A novel artificial intelligence (AI)-based model could accurately classify sarcomas among pediatric patients using digital pathology images alone, according to new findings presented by Thiesen et al at the 2025 American Association for Cancer Research (AACR) Annual Meeting (Abstract 2423/8).
Background
Sarcomas are rare and diverse tumors that can form in various types of soft tissue—including muscle, tendons, fat, blood or lymphatic vessels, nerves, or the tissue surrounding joints. The disease is classified into subtypes based on the tissue of origin and various molecular features.
“Accurate classification of a patient’s sarcoma subtype is an important step that helps guide and optimize treatment,” explained lead study author Adam Thiesen, MS, an MD/PhD candidate at the University of Connecticut Health and The Jackson Laboratory. “Unfortunately, the heterogeneity of sarcomas makes them particularly difficult to classify, often requiring complex molecular and genetic testing as well as external review by highly specialized pathologists who use pattern recognition skills honed through years of training to arrive at a diagnosis—resources that are not readily available in many health-care settings,” he added.
Study Methods and Results
In the study, researchers used 691 digital images of pathology slides, representing nine sarcoma subtypes, to train AI algorithms to recognize patterns associated with each subtype.
“By digitizing tissue pathology slides, we translated the visual data a pathologist normally studies into numerical data that a computer can analyze,” Mr. Thiesen explained. “Much like our cell phones can recognize a person’s face in photos and automatically generate an album of photos of that person, our AI-based models recognize certain tumor morphology patterns in the digitized slides and group them into diagnostic categories associated with specific sarcoma subtypes,” he noted.
The researchers developed and applied open-source software to harmonize the images collected from different institutions to account for variation in format, staining, and magnification. The harmonized images were then converted into small tiles before being fed into deep learning models that extracted numerical data for analysis by a novel statistical method—which generated summaries of each slide’s features that were evaluated by the trained AI algorithms to categorize each slide as a specific sarcoma subtype.
In validation experiments, the AI algorithms identified sarcoma subtypes with high accuracy. Notably, the AI-driven models correctly distinguished between:
- Ewing sarcoma and other sarcoma types in 92.2% of cases
- nonrhabdomyosarcoma soft tissue sarcomas and rhabdomyosarcoma soft tissue sarcomas in 93.8% of cases
- alveolar rhabdomyosarcoma and embryonal rhabdomyosarcoma in 95.1% of cases
- alveolar rhabdomyosarcoma, embryonal rhabdomyosarcoma, and spindle cell rhabdomyosarcoma in 87.3% of cases.
Conclusions
“Our findings demonstrate that AI-based models can accurately diagnose various subtypes of sarcoma [in pediatric patients] using only routine pathology images. This AI-driven model could help provide more pediatric patients access to quick, streamlined, and highly accurate cancer diagnoses regardless of their geographic location or health-care setting,” highlighted Mr. Thiesen. “Our models are built in such a way that new images can be added and trained with minimal computational equipment. After the standard data processing, clinicians could theoretically use our models on their own laptops, which could vastly increase accessibility even in under-resourced settings,” he emphasized.
A limitation of the study was that the number of available pathology images was smaller than the researchers would have wanted for training AI algorithms. However, given the rarity of sarcomas among pediatric patients, their imaging data set may be among the largest multicenter collections of these cancer types to date—representing multiple subtypes, anatomical locations, and patient demographics.
“We hope that, over time, additional groups will work with us to further increase the size of this data set,” Mr. Thiesen concluded.
Disclosure: The research in this study was supported by the National Institutes of Health, The Jackson Laboratory, and Hartford Hospital. For full disclosures of the study authors, visit abstractsonline.com.
