Researchers at Johns Hopkins Medicine have developed a machine learning model that accurately calculates the extent of tumor death in bone cancer patients. The model, trained on annotated pathology images, achieved 85% accuracy, which rose to 99% after removing an outlier. The innovative method reduces the workload for pathologists and has the potential to revolutionize the evaluation of osteosarcoma treatment outcomes.
Researchers from Johns Hopkins Medicine Developed a Machine Learning Model for Precise Osteosarcoma Necrosis Calculation
In the field of oncology, accurately assessing the effectiveness of chemotherapy on bone cancer patients is crucial for determining prognosis. A team of researchers at Johns Hopkins Medicine has made a groundbreaking advancement in this area. They have successfully developed and trained a machine learning model to calculate percent necrosis (PN), a key metric that indicates the extent of tumor death in patients with osteosarcoma. This innovative model achieves an impressive 85% accuracy compared to results obtained by a musculoskeletal pathologist. By removing a single outlier, accuracy soars to an astonishing 99%.
Traditionally, calculating PN has been a labor-intensive process that relies on extensive annotation data from musculoskeletal pathologists. It also suffers from low interobserver reliability, where two pathologists analyzing the same whole-slide images (WSIs) may arrive at different conclusions. Recognizing these challenges, the researchers identified the need for an alternative approach.
The team developed a weakly supervised machine learning model that requires minimal annotation data for training. This innovative methodology means that a musculoskeletal pathologist using the model for PN calculation would only need to provide partially annotated WSIs, significantly reducing their workload.
To construct this model, the team curated a comprehensive dataset from the pathology archives of Johns Hopkins’ renowned U.S. tertiary cancer center. The dataset exclusively included cases of intramedullary osteosarcoma in patients who underwent both chemotherapy and surgery at the center between 2011 and 2021.
A musculoskeletal pathologist carefully annotated three different tissue types on each collected WSI: active tumor, necrotic tumor, and non-tumor tissue. The pathologist also estimated the PN for each patient. With this valuable information, the team began the training phase.
The researchers explained the training process. They decided to train the model by teaching it to recognize image patterns. The WSIs were divided into thousands of small patches and grouped based on the pathologist’s labels. These grouped patches were then used to train the model. This approach was chosen to provide the model with a more robust frame of reference, avoiding potential oversights that could occur by solely using one large WSI.
After training, the model and the musculoskeletal pathologist evaluated two osteosarcoma patients using six WSIs. The results were remarkable, with an 85% positive correlation between the model’s PN calculations and tissue labeling compared to the pathologist’s findings. The only issue arose from occasional difficulties in properly identifying cartilage tissue, which led to an outlier due to an abundance of cartilage on one WSI. After removing this outlier, the correlation skyrocketed to an impressive 99%.
Practical Solutions and Value
If you want to evolve your company with AI and stay competitive, consider the machine learning model developed by researchers from Johns Hopkins Medicine for precise osteosarcoma necrosis calculation. This AI solution can revolutionize the evaluation of osteosarcoma treatment outcomes, providing accurate and efficient assessment of chemotherapy effectiveness.
To implement AI in your company, follow these practical steps:
- Identify Automation Opportunities: Locate key customer interaction points that can benefit from AI.
- Define KPIs: Ensure your AI endeavors have measurable impacts on business outcomes.
- Select an AI Solution: Choose tools that align with your needs and provide customization.
- Implement Gradually: Start with a pilot, gather data, and expand AI usage judiciously.
For AI KPI management advice and continuous insights into leveraging AI, connect with us at hello@itinai.com. Explore our AI Sales Bot at itinai.com/aisalesbot, designed to automate customer engagement and manage interactions across all customer journey stages.
