Full DICOM interoperability in a private dental imaging network: A 10-year longitudinal experience using open-source software Authors: Nielsen S. Pereira, Luis G. F. Santos Published: 2026 Dear Editors, Digital transformation in dental radiology often remains hindered by proprietary software ecosystems and isolated workstations. While medical imaging has long relied on the Digital Imaging and Communications in Medicine (DICOM) standard to ensure interoperability,1 dentistry continues to struggle with vendor lock-in and fragmented digital archives.23 To demonstrate that these barriers can be overcome, we present a 10-year longitudinal experience (2010–2020) of implementing a fully interoperable, exclusively open-source DICOM network in a private multi-site dental radiology practice in Rio de Janeiro, Brazil. Our infrastructure consisted of a central Picture Archiving and Communication System (PACS) deployed using DCM4CHEE45 running on a LAMP (Linux, Apache, MySQL, PHP) architecture. Local mini-PACS units were installed across six branch locations, performing automated nightly transfers to the central server via DICOM C-STORE. For remote interpretation, we utilized an OpenVPN-secured teleradiology setup. We strictly enforced a policy of acquiring only imaging equipment capable of native DICOM output, encompassing cone-beam computed tomography (CBCT), panoramic and cephalometric radiography, intraoral photostimulable phosphor (PSP) scanners, and clinical photography. Because many dental equipment manufacturers fail to provide seamless Health Level Seven (HL7) integration with legacy Radiology Information Systems (RIS)6, we developed a custom web-based Modality Worklist (MWL) using PHP and HL7 v2.x. We acknowledge that this implementation presents a high technical barrier, having required a senior in-house IT engineer with extensive experience in LAMP stack development, datacenter management, and secure network infrastructure. This critical middleware eliminated the need for manual demographic entry at the acquisition consoles, effectively preventing the generation of inconsistent metadata ("dirty data") that typically corrupts dental PACS archives. Over the 10-year period, this infrastructure processed referrals from approximately 27,400 dentists for 190,356 patients. The network successfully managed 3,101,377 DICOM files, totaling approximately 9.34 terabytes of data with zero downtime or data loss (Table 1). The only persistent limitation within the continuous DICOM chain was the proprietary nature of third-party cephalometric analysis software, which necessitated exporting 38,971 lateral cephalograms to non-DICOM formats (e.g., JPEG) for tracing. Beyond technical stability, the economic impact of adopting an open-source architecture was substantial. We estimate that avoiding proprietary PACS licensing fees, concurrent user charges, and annual maintenance contracts saved the network over US$ 200,000 during the decade. These resources were strategically redirected toward hardware upgrades, local 8-drive RAID-5 storage redundancy for immediate high availability, and a dedicated secondary server for automated daily backups to ensure robust disaster recovery. This longitudinal experience provides concrete evidence that the primary obstacle to achieving full digital interoperability in dentistry is neither technical nor financial. Open-source solutions are robust, highly scalable, and capable of managing millions of dental imaging files across distributed networks. Instead, the persistent fragmentation in dental imaging stems largely from commercial inertia and a lack of regulatory enforcement by professional councils regarding established standards. As the field looks toward the integration of Artificial Intelligence, which demands structured, standardized, and reproducible datasets, overcoming these interoperability barriers is no longer optional; it is the fundamental prerequisite for the next era of dental radiology.