Images acquired from medical procedures like X-rays, MRIs, CT scans, and ultrasounds often come in sizable files. The compression of such images is instrumental in minimizing storage demands, enhancing the utilization of disk space, and simplifying the administration of extensive datasets. Sending substantial medical image files can be a time-intensive process, particularly in situations involving telemedicine or remote consultations. Compression expedites the swifter transmission of images across networks, rendering it more feasible for remote medical consultations and collaborative efforts. Utilizing deep learning neural networks for the compression of medical images is a widely adopted and successful method. The InceptionV3 architecture, known simply as Inception, is a renowned deep learning model recognized for its outstanding performance across diverse computer vision applications. A tailor-made inception deep learning framework is proposed in this research work for the compression of medical images. The validation of the compression model by the performance metrics reveals its proficiency. The outcome of this research work paves the way toward the proficient transformation of big data in the healthcare sector.