18367546. METHOD AND SYSTEM FOR EVALUATING CLINICAL EFFICACY OF MULTI-LABEL MULTI-CLASS COMPUTATIONAL DIAGNOSTIC MODELS simplified abstract (Tata Consultancy Services Limited)
METHOD AND SYSTEM FOR EVALUATING CLINICAL EFFICACY OF MULTI-LABEL MULTI-CLASS COMPUTATIONAL DIAGNOSTIC MODELS
Organization Name
Tata Consultancy Services Limited
Inventor(s)
Sai Chander Racha of Hyderabad (IN)
Sundeep Khandelwal of Noida (IN)
Soumadeep Saha of Kolkata (IN)
METHOD AND SYSTEM FOR EVALUATING CLINICAL EFFICACY OF MULTI-LABEL MULTI-CLASS COMPUTATIONAL DIAGNOSTIC MODELS - A simplified explanation of the abstract
This abstract first appeared for US patent application 18367546 titled 'METHOD AND SYSTEM FOR EVALUATING CLINICAL EFFICACY OF MULTI-LABEL MULTI-CLASS COMPUTATIONAL DIAGNOSTIC MODELS
Simplified Explanation
The present invention provides a method and system for evaluating clinical efficacy of multi-label multi-class computational diagnostic models.
- The method involves obtaining a diagnosis for a given dataset of diagnostic samples from the diagnostic model, classifying the diagnosis as wrong, missed, over, or right, and calculating penalties based on these classifications.
- A contradiction matrix is used to calculate a second penalty for each diagnostic sample, which is then summed up with the first penalty to compute a pre-score for each sample.
- The diagnostic model is evaluated using a metric based on the sum of pre-scores, as well as scores from a perfect and a null multi-label multi-class computational diagnostic model.
---
- Potential Applications
This technology can be applied in the field of healthcare for evaluating the accuracy and efficacy of diagnostic models in clinical settings.
- Problems Solved
This technology addresses the limitations of conventional metrics in evaluating clinical diagnostic models by considering context-dependent clinical principles and capturing important features that should be present in a diagnostic model.
- Benefits
- Improved evaluation of diagnostic models in clinical practice - Enhanced accuracy and efficacy of multi-label multi-class computational diagnostic models - Better understanding of diagnostic performance in healthcare settings
- Potential Commercial Applications
- Healthcare institutions and clinics can utilize this technology to assess the performance of their diagnostic models and improve patient care outcomes.
- Possible Prior Art
There may be prior art related to evaluating diagnostic models in healthcare using different metrics or methodologies. Research papers or patents in the field of medical informatics or machine learning could potentially be relevant.
---
- Unanswered Questions
- How does this technology compare to existing methods for evaluating diagnostic models in clinical practice?
This article does not provide a direct comparison with existing methods or metrics commonly used in the evaluation of diagnostic models.
- What are the specific clinical scenarios or diseases where this technology has been tested or validated?
The article does not mention specific clinical scenarios or diseases where this technology has been applied or tested for validation.
Original Abstract Submitted
The present invention relates to the field of evaluating clinical diagnostic models. Conventional metrics does not consider context dependent clinical principles and is unable to capture critically important features that ought to be present in a diagnostic model. Thus, present disclosure provides a method and system for evaluating clinical efficacy of multi-label multi-class computational diagnostic models. Diagnosis for a given dataset of diagnostic samples is obtained from the diagnostic model which is then classified as wrong, missed, over or right diagnosis, based on which a first penalty is calculated. A second penalty is calculated for each diagnostic sample using a contradiction matrix. The first and second penalties are summed up to compute a pre-score for each diagnostic sample. Finally, the diagnostic model is evaluated using a metric that is based on sum of pre-scores, and scores from a perfect and a null multi-label multi-class computational diagnostic model.
