AI tool for detecting brain and spinal cord tumor| Explained
While speaking to India Today, IIT Madras Research Professors Medha Pandey and M. Michael Gromiha discussed various facets of the approach, the accuracy, and the use of AI in developing their machine learning tool that detects tumors in the human brain and spinal cord.
By Megha Chaturvedi: A widely used method for developing AI tools for tumor detection has been introduced by IIT Madras. The AI tool trains the machine learning model on a large dataset of mutations that have been labeled either as positive ('Driver'), meaning they cause tumors, or negative ('Passenger'), meaning they do not cause tumors.
The various characteristics and patterns introduced by IIT Madras researchers have brought different angles when discussed with India Today. The discussion was on the machine learning tool that detects tumors in the human brain and spinal cord.
AI TOOL FOR DETECTING BRAIN AND SPINAL CORD TUMOR: KNOW MORE
The machine learning model is trained to recognise patterns and characteristics in protein sequences linked with the dataset. Once trained, the model can be tested on a separate dataset with an unknown set of mutations to evaluate its performance.
If successful in the test dataset, the model can analyze new mutations to identify potential drivers. Building AI tools for tumor identification requires a combination of machine-learning approaches, biological understanding, and access to large clinical datasets.
These tools' accuracy and effectiveness depend on the quality and diversity of training data and the complexity of the machine learning model used for development and deployment.
WHAT IS THE ACCURACY COMPARED TO TRADITIONAL METHODS?
The accuracy and effectiveness of AI tools for tumor detection rely on the quality and quantity of data used. Our specific method considers the impact of neighboring residues and secondary structure info to extract features for each mutation, achieving 80% accuracy when classifying disease-causing and neutral mutations.
AI TOOL FOR DETECTING BRAIN AND SPINAL CORD TUMOR: HOW IT WORKS
The AI tools include very input features like amino acid sequence, mutations, physicochemical properties, and conservation of output on cancer-causing or neutral mutations. Once the method is developed on a web server named GBMDriver, it includes different machine-learning models trained on mutation data associated with brain tumors.
By supplying the UniProt ID and mutation of interest, a new user can determine the impact of any unknown mutation. The gained information will be helpful to design new therapeutic strategies for glioblastoma.
AI TOOL FOR DETECTING BRAIN AND SPINAL CORD TUMOR: LIMITATIONS
Because of advances in structural bioinformatics and the introduction of tools like AlphaFold, it is now possible to generate structure-based parameters. The combination of sequence and structural characteristics aids in the improvement of the approach, as does the inclusion of additional data.
AI TOOL FOR DETECTING BRAIN AND SPINAL CORD TUMOR: USAGE IN CLINICAL PRACTICE
The developed server may be used to discover probable driver mutations in brain tumors, allowing for the development of more advanced therapy techniques. If the driver mutations are detected at the right time, it can help improve patient survival rates.
AI TOOL FOR DETECTING BRAIN AND SPINAL CORD TUMOR: CHALLENGES
The main challenge in biological analytics is to preprocess data and extract important features. The extraction and preprocessed mutations from the COSMIC database are associated with glioblastoma, but finding the best feature combinations remains difficult.
WHAT TYPE OF BRAIN TUMOR DOES IT DETECT?
There are four phases of a brain tumor. The mutations linked with Glioblastoma Grade IV were the primary focus of the current investigation.
AI TOOL FOR DETECTING BRAIN AND SPINAL CORD TUMOR: FUTURE OBJECTIVES
AI is a new discipline that deals with massive amounts of data. The right application of AI approaches will undoubtedly reduce the time complexity and efficiency in identifying mutations unique to cancer kinds or tumor identification. The availability of a massive amount of data boosts dependability.
AI TOOL FOR DETECTING BRAIN AND SPINAL CORD TUMOR: MEDICAL DIAGNOSIS AND TREATMENT
The developed technique is an AI-driven strategy for prioritising driver mutations. These newly discovered mutations may be tested in experimental labs for confirmation.
AI TOOL FOR DETECTING BRAIN AND SPINAL CORD TUMOR: IMPACT ON NEUROSURGERY AND MEDICAL COMMUNITY
The development of AI tools is expected to have a significant impact on the field of neurosurgery and the broader medical community. The availability of experimental data is a determining factor for finding better results.
