<p>Parkinson’s disease (PD) is an age-related neurodegenerative disorder. Symptoms of PD develop slowly, sustain, and always deteriorate over time. These might create a feeling of social abandonment among the patients. Therefore, early detection and monitoring for the progression of PD are vital. Dopamine active transporter (DaTscan) is an imaging technology that has shown an essential feature concerning detecting and predicting neurodegenerative disorders, especially PD. The image analysis of the DaTscan is often used to monitor the severity of PD symptoms. DaTscan images can also detect any change in dopamine levels in its early stages, which helps it intervene and prevent it from going into uncontrollable stages. A novel model was proposed for the accurate detection of symptoms of PD using DaTscan images through a modified VGG19 (MVGG19) model. A group normalization layer is added after each pooling layer while global average pooling layers replace the fully connected layers. The model was trained on 1098 images of a DaTscan taken from the Parkinson’s Progression Markers Initiative (PPMI): 361 images of healthy patients, 364 images of patients with an early-stage disease, and 373 images of patients with an advanced-stage disease. The proposed model has been evaluated using different metrics. It achieves excellent results, with classification accuracy reaching 99.54%, sensitivity at 99.09%, F1-score at 99.31, precision at 99.54, and specificity at 99.77%. The results confirm that the DaTscan images are useful for detecting the severity of PD and that a combination of deep learning and neuroimaging could prove an accurate and cost-effective approach to PD detection.</p>