.59
Dermatosis papulosa nigra	100	99.95	2F21.0
Dermographism	84.62	99.71	EB01.0
Desquamation	100	99.62
Diabetic foot ulcer	99.03	82.83	BD54
Drug eruption	68.14	97.84	EH6Z
Dyshidrosis	91.18	99	EA85
Dyshidrotic eczema	94.74	99.78	EA85.0
Dysplastic nevus	96.97	96.81	XH9035
Dystrofic epidermolysis bullosa	100	99.61	EC32
Eccrine poroma	81.82	99.66	XH25Z9
Eczema	90.39	94.61	EA80
Eczema herpeticum	88.24	99.11	1F00.03
Eosinophilic folliculitis	100	99.66	ED9Y
Epidermal nevus	85.71	99.06	LC00.0
Epidermodysplasia verruciformis	81.82	99.23	x
Epidermoid cyst	83.45	97.86	EK70.0
Epidermolysis bullosa	84.51	98.69	EC3Z
Erysipelas	87.1	98.97	1B70.0
Erythema ab igne	75	99.79	EJ10
Erythema annulare centrifugum	66.67	99.16	EB11
Erythema chronic migrans	66.67	99.22	1C1G.0
Erythema craquele	100	99.41
Erythema elevatum diutinum	100	99.55	EF40.2Y
Erythema infectiosum	66.67	99.68	1F04
Erythema multiforme	91.18	97.78	EB12
Erythema nodosum	92.31	99.64	EB31
Erythrasma	100	99.37	1C44
Erythroderma	88.89	99.26	EB10
Erythrokeratoderma figurata variabilis	71.43	99.59	EC20.0Y
Exanthema	68.75	99.32
Excoriation disorder	85.61	96.93	6B25.1
Extramammary Paget disease	93.33	99.29	2E64.1
Favre Racouchot syndrome	90	99.6	EJ20.0
Fibroma	80.77	99.34	EK71.0
Filiform wart	95.83	98.9	1E80
Follicular mucinosis	84.62	99.61	EB90.1Y
Folliculitis	81.82	99.11
Folliculitis decalvans	83.33	99.74	ED70.50
Folliculitis keloidalis	100	99.79	EE60.Y
Fordyce spots	100	99.85	ED91.0
Furuncle	73.33	99.35	1B75.0
Geographic tongue	92.31	99.82	DA03.1
Gougerot Carteaud syndrome	100	99.6	ED51.Y
Granuloma annulare	89.36	97.84	EE80.0
Grover's disease	77.14	99.27	ED5Y
Guttate psoriasis	90.57	97.59	EA90.1
Haemangioma	89.09	97.3	XH5AW4
Hailey disease	92.68	99.51	EC20.2
Halo nevus	90.7	98.71	XH5971
Henoch Schonlein purpura	85.71	99.82	4A44.92
Herpes	85.19	97.7	1F00.Z
Herpes simplex	89.36	97.73	1F00.00
Herpes zoster	92.5	97.31	1E91.Z
Hidradenitis suppurativa	98.82	99.08	ED92.0
Hidrocystoma	88.24	99.34	EK70.3, XH7NR3
Histiocytoma	100	99.73	2F23.0, XH1UJ5
Ichthyosis	85.54	99.04	SB73
Impetigo	84.42	97.14	1B72
Infestation or insect bite	67.31	98.72	EK50.0
Intertrigo	84.62	99.49	EK02.20
Intradermal nevus	72.73	99.26	XH2MQ5, 2F20.0Y
Intraepidermal carcinoma	92.09	96.4	2E64.0Z
Junctional nevus	94.44	99.26	XH1M79
Juvenile xanthogranuloma	86.67	99.21	2B31.0
Keloid	85.71	99.48	EE60.0
Keloid acne	86.67	99.48	EE60.Y
Keratoacanthoma	91.67	95.6	2C31.1
Keratoderma palmaris et plantaris	94.44	99.4	EC20.32
Keratolysis exfoliativa	85.71	99.23	ED53
Keratosis palmaris et plantaris	92.86	99.66	EC20.32
Keratosis pilaris	88.1	99.5	ED56
Kerion	80.77	98.49	1F28.4
Koilonychia	88.89	99.72	EE10.0
Leishmaniasis	78.95	98.03	1F54
Lentigo	92	95.01	2F20.0Y, XH88L0, XH7B58, EJ20.1
Leprosy	89.38	98.93	1B20
Leukonychia	100	99.63	EC22.0, EE11.Y
Leukoplakia	71.43	99.35	DA01.00
Lichen amyloidosis	92.86	99.87	5D00.0
Lichen nitidus	92.59	99.23	EA92
Lichen planus	89.02	96.31	EA91
Lichen sclerosus	95.65	98.59	EB60
Lichen simplex chronicus	87.34	97.52	EA83.0Z
Lichen striatus	71.43	99.71	EA92
Linear IgA bullous dermatosis	83.33	99.64	EB42
Lipodermatosclerosis	90.91	99.61	BD74.2
Livedo reticularis	100	99.72	EG00, EF50, 6C45.2Z
Lyell's syndrome	92.86	99.58	EB13.1
Lymphangioma circunscriptum	100	99.37	LA90.10
Lymphocytic infiltration	93.75	99.12	EE90
Lymphocytoma	77.78	99.48	EE91
Lymphoma	62.5	99.63
Lymphomatoid papulosis	70	99.59	2B03.1, XH40C0
Malignant melanoma	91.29	91.26	XH4846
Melanocytic nevus	79.46	98	2F20
Melasma	92.31	99.34	ED60.1
Milia	84	99.48	EK70.Y
Miliaria	81.82	99.44	EE02
Molluscum contagiosum	88.62	98.3	1E76
Morphea	90	98.5	EB61
Mucous cyst	100	99.23	DA04.5
Mucous pemphigoid	100	99.76	EB41.1
Myiasis	80	99.81	1G01
Myxoid cyst	90	99.74	EK70.2
Necrobiosis lipoidica	86.54	99.08	EE80.1
Neurodermatitis	76.92	99.34	EA83.0Z
Neurofibromatosis	91.49	99.13	LD2D.1Z
Nevus	97.81	95.88	XH4L78
Nevus comedonicus	54.55	99.62	LC01
Nevus incipiens	92.31	98.61	2F20.0Y
Nevus of Ota	88.89	91.42	LC10
Nevus sebaceus of Jadassohn	90.32	99.33	LC02
Nevus spilus	84	99.44	XH40S8
Nodular prurigo	89.36	98.73	EC91.0
Non-specific lesion	99.44	97.02
Norwegian scabies	91.3	99.52	1G04.1
Onychodystrophy	84.62	99.08	EE10.5
Onychogryphosis	71.43	99.52	EE10.3
Onycholysis	95	99.18	EE10.2
Onychomycosis	96	98.15	EE12.1
Onychophagia	95	82.49	6B25.Y
Orf	100	99.58	1E75
Osler Weber Rendu disease	81.25	99.61	LA90.00
Palmoplantar psoriasis	96.08	97.62	EA90.42, EA90.5Y
Panniculitis	90	99.59	EF00
Papular acrodermatitis	79.31	99.41	EA12
Parapsoriasis	100	99.48	EA95, EK91.0
Paronychia	86.96	99.57	1F23.13, EE12.0, EE13.2, 1F00.0Y
Pearly penile papule	87.5	99.74	EK71.Z
Pellagra	94.12	99.55	5B5C.0
Pemphigoid gestationis	87.5	99.56	JA65.10
Pemphigus	72.73	98.5	EB40
Pemphigus foliaceus	67.74	98.86	EB40.1
Pemphigus vegetans	75	99.46	EB40.0Y
Perforating collagenosis	87.5	99.78	EE70.Y
Perifolliculitis capitis abscedens et suffodiens	100	99.88	ED70.51
Pigmentation	92.86	99.32	VV70
Pigmented purpura	92.86	99.53	EF40.0
Pincer nail syndrome	75	99.81	LD27.0Y
Pitted keratolysis	90.91	99.72	1C44
Pityriasis alba	100	99.73	EA88.4
Pityriasis lichenoides	83.33	99.16	EA93
Pityriasis rosea	89.11	97.83	EA10
Pityriasis rubra pilaris	77.14	99.05	EA94
Pityrosporum folliculitis	90.7	96.74	1F2D.1
Plane wart	90	99.03	1E81
Plantar wart	97.3	98.92	1E80.1
Plaque psoriasis	91.03	97.83	EA90.0
Polymorphic light eruption	60	99.26	EJ30.0
Porokeratosis	82.46	98.54	ED52
Pressure ulcer	97.97	98.85	EH90
Progressive symmetrical erythrokeratodermia	100	99.66	EC20.0Y
Pseudofolliculitis	90.48	99.72	ED9Y
Psoriasis	88.68	94.94	EA90
Purpura	73.33	99.58	EF40
Pustular psoriasis	91.67	97.73	EA90.4
Pyoderma gangrenosum	92.31	99.33	EB21
Pyogenic granuloma	91.89	98.01	2F26
Radiodermatitis	66.67	99.64	EJ7Z
Reiter's syndrome	100	99.46	FA11.2
Rhinophyma	93.55	99.08	ED90.02
Rosacea	95.19	94.6	ED90.0
Sarcoidosis	84.21	98.61	4B20
Scabies	73.15	98.34	1G04
Scalded skin syndrome	100	99.68	EA50.2
Scar	91.3	98.76	EH94
Schamberg's disease	97.22	99.29	EF40.0
Scleromyxedema	75	99.7	EB90.11
Sebaceous gland hyperplasia	83.87	98.81	ED91.1
Seborrheic dermatitis	89.14	97.77	EA81
Seborrheic keratosis	94.9	88.57	2F21.0
Sezary syndrome	75	99.61	2B02
Skin tag	92	98.77	EK71
Spitz nevus	93.18	98.35	XH2HG8
Squamous cell carcinoma	87	91.85	2C31, XH0945
Stasis dermatitis	95.2	98.79	EA86.0
Steatocystoma multiplex	100	99.64	2F22
Steroid acne	73.68	99.49	EH76.2
Striae	100	99.63	EE40.1
Stucco keratosis	92.59	98.96	2F21.Y
Syphilis	82.05	98.77	1A6Z
Syringoma	77.19	99.08	2F22, XH6325
Systemic scleroderma	71.43	99.8	4A42.Z
Telangiectasia	78.57	98.71	EF20.Z
Tinea	89.29	95.97	1F28.Z
Tinea nigra	96.43	99.74	1F2D.4
Tinea versicolor	94.24	98.41	1F2D.0
Tophus	75	99.73
Toxic epidermal necrolysis	100	99.83	EB13.1
Trichilemmal cyst	100	99.7	EK70.1
Tuberculoid leprosy	88.24	98.47	1B20.0
Tuberculosis	70	99.5
Tyloma (callus)	81.82	99.2	EH92.0Z
Urticaria	93.22	97.8	EB05
Urticaria pigmentosa	52.94	98.99	2A21.10, XH8VS0
Urticarial vasculitis	92.86	99.66	EF40.10
Varicella	89.74	98.82	1E90
Vascular lesion	78.57	99.69	XH8KN7
Vasculitis	88.89	98.88	4A44
Venous leg ulcer	97.92	99.24	BD74.3
Venous malformation	63.64	99.07	LA90.2, LC51, XH23S6
Vitiligo	89.25	96.3	ED63.0
Wart	93.96	95.77	4A00.Y
Wegener's granulomatosis	90.91	99.7	4A44.A1
Xanthoma	94.81	94.89	EB90.2
Xeroderma pigmentosum	85.71	99.88	LD27.1
Xerosis	90	99.79	ED54