Multicenter retrospective study of digital histopathology for automated classification of pediatric sarcoma

Pediatric sarcoma’s are rare and diverse tumors that can form in various types of soft tissue. Accurate classification of sarcoma subtypes in an important step to guide and optimize treatment. A multicenter retrospective study, presented today at the AACR Annual Meeting in Chicago, examined the potential of artificial intelligence to accurately identify pediatric sarcoma subtypes. The AI-driven models could correctly distinguish between Ewing sarcoma and other sarcoma types in 92.2% of cases; non-rhabdomyosarcoma and rhabdomyosarcoma soft tissue sarcoma in 93.8% of cases; alveolar rhabdomyosarcoma and embryonal rhabdomyosarcoma in 95.1% of cases; and alveolar rhabdomyosarcoma, embryonal rhabdomyosarcoma, and spindle cell rhabdomyosarcoma in 87.3% of cases. The authors conclude that digital pathology enables accurate automated classification of pediatric sarcomas.