Predicting BRAF Mutations in Melanoma Using Neural Networks
Author Information
Author(s): Dudin Oleksandr, Mintser Ozar, Gurianov Vitalii, Kobyliak Nazarii, Kaminskyi Dmytro, Matvieieva Alina, Shabalkov Roman, Mashukov Artem, Sulaieva Oksana
Primary Institution: Medical Laboratory CSD, Kyiv, Ukraine
Hypothesis
Can a model be developed to predict BRAF mutations in cutaneous melanoma using clinical and histological data?
Conclusion
The developed model can predict BRAF mutation status in melanoma patients, aiding in personalized treatment decisions.
Supporting Evidence
- BRAF mutations were found in 60.5% of the patients studied.
- The model achieved a sensitivity of 89.4% and specificity of 50.7%.
- Age and tumor location were significant predictors of BRAF mutation status.
- The V600E mutation was the most common subtype found in the cohort.
Takeaway
Researchers created a computer model that helps doctors figure out if a melanoma patient has a specific gene change called a BRAF mutation, which can help in choosing the right treatment.
Methodology
The study analyzed clinical and histological data from 2041 melanoma patients to develop a neural network model for predicting BRAF mutations.
Limitations
The study only assessed a subset of BRAF V600 variants and did not consider other genetic alterations.
Participant Demographics
The cohort included 991 males and 1050 females, with an average age of 54.2 years.
Statistical Information
P-Value
p<0.001
Confidence Interval
95% CI: 0.74–0.84
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website