BackDermatology

Advancements in psoriasis classification using custom transfer learning algorithms

Nature Scientific ReportsResearch Authors: L. Lakshmi, K. Dhana Sree Devi, KongaraSrinivasa Rao, M. Priyadharshini, V. Murugesh & Oleg RybinAIIM Authors: Shareen Chahal, Josh BronteApproved by President Reda RiffiPublication Date: 2/2/2026

Comprehensive Summary

Lakshmi et al. performed a study with the purpose of helping combat psoriasis, a prevalent skin disease, by creating a dataset for artificial intelligence (AI) algorithms to classify and diagnose the disease from images. Using ResNet50, InceptionResNetV2, and Inception V3, all three optimizers showed high accuracy during testing. Compared to previous works established in the study’s literature review, Inception V3 appeared to perform better than what has been done before. The high performance of the model serves as a promising feat that classifies lesions specifically and accurately.

Outcomes and Implications

Overall, the findings of Lakshmi et al. will assist in the future of psoriasis detection and treatment, as the researched models can more efficiently and accurately analyze images. The pressure on dermatologists may lighten as well, though a professional’s opinion would always more accurately reflect a proper diagnosis.

Our mission is to

Connect medicine with AI innovation.

No spam. Only the latest AI breakthroughs, simplified and relevant to your field.