MRI MRI proves to be a very helpful tool for physicians outside tertiary referral centers. Remarkably many physicians assess MR images themselves. In many outpatient clinics however, neurologists are the leading specialty in a multidisciplinary team and usually are a member of this team when other specialties are in the lead. Consultants neurology are trained in assessing MRI, consultants geriatric medicine and psychiatry usually are not. The still modest experience of many radiologists assessing potential signs of neurodegenerative disease, may be a reason that the physicians’ assessment is frequently used to make the consensus diagnosis. It will probably be helpful for radiologists to assess MRI in a standardized way, reporting on all relevant aspects of neurodegenerative disease by using scales to visually rate atrophy and white matter lesions. MRI proved to be useful in our practice, increasing confidence by around 10% on a VAS, which reflects a clinically significant increase.1-9 In our study the indications of confidence on the VAS will beforehand not be normally distributed. Very seldom a physician will only have a 10 or 20% confidence in a pre- or post MRI assessment. A 10% change in the left half of this distribution represents in a psychological sense a smaller change than the same amount in the right half of the distribution. In this way the result of a 10% increase represents a clinically relevant change. CSF The CSF tool also proves to be helpful in the diagnostic process outside a tertiary referral center. Results in the 3.1 and 3.2 study are comparable, with better results in the second study with regard to the number of diagnoses changed having CSF results available, as well as regarding the measure of increase in certainty. This is probably due to methodology; in 3.1 only for one patient no diagnosis was available pre-CSF. In 3.2 in a considerable number of patients a diagnosis was suspended. Only after presentation of CSF results a diagnosis was made. This may be a coincidence, it may also be the result of the availability of multiple biomarkers resulting in a tendency to postpone a diagnosis when in doubt. Furthermore, the mean age of subjects in 3.1 is 72, versus 67.9 years in 3.2. The 3.1 mean age falls in the oldest age category of 3.2, showing the least typical CSF profiles that match the clinical diagnosis. Neuropsychological tests In 4.1 we show that it pays off to adapt a test protocol to the needs of more senior subjects. These subjects need more time but also more encouragement to ensure the best possible performance. This sounds very straightforward but in everyday practice due to a lack of staff and consequently time, this condition is quite a challenge to meet. Number of missing data using the 15WT may not be reduced in AD dementia patients, even when a strict adherence to the test protocol is used. This illustrates the ineligibility for use of this test in this group of subjects. An interesting finding of the 4.2 study using the VAT extended version, is that memory performance of aMCI and AD dementia patients is much better when measured in a recognition mode in comparison to the direct free recall mode. This is important; aMCI and mild AD patients can learn new material in a visual associative learning paradigm. This presents caregivers and therapeutic professionals with an important tool to improve daily function and communication. It also questions the typical profile of AD patients, showing typically poor results in free recall and recognition conditions. A diagnosis of AD may still be taken into consideration despite normal or near normal recognition scores.