
New recommendations from the French Society of Radiology (Societé Française de Radiologie, SFR) for storing and archiving medical images not only define good storing and archiving practice, but promise to save time and money for hospitals. The guidance clearly distinguishes between storage of modifiable image data for current and recent clinical cases, and subsequent nonmodifiable archive material. The guidance also recommends separate systems for data handling, as well as automatic storage and archiving protocols to optimize the processes, minimizing labor and lowering archiving costs.
"Hospitals in France produce many terabytes of data per year. This is a lot of archiving space and a huge financial burden. Our goal was to decrease this," said Dr. Daniel Reizine, neuroradiologist at Lariboisière Hospital (University Hospital Paris 7), PACS coordinator for Paris Public Hospitals (Assistance Publique - Hôpitaux de Paris [AP-HP]) and member of the SFR's informatics working group charged with creating the good practice recommendations.
Dr. Daniel Reizine is one of the people charged with creating good practice recommendations for archiving medical images.
The guidance is for use in the selection and conservation of digital medical images, the objective being to store and archive information useful to the diagnosis and follow-up of patients in line with legal requirements.
For all hospitals, whether private or public, medical files must be kept archived for 20 years from the date of the patient's last consultation in the establishment or until the patient turns 28 if the legal duration terminates before then. If the patient dies within ten years of the last visit to the hospital, the file is to be kept for ten years from the patient's death. Time limits are suspended for any legal cases taken against establishment or staff practice.
For independent doctors privately treating outpatients, France's Medical Council (Le Conseil de l'Ordre) specified in May 2009 that although no legal text fixed the duration of archiving, they align themselves with the requirement of 20 years demanded of all hospitals.
Legally, "storing" can be defined as any action, tool, or method that captures electronic data and that allows later treatment of these contents. The stored data is not "fixed," and can be modified, deleted, or replaced, which means the duration of storing is not defined by law.
Archiving, on the other hand, includes any action, tool, or method that captures, identifies, selects, classes, and preserves electronic data securely so that it cannot be falsified, and which guarantees the integrity of the information for either informing, or for serving as evidence in legal cases. Although accessible, this data is fixed and cannot be modified, therefore archiving is subject to a legal framework that includes the length of time data must be preserved.
For efficiency purposes, SFR recommends automatic storage protocols should be in place at the time of production, these adapted to modality and clinical indication. A CT scan can generate around 20,000 images, but not all should be sent to the PACS, according to the SFR informatics group. Such automatic selection would save radiologists time and eventually reduce the number of images to be archived.
In addition, storage systems should provide the option of a reversible compression (without loss) in line with DICOM Lossless specification. Authorized users such as a clinician or radiologist should benefit from immediate access to all exams in the storage system, which might be standalone or shared between different centers for primary or secondary interpretation. Furthermore, the SFR advises a minimum of three years of storage, before they are manually or automatically expedited to the archive.
Medical image archiving systems also may be standalone or shared between establishments, and may house several types of data, according to the SFR. They should ensure images be kept for the legal duration, in DICOM format with reversible (Lossless) or nonreversible (Lossy) compression. There should be a different means of access to all archived images and an option for archiving non-DICOM data. All access to archived data must be managed and signaled. The system may or may not be independent from other integrated systems that feed it; however, it should be totally independent from the hospital's PACS and must be able to guarantee the security and integrity of data for the entire length of its retention, according to the SFR.
"Archives should be independent from PACS so that if you change the PACS system you don't need to systematically change the archiving system," Reizine said. "Imaging data transferred to the archives should only be a portion of the entire production, this automatically selected through adequate archiving protocols."










![Overview of the study design. (A) The fully automated deep learning framework was developed to estimate body composition (BC) (defined as subcutaneous adipose tissue [SAT] in liters; visceral adipose tissue [VAT] in liters; skeletal muscle [SM] in liters; SM fat fraction [SMFF] as a percentage; and intramuscular adipose tissue [IMAT] in deciliters) from MRI. The fully automated framework comprised one model (model 1) to quantify different BC measures (SAT, VAT, SM, SMFF, and IMAT) as three-dimensional (3D) measures from whole-body MRI scans. The second model (model 2) was trained to identify standardized anatomic landmarks along the craniocaudal body axis (z coordinate field), which allowed for subdividing the whole-body measures into different subregions typically examined on clinical routine MRI scans (chest, abdomen, and pelvis). (B) BC was quantified from whole-body MRI in over 66,000 individuals from two large population-based cohort studies, the UK Biobank (UKB) (36,317 individuals) and the German National Cohort (NAKO) (30,291 individuals). Bar graphs show age distribution by sex and cohort. BMI = body mass index. (C) After the performance assessment of the fully automated framework, the change in BC measures, distributions, and profiles across age decades were investigated. Age-, sex-, and height-adjusted body composition reference curves were calculated and made publicly available in a web-based z-score calculator (https://circ-ml.github.io).](https://img.auntminnieeurope.com/mindful/smg/workspaces/default/uploads/2026/05/body-comp.XgAjTfPj1W.jpg?auto=format%2Ccompress&fit=crop&h=112&q=70&w=112)






