Analysis of Trichoscopic Images with Deep Neural Networks for Diagnosis and Activity Assessment of Alopecia Areata – A Retrospective Study

January 2026 in “ JDDG Journal der Deutschen Dermatologischen Gesellschaft ”

Raffaele Dante Caposiena Caro, V P Orlova, Nicola di Meo, Iris Zalaudek

Alopecia areata trichoscopic images deep-learning models neural networks scalp conditions hair regrowth

TLDR Deep-learning models can effectively diagnose and assess Alopecia areata using scalp images.

This study developed a deep-learning framework using trichoscopic images to diagnose and assess the activity of Alopecia areata (AA) with a dataset involving 152 participants. The two-stage model achieved an accuracy of 88.92% in distinguishing AA from other conditions and 83.33% in assessing AA activity levels. EfficientNetB0 excelled in distinguishing AA, while DenseNet169 performed best in assessing activity levels. Despite the model's effectiveness, limitations include challenges in diagnosing androgenetic alopecia (AGA) due to its low representation in the dataset. The study underscores the potential of AI in enhancing AA diagnosis and monitoring, though further research is needed to address dataset balance and equipment reliance.

View this study on onlinelibrary.wiley.com →

Discuss this study in the Community →

Research cited in this study

31 / 31 results

research European Expert Consensus Statement on the Systemic Treatment of Alopecia Areata

4 citations , January 2024 in “JEADV. Journal of the European Academy of Dermatology and Venereology/Journal of the European Academy of Dermatology and Venereology”

Baricitinib and ritlecitinib are recommended for severe alopecia areata, with other treatments available off-label.

research Multi-Class Support Vector Machine Classification for Detecting Alopecia Areata and Scalp Diseases

5 citations , October 2023 in “International Journal on Recent and Innovation Trends in Computing and Communication”

The method accurately detects and classifies scalp diseases, including alopecia areata, with 89.3% accuracy.

research Ensemble of Pre-Learned Deep Learning Model and an Optimized LSTM for Alopecia Areata Classification

7 citations , October 2023 in “Journal of Intelligent & Fuzzy Systems”

The new model improves Alopecia Areata classification accuracy to 93.1%.

research Two-Dimensional Planimetry for Alopecia Areata Severity Evaluation Compared With Severity of Alopecia Tool: A Pilot Study

August 2023 in “Skin Research and Technology”

Measuring bald patch size can help grade hair loss severity, with photograph-based evaluation being more reliable.

research AB-MTE Deep Classifier Trained with AAGAN for the Identification and Classification of Alopecia Areata

5 citations , June 2023 in “Engineering Technology & Applied Science Research”

The AI model accurately classifies Alopecia Areata with 96.94% accuracy.

research Trichoscopy Patterns in Alopecia Areata: A Systematic Review and Meta-Analysis

2 citations , June 2023 in “Skin Research and Technology”

Yellow dots and short vellus hairs are the most common signs of Alopecia Areata (AA), and trichoscopy can help diagnose AA and track treatment progress.

research Attention Balanced Multi-Dimension Multi-Task Deep Learning for Alopecia Recognition

May 2023 in “Indian journal of science and technology”

The new deep learning system can accurately recognize hair loss conditions with a 95.11% success rate.

research Prediction of Alopecia Areata Using CNN

15 citations , May 2023

A new CNN model can detect Alopecia Areata with 98% accuracy.

research Trichoscopy: An Update

8 citations , February 2023 in “Actas Dermo-Sifiliográficas”

Dermatologists should learn trichoscopy to better diagnose hair loss conditions.

research FRCNN Based Deep Learning for Identification and Classification of Alopecia Areata

9 citations , February 2023

The model accurately detects alopecia areata with 84.3% accuracy.

research Hair and Scalp Disease Detection Using Machine Learning and Image Processing

3 citations , January 2023 in “European Journal of Information Technologies and Computer Science”

The machine learning model accurately detected hair loss and scalp diseases using processed images.

research Automating Hair Loss Labels for Universally Scoring Alopecia from Images

1 citations , December 2022 in “JAMA Dermatology”

The AI system HairComb accurately scores hair loss severity, matching dermatologist assessments.

research Evaluation of Automated Measurement of Hair Density Using Deep Neural Networks

8 citations , January 2022 in “Sensors”

Deep learning can accurately automate hair density measurement, with YOLOv4 performing best.

research Deep Learning-Based Trichoscopic Image Analysis and Quantitative Model for Predicting Basic and Specific Classification in Male Androgenetic Alopecia

December 2021 in “Acta dermato-venereologica”

A deep learning model accurately predicts male hair loss types using scalp images.

research Update on Trichoscopy: Integration of Terminology by Systematic Approach and Proposal of a Diagnostic Flowchart

8 citations , November 2021 in “The Journal of Dermatology”

The article simplifies trichoscopy terms and offers a new diagnostic flowchart.

research Alopecia Areata: An Update on Etiopathogenesis, Diagnosis, and Management

290 citations , August 2021 in “Clinical Reviews in Allergy & Immunology”

JAK inhibitors show promise for treating alopecia areata, but more research is needed.

research Classification Framework for Healthy Hair and Alopecia Areata: A Machine Learning Approach

8 citations , August 2021 in “Computational and Mathematical Methods in Medicine”

Machine learning can accurately identify Alopecia Areata, aiding in early detection and treatment of this hair loss condition.

research Trichoscopy of Alopecia Areata: Hair Loss Feature Extraction and Computation Using Grid Line Selection and Eigenvalue

10 citations , September 2020 in “Computational and Mathematical Methods in Medicine”

Researchers developed an algorithm for self-diagnosing scalp conditions with high accuracy using smart device-attached microscopes.

research Pre-Trained Classification of Scalp Conditions Using Image Processing

15 citations , August 2020 in “Indonesian Journal of Electrical Engineering and Computer Science”

The system can automatically classify scalp conditions with 85% accuracy.

research Fibrosing Alopecia in a Pattern Distribution

30 citations , January 2020 in “Journal of The American Academy of Dermatology”

Fibrosing alopecia in a pattern distribution is a hair loss condition often confused with other types, requiring early treatment but usually not resulting in significant hair regrowth.

research Alopecia Areata Predictive Score: A New Trichoscopy-Based Tool to Predict Treatment Outcome in Patients with Patchy Alopecia Areata

19 citations , July 2019 in “Journal of Cosmetic Dermatology”

A new tool can predict hair regrowth in patchy alopecia areata patients.

research Trichoscopy Tips

28 citations , August 2018 in “Dermatologic Clinics”

Trichoscopy is a useful tool for diagnosing hair and scalp issues, but doctors must be careful to avoid mistakes.

research Dynamic Trichoscopy: Advancements in Hair and Scalp Examination

14 citations , June 2018 in “JAMA Dermatology”

Trichoscopy is useful for diagnosing and monitoring hair and scalp conditions over time.

research Trichoscopy of Alopecia Areata: An Update

89 citations , March 2018 in “The Journal of Dermatology”

Trichoscopy helps diagnose and monitor alopecia areata by looking at a combination of specific hair and scalp features.

research Nail Changes in Alopecia Areata: An Update and Review

45 citations , January 2018 in “International Journal of Dermatology”

Nail problems are common in people with alopecia areata, often leading to cosmetic and functional issues, but more research is needed for treatment guidelines.

research Quantifying Alopecia Areata via Texture Analysis to Automate the SALT Score Computation

20 citations , December 2017 in “Journal of Investigative Dermatology Symposium Proceedings”

Researchers created a fast, accurate computer program to measure hair loss in alopecia areata patients.

research Evaluation and Diagnosis of the Hair Loss Patient

69 citations , August 2014 in “Journal of The American Academy of Dermatology”

Trichoscopy is a quick, cost-effective tool for diagnosing different hair loss conditions.

research Trichoscopy

42 citations , October 2012 in “Dermatologic Clinics”

Trichoscopy is a useful tool for diagnosing different hair and scalp diseases without surgery.

research Hair and Scalp Dermatoscopy

245 citations , March 2012 in “Journal of The American Academy of Dermatology”

Dermatoscopy is useful for identifying different hair and scalp conditions and can reduce the need for biopsies.

research Trichoscopy for Common Hair Loss Diseases: Algorithmic Method for Diagnosis

89 citations , December 2010 in “The Journal of Dermatology”

The conclusion is that an algorithm using trichoscopy helps diagnose different types of hair loss but may need updates and a biopsy if results are unclear.

research TrichoScan: A Novel Tool for the Analysis of Hair Growth In Vivo

70 citations , June 2003 in “Journal of Investigative Dermatology Symposium Proceedings”

TrichoScan is a reliable method for measuring hair growth and is useful for assessing hair loss treatments.