Green Innovations in Medical Imaging: A Pathway to Eco-Friendly Diagnostics
Nurul Huda(a*)- Deden Istiawan(b)- Ika Safitri Windiarti(c)-

a,b) Faculty of Science and Technology,Institut Teknologi Statistika dan Bisnis Muhammadiyah Semarang
Jl Prof Dr Hamka KM 1 Ngaliyan Semarang
nurul.huda[at]itesa.ac.id
c) Department oof Information Technology, Universiti Muhammadiyah Malaysia.

Abstract

1. Introduction:The carbon footprint of medical imaging varies widely across modalities- for instance, echocardiograms emit 2 kg of CO2 equivalent, while 3 Tesla MRI emits 200-300 kg. Abdominal CT produces 6 kg, and abdominal MRI produces 20 kg of CO2 equivalent per examination. In 2016, MRI and CT emissions accounted for 0.77% of global emissions across 120 countries, and medical imaging contributes approximately 1% to the 5-10% carbon footprint of global healthcare.

2. Purpose : Efforts can be made to promote climate-neutral choices in medical imaging. AI can assist in optimizing imaging protocols, reducing the need for excessive or unnecessary imaging tests, and minimizing the associated carbon footprint. By improving the accuracy of image analysis and diagnosis, AI can help avoid repeat imaging tests, reducing both costs and environmental impact
3. Method: Deep learning techniques, specifically Convolutional Neural Networks (CNN), are employed for image analysis and identification, with the model trained on the compiled dataset. The study then turns its focus to assessing the environmental impact by quantifying the carbon emissions associated with the CNN-based medical image identification process

4. Main Finding: The research aims to provide insights into the dual aspects of accuracy in medical image identification and the environmental implications of employing deep learning techniques on specific hardware configurations
5. Implication: This research finds that CNN models used for identifying and analyzing medical images produced carbon emissions average 0.00868 kg.

Keywords: Green technology, medical imaging, eco-friendly diagnostics, sustainability, carbon footprint.

Topic: Medicine & Health Sciences