Page 56 - AI for Good Innovate for Impact
P. 56

AI for Good Innovate for Impact



                      a broader populace. This technology can democratize healthcare solutions, making them
                      affordable and personalized for citizens across all ages, genders, economic strata, and social
                      backgrounds. By providing accessible and mobile-based AI solutions, public health education
                      in yoga care can be scaled to a population level, thereby alleviating the burden on hospitals
                      and optimizing the utilization of human and budgetary resources. Crucially, AI-powered yoga
                      applications will play a vital role in controlling non-communicable diseases (NCDs) such as
                      diabetes and cardiovascular conditions with unprecedented speed and scale, offering a cost-
                      effective solution particularly beneficial for developing countries. The availability of data-driven
                      health monitoring will also serve as a strong motivator for a new generation of yoga practitioners,
                      while governments can leverage this data to design and implement promotive healthcare
                      strategies and incentives to prevent diseases from reaching epidemic proportions [4]. Beyond
                      individual practice, AI-based yoga solutions will stimulate data-driven research, fostering the
                      development of disease-specific mobile applications that doctors can confidently recommend
                      as part of their prescriptions. The ability to track yoga pose improvement and correlate it
                      with disease lifecycle monitoring holds the promise of reducing the need for medication and
                      surgical interventions. Furthermore, this AI tool can be instrumental in objectively evaluating
                      yoga instructors and can serve as a valuable resource for addressing mental health issues within
                      educational institutions, as well as in professional sectors such as the military and police forces.
                      Ultimately, by popularizing yoga through accessible AI tools, we can significantly enhance
                      society's resilience and ability to manage future pandemic situations.


                      Use Case Status: Active, with ongoing deployment and iterative refinement�

                      Partners: Ministry of AYUSH [6], yoga & wellness industry, medical practitioners, and academic
                      institutes in the health domain.

                      2�2     Benefit of use case


                      AI can help bring affordable Healthcare solutions and ensure Good Health of all citizens – both
                      on the curative and preventive side. Personalized healthcare solutions will become available to
                      one and all. With availability of AI and Mobile based solution, Public health education in Yoga
                      care can be implemented at population scale, reducing the load on hospitals. This AI tool can
                      also be utilized for conducting examination of Yoga instructors in an objective manner, as well
                      as a solution in mental health issues at school, college level as well as at organization level in
                      sectors like military, police etc.

                      With popularization of Yoga, the ability to handle pandemic situations in the future will improve
                      in the future.


                      2�3     Future Work

                      •    Future work intends to expand the dataset to widen the scope of yoga poses and include
                           detailed information about each. A mobile app can also be developed for ease of use by
                           yoga practitioners.
                      •    The development of standards is paramount.  'Landmarks' essential for the determination
                           of yoga poses ought to be standardised from an interoperability perspective. These
                           landmarks may subsequently be converted back into an Avatar-based yoga instructor.
                           Initially, keypoints are extracted to delineate landmarks, which are then transmitted as
                           input to the classifier model for pose prediction and accuracy estimation. Should the
                           accuracy exceed seventy per cent, a timer shall commence counting the duration of the




                  20
   51   52   53   54   55   56   57   58   59   60   61