To assess the quality of standard clinical radiological reports in patients with MS and to compare these with a neuroradiological review, and clinically cleared automated software.

MRI scans are paramount to assess disease activity and for therapeutic decision-making in MS patients. As MS is an inflammatory disease, the follow-up of lesion load is crucial.

New and/or enlarging lesion count was assessed between baseline and one-year follow-up MRI scans of 100 MS patients (RRMS or SPSMS) originating from two centers. Assessment was done through standard Clinical Radiological (CR) reports, a structured Neuroradiologist Review (NR), local automated measurements, and central automated measurements using the CE/FDA cleared software icobrain.

There was a strong correlation between local and central count of new lesions (center 1: R=0.81, center 2: R=0.54, all p<0.001) and new/enlarging lesions (center 1: R=0.7, center 2: R=0.73, all p<0.001).

Both CR and NR reports indicated the same 33% of patients having new or enlarging lesions, yet incomplete separate counts. The automated measured proportion was 34% locally and 45% using icobrain. 

CR and NR specified presence of enlarging lesions in only 5% of patients, while automatic measurements detected the more likely 23% (local) and 39% (icobrain) of patients.

CR/NR reported presence of new/enlarging lesions agreed well with automated classification of new lesions (local: 87%, icobrain: 80%) and new/enlarging lesions (local: 88%, icobrain: 78%).

Automated measurements of new/enlarging lesion count correlated strongly with the CR reported number (icobrain: R=0.69, p<0.001, local: R=0.46, p=0.02), yet stronger so with the NR reported number (icobrain: R=0.81, p<0.001, local: R=0.38, p=0.03).

Structured radiological reporting adds to the accuracy of lesion counts over clinical radiological reporting, yet separate counting of enlarging lesions is cumbersome. Automated techniques agree well with manual counts both in terms of classification (presence/absence) as in absolute number of new/enlarging lesions.