Docutis

An open-source, structured dermatology reference and education toolkit.

View the live application

About

Docutis is an open-source educational and professional reference. It presents dermatology information in a searchable, structured format for physicians, medical trainees and other healthcare professionals. It is not a substitute for individual care, diagnosis, treatment or a current guideline.

Maturity: active development, and a public preview / pre-release. Some assets have a published clinician review. Most content is still review required. This site is not production-ready, not clinically validated decision support, and not a statement that every record has been clinician reviewed.

Live site: https://mnasuharas.github.io/Docutis/

Clinical review is version-bound and independently tracked for disease records, quiz items, visual assets, follow-up protocols and case-based learning units. The public dashboard and review-status.json distinguish human review from automated validation. AI interpretation is not clinician review. Release notes for older tags are point-in-time and may not match current main.

Release state: Goals 18 to 28 on main and the live site

• Previous preview (tag v0.2.0-preview.1, commit c28f1de): 50 condition records, the five pilot cases, and 28 public review units, of which 5 are clinician reviewed and 23 are review required.
• Current main and the live site (Goals 18 to 28, developed in stacked pull requests #24 to #30 and #32 to #35 and released through one integration pull request, #36; no new version tag): 51 cases, 22 true clinical-dermoscopic pairs, and 74 review units, of which 5 are clinician reviewed and 69 are review required. Publication is not clinical approval. Clinical review is deferred: no Goal 18 to 27 case or pattern is clinician reviewed, and no physician decision was created for them. Goal 28 adds no case and no review unit.

Goal 25 adds six histopathology-labelled clinical and dermoscopic pairs from CC BY 4.0 open-access articles: 4 histopathology-confirmed true paired melanoma cases (melanoma in situ, amelanotic nodular melanoma, acral melanoma in situ, and lentigo maligna) and 2 benign mimics that share a feature with a melanoma case (a nevus with cytologic atypia and a facial solar lentigo). Three confirmations come from a figure-caption sentence about that lesion (melanoma in situ, nevus with cytologic atypia, nodular melanoma) and three from a study-level methods statement (acral melanoma in situ, lentigo maligna, solar lentigo); each case record names its basis in histopathologySource. Panels were cropped from composite figures with recorded pixel boxes and hashes. Every candidate checked, including 31 Goal 25 candidates, is in acquisition-ledger.js; rejected rows never become cases.

Goal 26 adds an acral and nail-unit contrast set of seven clinical and dermoscopic pairs from CC BY 4.0 open-access articles: a heel melanoma and a heel nevus from one figure, an acral melanoma in situ with an irregular stroke pattern, an acral subcorneal haematoma, two childhood nail-band nevi, and an adult great-toenail melanoma. Five have a histopathology statement (three study-level, two in the case text); the haematoma and the nail melanoma stay at a clinical or caption diagnosis. No Hutchinson sign, migration, or change is claimed from a single frame. The ledger records 24 Goal 26 candidates.

Goal 27 adds six clinical and dermoscopic pairs from CC BY 4.0 open-access articles to repeat the main acral and nail discriminators in new patients: two benign sole cases with a parallel furrow pattern (a cluster of nevi with histopathology from a later biopsy, and a nevus verified by five years of stability), a palm melanoma in situ that shows a furrow pattern beside an eccentric blotch, an independent second parallel ridge example (a heel melanoma in situ), the first adult benign nail band (a nail matrix nevus), and the first nail haemorrhage with onychoscopy. Four have a lesion-specific histopathology statement; the stable nevus and the haemorrhage do not. No histopathology-confirmed nail melanoma with onychoscopy was found under an allowed license. One dermoscopic structure was added (parallel furrow pattern). The ledger records 40 Goal 27 candidates. node scripts/paired-modality.js now separates positive observations from independent positive cases for each structure.

Goal 28 adds mixed-case screening training: unknown lesions from the existing library, one at a time, with observation, a working impression, staged dermoscopy, a deliberate reveal of the source diagnosis and its verification, stored mimic and contrast teaching, and a qualitative debrief. It adds no case and no clinical claim. Of 51 cases, 39 are core training, 12 context only (including the caption-only nail melanoma and both nail haemorrhages), and 0 excluded. Sessions are curated for teaching coverage, not prevalence. There is no score, probability, disposition, or pass mark, and choices stay in the open tab. See TRAINING.md and TRAINING_ELIGIBILITY.md.

This is the beginning of the README. Read the rest at the original. Read the full README ↗