Kalibrasi Alat Ukur Hemoglobin Non-Invasive

Abstract

Di seluruh dunia prosentase penderita anemia, yaitu suatu keadaan rendahnya kadar hemoglobin dalam darah, masih tinggi. Oleh karena itu alat ukur hemoglobin sangat diperlukan. Namun sampai sekarang kebanyakan alat ukur hemoglobin masih menggunakan metoda invasive yang mengambil sampel darah dengan melukai pasien. Walaupun memiliki ketelitian yang sangat tinggi namun metoda invasive bisa menyebabkan ketidaknyamanan atau trauma bagi sebagian orang. Oleh karena itu dibutuhkan alat ukur hemoglobin non-invasive yang mudah digunakan dan hasilnya bisa diandalkan. Penelitian ini melakukan kalibrasi sekaligus pengembangan terhadap alat ukur hemoglobin non-invasive hasil penelitan sebelumnya. Alat ini bekerja menggunakan prinsip reflektansi dimana LED pada probe memancarkan cahaya yang menembus kulit hingga mencapai pembuluh vena yang berisi darah yang mengandung hemoglobin. Pantulan cahaya dari vena diterima sensor photodiode, yang kemudian dikonversi menjadi data digital oleh ADS1115 dan disimpan dalam bentuk database di dalam Raspberry Pi. Data kemudian diolah menggunakan regresi linear berganda untuk kalibrasi dengan data pengukuran hemoglobin metoda SLS sebagai ground truth sehingga didapatkan persamaan linear berganda. Nilai Root Mean Squared Error (RMSE) pada tahap kalibrasi didapatkan sebesar 1,332, yang berarti memenuhi syarat karena berdasarkan Jurnal Biomedical and Health Informatics persyaratan alat ukur hemoglobin nilai maksimal RMSE sebesar 1,352. Selanjutnya persamaan ini kemudian dimasukkan pada alat untuk proses validasi. Hasil validasi yang mendapatkan nilai RSME sebesar 1,264 yang berarti cukup rendah, atau alatnya cukup teliti.

Throughout the world the percentage of people with anemia, which is a condition of low levels of hemoglobin in the blood, is still high. Therefore a hemoglobin measuring tool is needed. However, until now most hemoglobin measuring tools still use invasive methods which take blood samples by injuring the patient. Even though it has very high accuracy, the invasive method can cause discomfort or trauma for some people. Therefore, a non-invasive hemoglobin measuring tool is needed that is easy to use and has reliable results. This study calibrated as well as developed a non-invasive hemoglobin measuring tools resulting from previous research. This tool works using the principle of reflectance where the LED on the probe emits light that penetrates the skin until it reaches the veins which contain blood containing hemoglobin. The light reflected from the veins is received by the photodiode sensor, which is then converted into digital data by the ADS1115 and stored in a database on the Raspberry Pi. The data was then processed using multiple linear regression for calibration with hemoglobin measurement data using the SLS method as the ground truth so that a multiple linear equation was obtained. The Root Mean Squared Error (RMSE) value at the calibration stage was 1.332, which means that it meets the requirements because based on the Biomedical and Health Informatics Journal the maximum RMSE value for the hemoglobin measuring instrument is 1.352. Furthermore, this equation is then entered into the tool for the validation process. The validation results obtained an RSME value of 1,264 which means it is quite low, or the tool is quite thorough.