With the prevalence of plastics exceeding over 368 million tons yearly, microplastic (MP) pollution has grown to an extent where air, water, soil, and living organisms have all been tested where microplastics have been detected. No less are these particles of size < 5 mm harmful to humans than all other components of the environment. Toxicity research on microplastics has shown that exposure to microplastics may cause liver infection, intestinal injuries, and floral imbalance, leading to many other potential health hazards. This paper presents a new model, the MicroDetect-Net (MDN), which applies fluorescence microscopy with Nile red-dye staining and deep learning to scan blood samples for microplastics. Although clam blood has certain limits in replicating real blood, the study opens avenues for this approach with human blood, with the samples being more consistent for preliminary data collection. The MDN model uses dataset preparation, fluorescence imaging, and segmentation using a convolutional neural network (CNN) to localize and count the microplastic fragments. The combination of CNN and Nile red dye for segmentation produced strong image detection and accuracy. MDN was evaluated on a dataset of 276 Nile red-stained fluorescent blood images and achieved an accuracy of 92%. Robust performance with an Intersection of Union (IoU) of 87.4%, F1 score of 92.1%, Precision of 90.6%, and Recall of 93.7% was observed. These metrics prove MDN’s effectiveness in the detection of microplastics.