<p>Sarcoidosis is a multisystem granulomatous disease with an unknown cause distinguished by the development of noncaseating granulomas in implicated organs. In patients with sarcoidosis, ventricular arrhythmias, and atrioventricular blocks can be deadly and result in sudden death. Clinically cardiac sarcoidosis affects five percent of sarcoidosis patients. Autopsy reports and imaging investigations, however, have revealed a higher frequency of cardiac involvement. Early detection of sarcoidosis through precise diagnosis, especially in cases with no evident symptoms, may reduce the patient’s mortality rate. This condition is primarily diagnosed via chest X-ray images. As a result, this research proposes a novel detection and classification approach for rapid diagnosis of sarcoidosis utilizing patient chest X-ray data. To diagnose sarcoidosis from chest X-ray images, we used state-of-the-art models like Inception and Residual Network-50, a handcrafted deep learning method, and a Quantum Support vector machine classifier. This study offers a convolutional neural network and histogram of oriented gradients method to aid medical experts in identifying sarcoidosis disease. The anisotropic diffusion filtering approach was used to improve image edge preservation, reduce noise, and augmentation to enhance the image. Gradient-weighted Class Activation Mapping was employed to illustrate the important activation areas that influenced the model’s decision. After evaluating the convolutional neural network model, it achieved 99.7% training accuracy and 98% test accuracy, while the histogram of oriented gradients achieved 100% training accuracy and 98% test accuracy.</p>