Cell Phone Based Remote Early Detection of Respiratory Disorders for Rural Children Using Modified Stethoscope

This work embodies the development of simple and robust technique for early detection of respiratory disorders in the children living in the rural areas of resource poor countries like India by acquisition of the lung sound recorded by cell phone based modified stethoscope by local health worker. The recorded file is transmitted along with case history to email address of health care center volunteer. The analysis of received lung sound file is done at health care centre. The analysis involves extraction of Mel frequency cepstral coefficients of lung sound. The coefficients and their statistical parameters are passed on to neural network for detection. The network was tested for 40 subjects containing equal mix of normal and bronchitis cases. The method detects the disease with an accuracy of 90% and with specificity of 95%. The overall accuracy of this method is 92.5%. The report (lung sound, analysis results and case history) is checked by medical practitioner at health-care centre on daily basis and necessary subscription is sent to designated rural local health care worker. Samsung cell phone model Galaxy Y was used for implementation. This work has great significance for evolving solutions towards rural health care of undeveloped and developing countries.