DICOM (Digital Imaging and Communications in Medicine) is a standard format for medical images, often containing sensitive patient information. To protect patient privacy, anonymizing DICOM files before sharing or storage is essential, this work introduces a novel DICOM medical image anonymization approach that employs Optical Character Recognition (OCR) to identify personally identifiable information (PII) within both tags and textual content embedded within images. By employing Named Entity Recognition (NER) techniques, the system accurately detects PII in textual data. The proposed method operates independently of metadata, enabling the anonymization of sensitive information in both tags and pixel-level image content. This approach can help efficient and automated anonymization of medical image files, significantly enhancing patient privacy.